End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.2,
|
| 3 |
+
"eval_accuracy": 0.9929988331388565,
|
| 4 |
+
"eval_loss": 0.025693388655781746,
|
| 5 |
+
"eval_runtime": 981.9678,
|
| 6 |
+
"eval_samples_per_second": 0.873,
|
| 7 |
+
"eval_steps_per_second": 0.219
|
| 8 |
+
}
|
runs/Apr23_02-34-00_blackhorse/events.out.tfevents.1713846713.blackhorse.705231.4
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:91a4dacc5faaecd59deac09679479684329d95eedcfd79d8ca70a8173cdca464
|
| 3 |
+
size 734
|
test_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.2,
|
| 3 |
+
"eval_accuracy": 0.9929988331388565,
|
| 4 |
+
"eval_loss": 0.025693388655781746,
|
| 5 |
+
"eval_runtime": 981.9678,
|
| 6 |
+
"eval_samples_per_second": 0.873,
|
| 7 |
+
"eval_steps_per_second": 0.219
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3019 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9985528219971056,
|
| 3 |
+
"best_model_checkpoint": "videomae-base-finetuned-isl-numbers_aug/checkpoint-4180",
|
| 4 |
+
"epoch": 4.2,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4180,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0023923444976076554,
|
| 13 |
+
"grad_norm": 14.015558242797852,
|
| 14 |
+
"learning_rate": 1.1961722488038277e-06,
|
| 15 |
+
"loss": 1.36,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.004784688995215311,
|
| 20 |
+
"grad_norm": 13.480476379394531,
|
| 21 |
+
"learning_rate": 2.3923444976076554e-06,
|
| 22 |
+
"loss": 1.4805,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.007177033492822967,
|
| 27 |
+
"grad_norm": 24.253854751586914,
|
| 28 |
+
"learning_rate": 3.5885167464114835e-06,
|
| 29 |
+
"loss": 1.1355,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.009569377990430622,
|
| 34 |
+
"grad_norm": 24.031822204589844,
|
| 35 |
+
"learning_rate": 4.784688995215311e-06,
|
| 36 |
+
"loss": 1.2175,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.011961722488038277,
|
| 41 |
+
"grad_norm": 33.942161560058594,
|
| 42 |
+
"learning_rate": 5.980861244019139e-06,
|
| 43 |
+
"loss": 1.5014,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.014354066985645933,
|
| 48 |
+
"grad_norm": 25.417530059814453,
|
| 49 |
+
"learning_rate": 7.177033492822967e-06,
|
| 50 |
+
"loss": 1.0382,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.01674641148325359,
|
| 55 |
+
"grad_norm": 33.37525939941406,
|
| 56 |
+
"learning_rate": 8.373205741626795e-06,
|
| 57 |
+
"loss": 0.818,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.019138755980861243,
|
| 62 |
+
"grad_norm": 11.674866676330566,
|
| 63 |
+
"learning_rate": 9.569377990430622e-06,
|
| 64 |
+
"loss": 1.0039,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.0215311004784689,
|
| 69 |
+
"grad_norm": 29.53816795349121,
|
| 70 |
+
"learning_rate": 1.0765550239234451e-05,
|
| 71 |
+
"loss": 1.0634,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.023923444976076555,
|
| 76 |
+
"grad_norm": 32.427513122558594,
|
| 77 |
+
"learning_rate": 1.1961722488038278e-05,
|
| 78 |
+
"loss": 1.1812,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.02631578947368421,
|
| 83 |
+
"grad_norm": 47.19343566894531,
|
| 84 |
+
"learning_rate": 1.3157894736842106e-05,
|
| 85 |
+
"loss": 0.9474,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.028708133971291867,
|
| 90 |
+
"grad_norm": 21.350784301757812,
|
| 91 |
+
"learning_rate": 1.4354066985645934e-05,
|
| 92 |
+
"loss": 0.7303,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.03110047846889952,
|
| 97 |
+
"grad_norm": 27.72856903076172,
|
| 98 |
+
"learning_rate": 1.555023923444976e-05,
|
| 99 |
+
"loss": 0.7567,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.03349282296650718,
|
| 104 |
+
"grad_norm": 24.274192810058594,
|
| 105 |
+
"learning_rate": 1.674641148325359e-05,
|
| 106 |
+
"loss": 0.5606,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.03588516746411483,
|
| 111 |
+
"grad_norm": 52.520484924316406,
|
| 112 |
+
"learning_rate": 1.7942583732057417e-05,
|
| 113 |
+
"loss": 0.718,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.03827751196172249,
|
| 118 |
+
"grad_norm": 37.529579162597656,
|
| 119 |
+
"learning_rate": 1.9138755980861243e-05,
|
| 120 |
+
"loss": 0.8521,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.04066985645933014,
|
| 125 |
+
"grad_norm": 4.263017177581787,
|
| 126 |
+
"learning_rate": 2.0334928229665073e-05,
|
| 127 |
+
"loss": 0.6431,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.0430622009569378,
|
| 132 |
+
"grad_norm": 15.213534355163574,
|
| 133 |
+
"learning_rate": 2.1531100478468903e-05,
|
| 134 |
+
"loss": 0.7901,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.045454545454545456,
|
| 139 |
+
"grad_norm": 31.36748504638672,
|
| 140 |
+
"learning_rate": 2.272727272727273e-05,
|
| 141 |
+
"loss": 0.6664,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.04784688995215311,
|
| 146 |
+
"grad_norm": 16.556751251220703,
|
| 147 |
+
"learning_rate": 2.3923444976076556e-05,
|
| 148 |
+
"loss": 0.5235,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.050239234449760764,
|
| 153 |
+
"grad_norm": 46.87775421142578,
|
| 154 |
+
"learning_rate": 2.5119617224880382e-05,
|
| 155 |
+
"loss": 0.8653,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.05263157894736842,
|
| 160 |
+
"grad_norm": 26.508085250854492,
|
| 161 |
+
"learning_rate": 2.6315789473684212e-05,
|
| 162 |
+
"loss": 0.8489,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.05502392344497608,
|
| 167 |
+
"grad_norm": 25.13356590270996,
|
| 168 |
+
"learning_rate": 2.751196172248804e-05,
|
| 169 |
+
"loss": 0.6788,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.05741626794258373,
|
| 174 |
+
"grad_norm": 31.49532699584961,
|
| 175 |
+
"learning_rate": 2.8708133971291868e-05,
|
| 176 |
+
"loss": 0.9036,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.05980861244019139,
|
| 181 |
+
"grad_norm": 42.16746520996094,
|
| 182 |
+
"learning_rate": 2.9904306220095695e-05,
|
| 183 |
+
"loss": 0.9568,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.06220095693779904,
|
| 188 |
+
"grad_norm": 22.292442321777344,
|
| 189 |
+
"learning_rate": 3.110047846889952e-05,
|
| 190 |
+
"loss": 0.3915,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.0645933014354067,
|
| 195 |
+
"grad_norm": 31.647417068481445,
|
| 196 |
+
"learning_rate": 3.229665071770335e-05,
|
| 197 |
+
"loss": 0.8292,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.06698564593301436,
|
| 202 |
+
"grad_norm": 31.727983474731445,
|
| 203 |
+
"learning_rate": 3.349282296650718e-05,
|
| 204 |
+
"loss": 0.8209,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.06937799043062201,
|
| 209 |
+
"grad_norm": 24.073183059692383,
|
| 210 |
+
"learning_rate": 3.4688995215311004e-05,
|
| 211 |
+
"loss": 0.4971,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.07177033492822966,
|
| 216 |
+
"grad_norm": 19.445539474487305,
|
| 217 |
+
"learning_rate": 3.5885167464114834e-05,
|
| 218 |
+
"loss": 0.7874,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.07416267942583732,
|
| 223 |
+
"grad_norm": 19.681711196899414,
|
| 224 |
+
"learning_rate": 3.7081339712918663e-05,
|
| 225 |
+
"loss": 0.6186,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.07655502392344497,
|
| 230 |
+
"grad_norm": 19.926891326904297,
|
| 231 |
+
"learning_rate": 3.8277511961722486e-05,
|
| 232 |
+
"loss": 0.5922,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.07894736842105263,
|
| 237 |
+
"grad_norm": 26.6019229888916,
|
| 238 |
+
"learning_rate": 3.9473684210526316e-05,
|
| 239 |
+
"loss": 0.4758,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.08133971291866028,
|
| 244 |
+
"grad_norm": 29.542890548706055,
|
| 245 |
+
"learning_rate": 4.0669856459330146e-05,
|
| 246 |
+
"loss": 0.6328,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.08373205741626795,
|
| 251 |
+
"grad_norm": 27.611955642700195,
|
| 252 |
+
"learning_rate": 4.1866028708133976e-05,
|
| 253 |
+
"loss": 0.8429,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.0861244019138756,
|
| 258 |
+
"grad_norm": 26.79853057861328,
|
| 259 |
+
"learning_rate": 4.3062200956937806e-05,
|
| 260 |
+
"loss": 0.7488,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.08851674641148326,
|
| 265 |
+
"grad_norm": 20.399744033813477,
|
| 266 |
+
"learning_rate": 4.425837320574163e-05,
|
| 267 |
+
"loss": 0.9143,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.09090909090909091,
|
| 272 |
+
"grad_norm": 31.50543785095215,
|
| 273 |
+
"learning_rate": 4.545454545454546e-05,
|
| 274 |
+
"loss": 0.9285,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.09330143540669857,
|
| 279 |
+
"grad_norm": 12.433365821838379,
|
| 280 |
+
"learning_rate": 4.665071770334928e-05,
|
| 281 |
+
"loss": 0.729,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.09569377990430622,
|
| 286 |
+
"grad_norm": 5.882745742797852,
|
| 287 |
+
"learning_rate": 4.784688995215311e-05,
|
| 288 |
+
"loss": 0.6221,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.09808612440191387,
|
| 293 |
+
"grad_norm": 47.791988372802734,
|
| 294 |
+
"learning_rate": 4.904306220095694e-05,
|
| 295 |
+
"loss": 0.697,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.10047846889952153,
|
| 300 |
+
"grad_norm": 32.664573669433594,
|
| 301 |
+
"learning_rate": 4.997341839447103e-05,
|
| 302 |
+
"loss": 0.6012,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.10287081339712918,
|
| 307 |
+
"grad_norm": 5.583775997161865,
|
| 308 |
+
"learning_rate": 4.984051036682616e-05,
|
| 309 |
+
"loss": 0.4423,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.10526315789473684,
|
| 314 |
+
"grad_norm": 19.180892944335938,
|
| 315 |
+
"learning_rate": 4.970760233918128e-05,
|
| 316 |
+
"loss": 0.7457,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.1076555023923445,
|
| 321 |
+
"grad_norm": 16.862192153930664,
|
| 322 |
+
"learning_rate": 4.9574694311536416e-05,
|
| 323 |
+
"loss": 0.8563,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.11004784688995216,
|
| 328 |
+
"grad_norm": 36.542415618896484,
|
| 329 |
+
"learning_rate": 4.944178628389155e-05,
|
| 330 |
+
"loss": 0.6835,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.11244019138755981,
|
| 335 |
+
"grad_norm": 26.89089584350586,
|
| 336 |
+
"learning_rate": 4.930887825624668e-05,
|
| 337 |
+
"loss": 0.5772,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.11483253588516747,
|
| 342 |
+
"grad_norm": 64.69689178466797,
|
| 343 |
+
"learning_rate": 4.917597022860181e-05,
|
| 344 |
+
"loss": 0.6574,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.11722488038277512,
|
| 349 |
+
"grad_norm": 32.71666717529297,
|
| 350 |
+
"learning_rate": 4.904306220095694e-05,
|
| 351 |
+
"loss": 0.7411,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.11961722488038277,
|
| 356 |
+
"grad_norm": 28.70256805419922,
|
| 357 |
+
"learning_rate": 4.8910154173312074e-05,
|
| 358 |
+
"loss": 0.7735,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.12200956937799043,
|
| 363 |
+
"grad_norm": 47.490455627441406,
|
| 364 |
+
"learning_rate": 4.87772461456672e-05,
|
| 365 |
+
"loss": 0.9173,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.12440191387559808,
|
| 370 |
+
"grad_norm": 0.40462446212768555,
|
| 371 |
+
"learning_rate": 4.8644338118022334e-05,
|
| 372 |
+
"loss": 0.3499,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.12679425837320574,
|
| 377 |
+
"grad_norm": 13.079553604125977,
|
| 378 |
+
"learning_rate": 4.8511430090377467e-05,
|
| 379 |
+
"loss": 0.6362,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.1291866028708134,
|
| 384 |
+
"grad_norm": 16.84446907043457,
|
| 385 |
+
"learning_rate": 4.837852206273259e-05,
|
| 386 |
+
"loss": 0.6439,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.13157894736842105,
|
| 391 |
+
"grad_norm": 1.5127249956130981,
|
| 392 |
+
"learning_rate": 4.824561403508772e-05,
|
| 393 |
+
"loss": 0.688,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.1339712918660287,
|
| 398 |
+
"grad_norm": 5.6877923011779785,
|
| 399 |
+
"learning_rate": 4.811270600744285e-05,
|
| 400 |
+
"loss": 0.4104,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.13636363636363635,
|
| 405 |
+
"grad_norm": 12.39535140991211,
|
| 406 |
+
"learning_rate": 4.797979797979798e-05,
|
| 407 |
+
"loss": 0.5339,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.13875598086124402,
|
| 412 |
+
"grad_norm": 55.42013168334961,
|
| 413 |
+
"learning_rate": 4.784688995215311e-05,
|
| 414 |
+
"loss": 0.4111,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.14114832535885166,
|
| 419 |
+
"grad_norm": 26.363779067993164,
|
| 420 |
+
"learning_rate": 4.7713981924508244e-05,
|
| 421 |
+
"loss": 0.8168,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.14354066985645933,
|
| 426 |
+
"grad_norm": 22.93970489501953,
|
| 427 |
+
"learning_rate": 4.758107389686337e-05,
|
| 428 |
+
"loss": 0.6611,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.145933014354067,
|
| 433 |
+
"grad_norm": 17.63263511657715,
|
| 434 |
+
"learning_rate": 4.7448165869218504e-05,
|
| 435 |
+
"loss": 0.6952,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.14832535885167464,
|
| 440 |
+
"grad_norm": 42.79911804199219,
|
| 441 |
+
"learning_rate": 4.731525784157364e-05,
|
| 442 |
+
"loss": 0.6323,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.1507177033492823,
|
| 447 |
+
"grad_norm": 52.291927337646484,
|
| 448 |
+
"learning_rate": 4.718234981392876e-05,
|
| 449 |
+
"loss": 0.5826,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.15311004784688995,
|
| 454 |
+
"grad_norm": 5.49782133102417,
|
| 455 |
+
"learning_rate": 4.7049441786283896e-05,
|
| 456 |
+
"loss": 0.6368,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.15550239234449761,
|
| 461 |
+
"grad_norm": 24.484716415405273,
|
| 462 |
+
"learning_rate": 4.691653375863902e-05,
|
| 463 |
+
"loss": 0.767,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.15789473684210525,
|
| 468 |
+
"grad_norm": 43.46366882324219,
|
| 469 |
+
"learning_rate": 4.678362573099415e-05,
|
| 470 |
+
"loss": 0.6198,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.16028708133971292,
|
| 475 |
+
"grad_norm": 0.9531241059303284,
|
| 476 |
+
"learning_rate": 4.665071770334928e-05,
|
| 477 |
+
"loss": 0.3393,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.16267942583732056,
|
| 482 |
+
"grad_norm": 23.464101791381836,
|
| 483 |
+
"learning_rate": 4.6517809675704415e-05,
|
| 484 |
+
"loss": 0.3039,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.16507177033492823,
|
| 489 |
+
"grad_norm": 61.715240478515625,
|
| 490 |
+
"learning_rate": 4.638490164805955e-05,
|
| 491 |
+
"loss": 0.635,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.1674641148325359,
|
| 496 |
+
"grad_norm": 4.636231422424316,
|
| 497 |
+
"learning_rate": 4.6251993620414674e-05,
|
| 498 |
+
"loss": 0.4028,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.16985645933014354,
|
| 503 |
+
"grad_norm": 19.922252655029297,
|
| 504 |
+
"learning_rate": 4.611908559276981e-05,
|
| 505 |
+
"loss": 0.4871,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.1722488038277512,
|
| 510 |
+
"grad_norm": 25.78838348388672,
|
| 511 |
+
"learning_rate": 4.598617756512494e-05,
|
| 512 |
+
"loss": 0.7949,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.17464114832535885,
|
| 517 |
+
"grad_norm": 8.514424324035645,
|
| 518 |
+
"learning_rate": 4.5853269537480066e-05,
|
| 519 |
+
"loss": 0.6953,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.17703349282296652,
|
| 524 |
+
"grad_norm": 44.999107360839844,
|
| 525 |
+
"learning_rate": 4.57203615098352e-05,
|
| 526 |
+
"loss": 0.7568,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.17942583732057416,
|
| 531 |
+
"grad_norm": 33.99626541137695,
|
| 532 |
+
"learning_rate": 4.5587453482190326e-05,
|
| 533 |
+
"loss": 0.3274,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.18181818181818182,
|
| 538 |
+
"grad_norm": 1.6210479736328125,
|
| 539 |
+
"learning_rate": 4.545454545454546e-05,
|
| 540 |
+
"loss": 0.2306,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.18421052631578946,
|
| 545 |
+
"grad_norm": 6.679223537445068,
|
| 546 |
+
"learning_rate": 4.5321637426900585e-05,
|
| 547 |
+
"loss": 0.6284,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.18660287081339713,
|
| 552 |
+
"grad_norm": 43.11864471435547,
|
| 553 |
+
"learning_rate": 4.518872939925572e-05,
|
| 554 |
+
"loss": 0.9184,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.18899521531100477,
|
| 559 |
+
"grad_norm": 39.860008239746094,
|
| 560 |
+
"learning_rate": 4.5055821371610844e-05,
|
| 561 |
+
"loss": 0.3346,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.19138755980861244,
|
| 566 |
+
"grad_norm": 58.848724365234375,
|
| 567 |
+
"learning_rate": 4.492291334396598e-05,
|
| 568 |
+
"loss": 0.7516,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.1937799043062201,
|
| 573 |
+
"grad_norm": 1.0923501253128052,
|
| 574 |
+
"learning_rate": 4.479000531632111e-05,
|
| 575 |
+
"loss": 0.5148,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.19617224880382775,
|
| 580 |
+
"grad_norm": 24.75638198852539,
|
| 581 |
+
"learning_rate": 4.4657097288676236e-05,
|
| 582 |
+
"loss": 0.5507,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.19856459330143542,
|
| 587 |
+
"grad_norm": 0.345415860414505,
|
| 588 |
+
"learning_rate": 4.452418926103137e-05,
|
| 589 |
+
"loss": 0.1649,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.2,
|
| 594 |
+
"eval_accuracy": 0.8552821997105644,
|
| 595 |
+
"eval_loss": 0.4412595331668854,
|
| 596 |
+
"eval_runtime": 761.629,
|
| 597 |
+
"eval_samples_per_second": 0.907,
|
| 598 |
+
"eval_steps_per_second": 0.227,
|
| 599 |
+
"step": 836
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.000956937799043,
|
| 603 |
+
"grad_norm": 55.6067008972168,
|
| 604 |
+
"learning_rate": 4.43912812333865e-05,
|
| 605 |
+
"loss": 0.3236,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.0033492822966508,
|
| 610 |
+
"grad_norm": 0.7264971733093262,
|
| 611 |
+
"learning_rate": 4.425837320574163e-05,
|
| 612 |
+
"loss": 0.3854,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.0057416267942583,
|
| 617 |
+
"grad_norm": 0.12801001965999603,
|
| 618 |
+
"learning_rate": 4.412546517809676e-05,
|
| 619 |
+
"loss": 0.2788,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.008133971291866,
|
| 624 |
+
"grad_norm": 29.689430236816406,
|
| 625 |
+
"learning_rate": 4.399255715045189e-05,
|
| 626 |
+
"loss": 0.5362,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.0105263157894737,
|
| 631 |
+
"grad_norm": 3.776660442352295,
|
| 632 |
+
"learning_rate": 4.3859649122807014e-05,
|
| 633 |
+
"loss": 0.2869,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.0129186602870814,
|
| 638 |
+
"grad_norm": 0.310320645570755,
|
| 639 |
+
"learning_rate": 4.372674109516215e-05,
|
| 640 |
+
"loss": 0.3163,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.015311004784689,
|
| 645 |
+
"grad_norm": 25.65765953063965,
|
| 646 |
+
"learning_rate": 4.359383306751728e-05,
|
| 647 |
+
"loss": 0.1512,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.0177033492822967,
|
| 652 |
+
"grad_norm": 68.732666015625,
|
| 653 |
+
"learning_rate": 4.346092503987241e-05,
|
| 654 |
+
"loss": 0.7086,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.0200956937799044,
|
| 659 |
+
"grad_norm": 18.452295303344727,
|
| 660 |
+
"learning_rate": 4.332801701222754e-05,
|
| 661 |
+
"loss": 0.3176,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.022488038277512,
|
| 666 |
+
"grad_norm": 1.5882939100265503,
|
| 667 |
+
"learning_rate": 4.319510898458267e-05,
|
| 668 |
+
"loss": 0.3374,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.0248803827751196,
|
| 673 |
+
"grad_norm": 22.042537689208984,
|
| 674 |
+
"learning_rate": 4.3062200956937806e-05,
|
| 675 |
+
"loss": 0.1728,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.0272727272727273,
|
| 680 |
+
"grad_norm": 1.0320727825164795,
|
| 681 |
+
"learning_rate": 4.292929292929293e-05,
|
| 682 |
+
"loss": 0.2954,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.0296650717703348,
|
| 687 |
+
"grad_norm": 27.908615112304688,
|
| 688 |
+
"learning_rate": 4.2796384901648065e-05,
|
| 689 |
+
"loss": 0.1296,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.0320574162679426,
|
| 694 |
+
"grad_norm": 39.791996002197266,
|
| 695 |
+
"learning_rate": 4.266347687400319e-05,
|
| 696 |
+
"loss": 0.0841,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.0344497607655503,
|
| 701 |
+
"grad_norm": 1.0737611055374146,
|
| 702 |
+
"learning_rate": 4.253056884635832e-05,
|
| 703 |
+
"loss": 0.4375,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.0368421052631578,
|
| 708 |
+
"grad_norm": 40.09695053100586,
|
| 709 |
+
"learning_rate": 4.239766081871345e-05,
|
| 710 |
+
"loss": 0.2464,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.0392344497607655,
|
| 715 |
+
"grad_norm": 5.695016384124756,
|
| 716 |
+
"learning_rate": 4.2264752791068584e-05,
|
| 717 |
+
"loss": 0.3341,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.0416267942583732,
|
| 722 |
+
"grad_norm": 12.370247840881348,
|
| 723 |
+
"learning_rate": 4.213184476342371e-05,
|
| 724 |
+
"loss": 0.3143,
|
| 725 |
+
"step": 1010
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.044019138755981,
|
| 729 |
+
"grad_norm": 0.9093998670578003,
|
| 730 |
+
"learning_rate": 4.199893673577884e-05,
|
| 731 |
+
"loss": 0.2729,
|
| 732 |
+
"step": 1020
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.0464114832535885,
|
| 736 |
+
"grad_norm": 7.610515117645264,
|
| 737 |
+
"learning_rate": 4.1866028708133976e-05,
|
| 738 |
+
"loss": 0.1896,
|
| 739 |
+
"step": 1030
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.0488038277511962,
|
| 743 |
+
"grad_norm": 7.180859088897705,
|
| 744 |
+
"learning_rate": 4.17331206804891e-05,
|
| 745 |
+
"loss": 0.4265,
|
| 746 |
+
"step": 1040
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.051196172248804,
|
| 750 |
+
"grad_norm": 0.1555383950471878,
|
| 751 |
+
"learning_rate": 4.1600212652844235e-05,
|
| 752 |
+
"loss": 0.0456,
|
| 753 |
+
"step": 1050
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.0535885167464114,
|
| 757 |
+
"grad_norm": 0.6277172565460205,
|
| 758 |
+
"learning_rate": 4.146730462519937e-05,
|
| 759 |
+
"loss": 0.3013,
|
| 760 |
+
"step": 1060
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.0559808612440191,
|
| 764 |
+
"grad_norm": 26.010774612426758,
|
| 765 |
+
"learning_rate": 4.1334396597554494e-05,
|
| 766 |
+
"loss": 0.1184,
|
| 767 |
+
"step": 1070
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.0583732057416269,
|
| 771 |
+
"grad_norm": 20.830724716186523,
|
| 772 |
+
"learning_rate": 4.120148856990963e-05,
|
| 773 |
+
"loss": 0.2435,
|
| 774 |
+
"step": 1080
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.0607655502392344,
|
| 778 |
+
"grad_norm": 43.08734893798828,
|
| 779 |
+
"learning_rate": 4.1068580542264754e-05,
|
| 780 |
+
"loss": 0.1626,
|
| 781 |
+
"step": 1090
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.063157894736842,
|
| 785 |
+
"grad_norm": 1.0129386186599731,
|
| 786 |
+
"learning_rate": 4.093567251461988e-05,
|
| 787 |
+
"loss": 0.1131,
|
| 788 |
+
"step": 1100
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.0655502392344498,
|
| 792 |
+
"grad_norm": 74.8167495727539,
|
| 793 |
+
"learning_rate": 4.080276448697501e-05,
|
| 794 |
+
"loss": 0.3853,
|
| 795 |
+
"step": 1110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.0679425837320573,
|
| 799 |
+
"grad_norm": 0.1025879979133606,
|
| 800 |
+
"learning_rate": 4.0669856459330146e-05,
|
| 801 |
+
"loss": 0.1161,
|
| 802 |
+
"step": 1120
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.070334928229665,
|
| 806 |
+
"grad_norm": 37.73963165283203,
|
| 807 |
+
"learning_rate": 4.053694843168527e-05,
|
| 808 |
+
"loss": 0.4214,
|
| 809 |
+
"step": 1130
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.0727272727272728,
|
| 813 |
+
"grad_norm": 0.028422629460692406,
|
| 814 |
+
"learning_rate": 4.0404040404040405e-05,
|
| 815 |
+
"loss": 0.2753,
|
| 816 |
+
"step": 1140
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.0751196172248805,
|
| 820 |
+
"grad_norm": 71.2274398803711,
|
| 821 |
+
"learning_rate": 4.027113237639554e-05,
|
| 822 |
+
"loss": 0.2009,
|
| 823 |
+
"step": 1150
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.077511961722488,
|
| 827 |
+
"grad_norm": 39.377342224121094,
|
| 828 |
+
"learning_rate": 4.0138224348750665e-05,
|
| 829 |
+
"loss": 0.3117,
|
| 830 |
+
"step": 1160
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.0799043062200957,
|
| 834 |
+
"grad_norm": 0.112571582198143,
|
| 835 |
+
"learning_rate": 4.00053163211058e-05,
|
| 836 |
+
"loss": 0.1175,
|
| 837 |
+
"step": 1170
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.0822966507177034,
|
| 841 |
+
"grad_norm": 25.220449447631836,
|
| 842 |
+
"learning_rate": 3.987240829346093e-05,
|
| 843 |
+
"loss": 0.3254,
|
| 844 |
+
"step": 1180
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.084688995215311,
|
| 848 |
+
"grad_norm": 50.95032501220703,
|
| 849 |
+
"learning_rate": 3.973950026581606e-05,
|
| 850 |
+
"loss": 0.2824,
|
| 851 |
+
"step": 1190
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.0870813397129186,
|
| 855 |
+
"grad_norm": 24.743955612182617,
|
| 856 |
+
"learning_rate": 3.960659223817118e-05,
|
| 857 |
+
"loss": 0.2875,
|
| 858 |
+
"step": 1200
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.0894736842105264,
|
| 862 |
+
"grad_norm": 57.72587966918945,
|
| 863 |
+
"learning_rate": 3.9473684210526316e-05,
|
| 864 |
+
"loss": 0.2018,
|
| 865 |
+
"step": 1210
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.0918660287081339,
|
| 869 |
+
"grad_norm": 0.6570279598236084,
|
| 870 |
+
"learning_rate": 3.934077618288145e-05,
|
| 871 |
+
"loss": 0.4795,
|
| 872 |
+
"step": 1220
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.0942583732057416,
|
| 876 |
+
"grad_norm": 2.577521324157715,
|
| 877 |
+
"learning_rate": 3.9207868155236576e-05,
|
| 878 |
+
"loss": 0.117,
|
| 879 |
+
"step": 1230
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.0966507177033493,
|
| 883 |
+
"grad_norm": 0.06854643672704697,
|
| 884 |
+
"learning_rate": 3.907496012759171e-05,
|
| 885 |
+
"loss": 0.209,
|
| 886 |
+
"step": 1240
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.0990430622009568,
|
| 890 |
+
"grad_norm": 0.17914560437202454,
|
| 891 |
+
"learning_rate": 3.894205209994684e-05,
|
| 892 |
+
"loss": 0.5758,
|
| 893 |
+
"step": 1250
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.1014354066985645,
|
| 897 |
+
"grad_norm": 6.280195236206055,
|
| 898 |
+
"learning_rate": 3.880914407230197e-05,
|
| 899 |
+
"loss": 0.1779,
|
| 900 |
+
"step": 1260
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.1038277511961723,
|
| 904 |
+
"grad_norm": 1.3312019109725952,
|
| 905 |
+
"learning_rate": 3.86762360446571e-05,
|
| 906 |
+
"loss": 0.1311,
|
| 907 |
+
"step": 1270
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.10622009569378,
|
| 911 |
+
"grad_norm": 9.676103591918945,
|
| 912 |
+
"learning_rate": 3.8543328017012234e-05,
|
| 913 |
+
"loss": 0.1207,
|
| 914 |
+
"step": 1280
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.1086124401913875,
|
| 918 |
+
"grad_norm": 104.29061889648438,
|
| 919 |
+
"learning_rate": 3.841041998936736e-05,
|
| 920 |
+
"loss": 0.3504,
|
| 921 |
+
"step": 1290
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.1110047846889952,
|
| 925 |
+
"grad_norm": 0.15878012776374817,
|
| 926 |
+
"learning_rate": 3.8277511961722486e-05,
|
| 927 |
+
"loss": 0.1365,
|
| 928 |
+
"step": 1300
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.113397129186603,
|
| 932 |
+
"grad_norm": 0.07822709530591965,
|
| 933 |
+
"learning_rate": 3.814460393407762e-05,
|
| 934 |
+
"loss": 0.4698,
|
| 935 |
+
"step": 1310
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.1157894736842104,
|
| 939 |
+
"grad_norm": 0.5035278797149658,
|
| 940 |
+
"learning_rate": 3.8011695906432746e-05,
|
| 941 |
+
"loss": 0.2533,
|
| 942 |
+
"step": 1320
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.1181818181818182,
|
| 946 |
+
"grad_norm": 35.283233642578125,
|
| 947 |
+
"learning_rate": 3.787878787878788e-05,
|
| 948 |
+
"loss": 0.3578,
|
| 949 |
+
"step": 1330
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.120574162679426,
|
| 953 |
+
"grad_norm": 55.772911071777344,
|
| 954 |
+
"learning_rate": 3.774587985114301e-05,
|
| 955 |
+
"loss": 0.3235,
|
| 956 |
+
"step": 1340
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.1229665071770334,
|
| 960 |
+
"grad_norm": 0.025184158235788345,
|
| 961 |
+
"learning_rate": 3.761297182349814e-05,
|
| 962 |
+
"loss": 0.0543,
|
| 963 |
+
"step": 1350
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.1253588516746411,
|
| 967 |
+
"grad_norm": 0.04470227286219597,
|
| 968 |
+
"learning_rate": 3.748006379585327e-05,
|
| 969 |
+
"loss": 0.0378,
|
| 970 |
+
"step": 1360
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.1277511961722488,
|
| 974 |
+
"grad_norm": 0.10474701225757599,
|
| 975 |
+
"learning_rate": 3.7347155768208404e-05,
|
| 976 |
+
"loss": 0.2598,
|
| 977 |
+
"step": 1370
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.1301435406698563,
|
| 981 |
+
"grad_norm": 0.1666097193956375,
|
| 982 |
+
"learning_rate": 3.721424774056353e-05,
|
| 983 |
+
"loss": 0.3193,
|
| 984 |
+
"step": 1380
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.132535885167464,
|
| 988 |
+
"grad_norm": 0.2347351312637329,
|
| 989 |
+
"learning_rate": 3.7081339712918663e-05,
|
| 990 |
+
"loss": 0.1546,
|
| 991 |
+
"step": 1390
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.1349282296650718,
|
| 995 |
+
"grad_norm": 2.922492027282715,
|
| 996 |
+
"learning_rate": 3.6948431685273796e-05,
|
| 997 |
+
"loss": 0.0459,
|
| 998 |
+
"step": 1400
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.1373205741626795,
|
| 1002 |
+
"grad_norm": 90.11852264404297,
|
| 1003 |
+
"learning_rate": 3.681552365762892e-05,
|
| 1004 |
+
"loss": 0.1136,
|
| 1005 |
+
"step": 1410
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.139712918660287,
|
| 1009 |
+
"grad_norm": 0.0664602667093277,
|
| 1010 |
+
"learning_rate": 3.668261562998405e-05,
|
| 1011 |
+
"loss": 0.1917,
|
| 1012 |
+
"step": 1420
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.1421052631578947,
|
| 1016 |
+
"grad_norm": 28.271459579467773,
|
| 1017 |
+
"learning_rate": 3.654970760233918e-05,
|
| 1018 |
+
"loss": 0.0633,
|
| 1019 |
+
"step": 1430
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.1444976076555025,
|
| 1023 |
+
"grad_norm": 48.338539123535156,
|
| 1024 |
+
"learning_rate": 3.6416799574694315e-05,
|
| 1025 |
+
"loss": 0.1383,
|
| 1026 |
+
"step": 1440
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.14688995215311,
|
| 1030 |
+
"grad_norm": 1.0220551490783691,
|
| 1031 |
+
"learning_rate": 3.628389154704944e-05,
|
| 1032 |
+
"loss": 0.2675,
|
| 1033 |
+
"step": 1450
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.1492822966507177,
|
| 1037 |
+
"grad_norm": 0.07435596734285355,
|
| 1038 |
+
"learning_rate": 3.6150983519404574e-05,
|
| 1039 |
+
"loss": 0.0058,
|
| 1040 |
+
"step": 1460
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.1516746411483254,
|
| 1044 |
+
"grad_norm": 0.025288909673690796,
|
| 1045 |
+
"learning_rate": 3.601807549175971e-05,
|
| 1046 |
+
"loss": 0.3105,
|
| 1047 |
+
"step": 1470
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.1540669856459331,
|
| 1051 |
+
"grad_norm": 3.0553793907165527,
|
| 1052 |
+
"learning_rate": 3.5885167464114834e-05,
|
| 1053 |
+
"loss": 0.1359,
|
| 1054 |
+
"step": 1480
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.1564593301435406,
|
| 1058 |
+
"grad_norm": 12.173036575317383,
|
| 1059 |
+
"learning_rate": 3.5752259436469967e-05,
|
| 1060 |
+
"loss": 0.1923,
|
| 1061 |
+
"step": 1490
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.1588516746411484,
|
| 1065 |
+
"grad_norm": 0.060364119708538055,
|
| 1066 |
+
"learning_rate": 3.56193514088251e-05,
|
| 1067 |
+
"loss": 0.0117,
|
| 1068 |
+
"step": 1500
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.1612440191387559,
|
| 1072 |
+
"grad_norm": 0.03194255381822586,
|
| 1073 |
+
"learning_rate": 3.5486443381180226e-05,
|
| 1074 |
+
"loss": 0.004,
|
| 1075 |
+
"step": 1510
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.1636363636363636,
|
| 1079 |
+
"grad_norm": 81.4512710571289,
|
| 1080 |
+
"learning_rate": 3.535353535353535e-05,
|
| 1081 |
+
"loss": 0.0776,
|
| 1082 |
+
"step": 1520
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.1660287081339713,
|
| 1086 |
+
"grad_norm": 0.016529429703950882,
|
| 1087 |
+
"learning_rate": 3.5220627325890485e-05,
|
| 1088 |
+
"loss": 0.3877,
|
| 1089 |
+
"step": 1530
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.168421052631579,
|
| 1093 |
+
"grad_norm": 0.6734187602996826,
|
| 1094 |
+
"learning_rate": 3.508771929824561e-05,
|
| 1095 |
+
"loss": 0.1606,
|
| 1096 |
+
"step": 1540
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 1.1708133971291865,
|
| 1100 |
+
"grad_norm": 0.019191304221749306,
|
| 1101 |
+
"learning_rate": 3.4954811270600744e-05,
|
| 1102 |
+
"loss": 0.1287,
|
| 1103 |
+
"step": 1550
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.1732057416267943,
|
| 1107 |
+
"grad_norm": 53.84104919433594,
|
| 1108 |
+
"learning_rate": 3.482190324295588e-05,
|
| 1109 |
+
"loss": 0.1832,
|
| 1110 |
+
"step": 1560
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.175598086124402,
|
| 1114 |
+
"grad_norm": 30.083900451660156,
|
| 1115 |
+
"learning_rate": 3.4688995215311004e-05,
|
| 1116 |
+
"loss": 0.3451,
|
| 1117 |
+
"step": 1570
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 1.1779904306220095,
|
| 1121 |
+
"grad_norm": 9.459929466247559,
|
| 1122 |
+
"learning_rate": 3.455608718766614e-05,
|
| 1123 |
+
"loss": 0.5572,
|
| 1124 |
+
"step": 1580
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.1803827751196172,
|
| 1128 |
+
"grad_norm": 0.8345646262168884,
|
| 1129 |
+
"learning_rate": 3.442317916002127e-05,
|
| 1130 |
+
"loss": 0.0222,
|
| 1131 |
+
"step": 1590
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.182775119617225,
|
| 1135 |
+
"grad_norm": 29.698089599609375,
|
| 1136 |
+
"learning_rate": 3.4290271132376396e-05,
|
| 1137 |
+
"loss": 0.1476,
|
| 1138 |
+
"step": 1600
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 1.1851674641148326,
|
| 1142 |
+
"grad_norm": 40.125308990478516,
|
| 1143 |
+
"learning_rate": 3.415736310473153e-05,
|
| 1144 |
+
"loss": 0.1169,
|
| 1145 |
+
"step": 1610
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.1875598086124401,
|
| 1149 |
+
"grad_norm": 0.015308327041566372,
|
| 1150 |
+
"learning_rate": 3.402445507708666e-05,
|
| 1151 |
+
"loss": 0.1322,
|
| 1152 |
+
"step": 1620
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 1.1899521531100479,
|
| 1156 |
+
"grad_norm": 57.31826400756836,
|
| 1157 |
+
"learning_rate": 3.389154704944179e-05,
|
| 1158 |
+
"loss": 0.2188,
|
| 1159 |
+
"step": 1630
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 1.1923444976076556,
|
| 1163 |
+
"grad_norm": 2.116442918777466,
|
| 1164 |
+
"learning_rate": 3.3758639021796915e-05,
|
| 1165 |
+
"loss": 0.0865,
|
| 1166 |
+
"step": 1640
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.194736842105263,
|
| 1170 |
+
"grad_norm": 2.048318862915039,
|
| 1171 |
+
"learning_rate": 3.362573099415205e-05,
|
| 1172 |
+
"loss": 0.0313,
|
| 1173 |
+
"step": 1650
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 1.1971291866028708,
|
| 1177 |
+
"grad_norm": 1.6264055967330933,
|
| 1178 |
+
"learning_rate": 3.349282296650718e-05,
|
| 1179 |
+
"loss": 0.0748,
|
| 1180 |
+
"step": 1660
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.1995215311004785,
|
| 1184 |
+
"grad_norm": 0.020111849531531334,
|
| 1185 |
+
"learning_rate": 3.335991493886231e-05,
|
| 1186 |
+
"loss": 0.1242,
|
| 1187 |
+
"step": 1670
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 1.2,
|
| 1191 |
+
"eval_accuracy": 0.9681620839363242,
|
| 1192 |
+
"eval_loss": 0.11400265246629715,
|
| 1193 |
+
"eval_runtime": 833.2657,
|
| 1194 |
+
"eval_samples_per_second": 0.829,
|
| 1195 |
+
"eval_steps_per_second": 0.208,
|
| 1196 |
+
"step": 1672
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.001913875598086,
|
| 1200 |
+
"grad_norm": 0.04623245820403099,
|
| 1201 |
+
"learning_rate": 3.322700691121744e-05,
|
| 1202 |
+
"loss": 0.2269,
|
| 1203 |
+
"step": 1680
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.004306220095694,
|
| 1207 |
+
"grad_norm": 0.04105829820036888,
|
| 1208 |
+
"learning_rate": 3.309409888357257e-05,
|
| 1209 |
+
"loss": 0.0314,
|
| 1210 |
+
"step": 1690
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 2.0066985645933015,
|
| 1214 |
+
"grad_norm": 0.012316164560616016,
|
| 1215 |
+
"learning_rate": 3.29611908559277e-05,
|
| 1216 |
+
"loss": 0.0812,
|
| 1217 |
+
"step": 1700
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 2.0090909090909093,
|
| 1221 |
+
"grad_norm": 0.014452405273914337,
|
| 1222 |
+
"learning_rate": 3.282828282828283e-05,
|
| 1223 |
+
"loss": 0.095,
|
| 1224 |
+
"step": 1710
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 2.0114832535885165,
|
| 1228 |
+
"grad_norm": 0.02369452267885208,
|
| 1229 |
+
"learning_rate": 3.2695374800637965e-05,
|
| 1230 |
+
"loss": 0.0393,
|
| 1231 |
+
"step": 1720
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 2.0138755980861243,
|
| 1235 |
+
"grad_norm": 0.05732015147805214,
|
| 1236 |
+
"learning_rate": 3.256246677299309e-05,
|
| 1237 |
+
"loss": 0.0019,
|
| 1238 |
+
"step": 1730
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.016267942583732,
|
| 1242 |
+
"grad_norm": 0.23258601129055023,
|
| 1243 |
+
"learning_rate": 3.242955874534822e-05,
|
| 1244 |
+
"loss": 0.1842,
|
| 1245 |
+
"step": 1740
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 2.0186602870813397,
|
| 1249 |
+
"grad_norm": 0.09864845871925354,
|
| 1250 |
+
"learning_rate": 3.229665071770335e-05,
|
| 1251 |
+
"loss": 0.0921,
|
| 1252 |
+
"step": 1750
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 2.0210526315789474,
|
| 1256 |
+
"grad_norm": 472.58856201171875,
|
| 1257 |
+
"learning_rate": 3.216374269005848e-05,
|
| 1258 |
+
"loss": 0.1973,
|
| 1259 |
+
"step": 1760
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 2.023444976076555,
|
| 1263 |
+
"grad_norm": 0.010467225685715675,
|
| 1264 |
+
"learning_rate": 3.203083466241361e-05,
|
| 1265 |
+
"loss": 0.0142,
|
| 1266 |
+
"step": 1770
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 2.025837320574163,
|
| 1270 |
+
"grad_norm": 34.93497848510742,
|
| 1271 |
+
"learning_rate": 3.189792663476874e-05,
|
| 1272 |
+
"loss": 0.1638,
|
| 1273 |
+
"step": 1780
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 2.02822966507177,
|
| 1277 |
+
"grad_norm": 108.92484283447266,
|
| 1278 |
+
"learning_rate": 3.176501860712387e-05,
|
| 1279 |
+
"loss": 0.2049,
|
| 1280 |
+
"step": 1790
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.030622009569378,
|
| 1284 |
+
"grad_norm": 0.12743550539016724,
|
| 1285 |
+
"learning_rate": 3.1632110579479e-05,
|
| 1286 |
+
"loss": 0.0541,
|
| 1287 |
+
"step": 1800
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 2.0330143540669856,
|
| 1291 |
+
"grad_norm": 0.01830146089196205,
|
| 1292 |
+
"learning_rate": 3.1499202551834136e-05,
|
| 1293 |
+
"loss": 0.0266,
|
| 1294 |
+
"step": 1810
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 2.0354066985645933,
|
| 1298 |
+
"grad_norm": 0.07744389772415161,
|
| 1299 |
+
"learning_rate": 3.136629452418926e-05,
|
| 1300 |
+
"loss": 0.2081,
|
| 1301 |
+
"step": 1820
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 2.037799043062201,
|
| 1305 |
+
"grad_norm": 0.09580251574516296,
|
| 1306 |
+
"learning_rate": 3.1233386496544395e-05,
|
| 1307 |
+
"loss": 0.1809,
|
| 1308 |
+
"step": 1830
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 2.0401913875598088,
|
| 1312 |
+
"grad_norm": 1.139581561088562,
|
| 1313 |
+
"learning_rate": 3.110047846889952e-05,
|
| 1314 |
+
"loss": 0.0047,
|
| 1315 |
+
"step": 1840
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 2.042583732057416,
|
| 1319 |
+
"grad_norm": 109.73816680908203,
|
| 1320 |
+
"learning_rate": 3.0967570441254654e-05,
|
| 1321 |
+
"loss": 0.1755,
|
| 1322 |
+
"step": 1850
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.044976076555024,
|
| 1326 |
+
"grad_norm": 0.021538248285651207,
|
| 1327 |
+
"learning_rate": 3.083466241360978e-05,
|
| 1328 |
+
"loss": 0.146,
|
| 1329 |
+
"step": 1860
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 2.0473684210526315,
|
| 1333 |
+
"grad_norm": 0.06147049739956856,
|
| 1334 |
+
"learning_rate": 3.0701754385964913e-05,
|
| 1335 |
+
"loss": 0.1163,
|
| 1336 |
+
"step": 1870
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 2.0497607655502392,
|
| 1340 |
+
"grad_norm": 0.05214022099971771,
|
| 1341 |
+
"learning_rate": 3.056884635832004e-05,
|
| 1342 |
+
"loss": 0.0021,
|
| 1343 |
+
"step": 1880
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 2.052153110047847,
|
| 1347 |
+
"grad_norm": 0.006269617471843958,
|
| 1348 |
+
"learning_rate": 3.0435938330675173e-05,
|
| 1349 |
+
"loss": 0.0012,
|
| 1350 |
+
"step": 1890
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 2.0545454545454547,
|
| 1354 |
+
"grad_norm": 3.528353691101074,
|
| 1355 |
+
"learning_rate": 3.0303030303030306e-05,
|
| 1356 |
+
"loss": 0.0097,
|
| 1357 |
+
"step": 1900
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 2.0569377990430624,
|
| 1361 |
+
"grad_norm": 0.43716534972190857,
|
| 1362 |
+
"learning_rate": 3.017012227538544e-05,
|
| 1363 |
+
"loss": 0.0138,
|
| 1364 |
+
"step": 1910
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 2.0593301435406697,
|
| 1368 |
+
"grad_norm": 0.008898725733160973,
|
| 1369 |
+
"learning_rate": 3.0037214247740565e-05,
|
| 1370 |
+
"loss": 0.1143,
|
| 1371 |
+
"step": 1920
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 2.0617224880382774,
|
| 1375 |
+
"grad_norm": 18.777921676635742,
|
| 1376 |
+
"learning_rate": 2.9904306220095695e-05,
|
| 1377 |
+
"loss": 0.0059,
|
| 1378 |
+
"step": 1930
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 2.064114832535885,
|
| 1382 |
+
"grad_norm": 0.010017622262239456,
|
| 1383 |
+
"learning_rate": 2.9771398192450828e-05,
|
| 1384 |
+
"loss": 0.0021,
|
| 1385 |
+
"step": 1940
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 2.066507177033493,
|
| 1389 |
+
"grad_norm": 0.028644366189837456,
|
| 1390 |
+
"learning_rate": 2.9638490164805954e-05,
|
| 1391 |
+
"loss": 0.0024,
|
| 1392 |
+
"step": 1950
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 2.0688995215311006,
|
| 1396 |
+
"grad_norm": 0.018241288140416145,
|
| 1397 |
+
"learning_rate": 2.9505582137161087e-05,
|
| 1398 |
+
"loss": 0.2653,
|
| 1399 |
+
"step": 1960
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 2.0712918660287083,
|
| 1403 |
+
"grad_norm": 0.027866262942552567,
|
| 1404 |
+
"learning_rate": 2.937267410951622e-05,
|
| 1405 |
+
"loss": 0.0659,
|
| 1406 |
+
"step": 1970
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 2.0736842105263156,
|
| 1410 |
+
"grad_norm": 0.0228541512042284,
|
| 1411 |
+
"learning_rate": 2.9239766081871346e-05,
|
| 1412 |
+
"loss": 0.0024,
|
| 1413 |
+
"step": 1980
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 2.0760765550239233,
|
| 1417 |
+
"grad_norm": 1.270308494567871,
|
| 1418 |
+
"learning_rate": 2.9106858054226476e-05,
|
| 1419 |
+
"loss": 0.1811,
|
| 1420 |
+
"step": 1990
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 2.078468899521531,
|
| 1424 |
+
"grad_norm": 0.03892623633146286,
|
| 1425 |
+
"learning_rate": 2.897395002658161e-05,
|
| 1426 |
+
"loss": 0.2084,
|
| 1427 |
+
"step": 2000
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 2.0808612440191387,
|
| 1431 |
+
"grad_norm": 145.8848876953125,
|
| 1432 |
+
"learning_rate": 2.8841041998936735e-05,
|
| 1433 |
+
"loss": 0.064,
|
| 1434 |
+
"step": 2010
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 2.0832535885167465,
|
| 1438 |
+
"grad_norm": 0.01369253359735012,
|
| 1439 |
+
"learning_rate": 2.8708133971291868e-05,
|
| 1440 |
+
"loss": 0.0026,
|
| 1441 |
+
"step": 2020
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 2.085645933014354,
|
| 1445 |
+
"grad_norm": 0.06184624508023262,
|
| 1446 |
+
"learning_rate": 2.8575225943646998e-05,
|
| 1447 |
+
"loss": 0.0449,
|
| 1448 |
+
"step": 2030
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 2.088038277511962,
|
| 1452 |
+
"grad_norm": 0.018334228545427322,
|
| 1453 |
+
"learning_rate": 2.8442317916002124e-05,
|
| 1454 |
+
"loss": 0.039,
|
| 1455 |
+
"step": 2040
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 2.090430622009569,
|
| 1459 |
+
"grad_norm": 0.01396853942424059,
|
| 1460 |
+
"learning_rate": 2.8309409888357257e-05,
|
| 1461 |
+
"loss": 0.0009,
|
| 1462 |
+
"step": 2050
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 2.092822966507177,
|
| 1466 |
+
"grad_norm": 9.39311408996582,
|
| 1467 |
+
"learning_rate": 2.817650186071239e-05,
|
| 1468 |
+
"loss": 0.0502,
|
| 1469 |
+
"step": 2060
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 2.0952153110047846,
|
| 1473 |
+
"grad_norm": 0.01102653332054615,
|
| 1474 |
+
"learning_rate": 2.8043593833067516e-05,
|
| 1475 |
+
"loss": 0.0142,
|
| 1476 |
+
"step": 2070
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 2.0976076555023924,
|
| 1480 |
+
"grad_norm": 0.030104391276836395,
|
| 1481 |
+
"learning_rate": 2.791068580542265e-05,
|
| 1482 |
+
"loss": 0.0014,
|
| 1483 |
+
"step": 2080
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 2.1,
|
| 1487 |
+
"grad_norm": 2.8127565383911133,
|
| 1488 |
+
"learning_rate": 2.777777777777778e-05,
|
| 1489 |
+
"loss": 0.0858,
|
| 1490 |
+
"step": 2090
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 2.102392344497608,
|
| 1494 |
+
"grad_norm": 0.022678395733237267,
|
| 1495 |
+
"learning_rate": 2.7644869750132905e-05,
|
| 1496 |
+
"loss": 0.0855,
|
| 1497 |
+
"step": 2100
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 2.104784688995215,
|
| 1501 |
+
"grad_norm": 0.011784915812313557,
|
| 1502 |
+
"learning_rate": 2.751196172248804e-05,
|
| 1503 |
+
"loss": 0.2841,
|
| 1504 |
+
"step": 2110
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 2.107177033492823,
|
| 1508 |
+
"grad_norm": 0.01566784456372261,
|
| 1509 |
+
"learning_rate": 2.737905369484317e-05,
|
| 1510 |
+
"loss": 0.4591,
|
| 1511 |
+
"step": 2120
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 2.1095693779904305,
|
| 1515 |
+
"grad_norm": 0.22584687173366547,
|
| 1516 |
+
"learning_rate": 2.7246145667198298e-05,
|
| 1517 |
+
"loss": 0.1164,
|
| 1518 |
+
"step": 2130
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 2.1119617224880383,
|
| 1522 |
+
"grad_norm": 0.293266624212265,
|
| 1523 |
+
"learning_rate": 2.711323763955343e-05,
|
| 1524 |
+
"loss": 0.0014,
|
| 1525 |
+
"step": 2140
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 2.114354066985646,
|
| 1529 |
+
"grad_norm": 0.04173046350479126,
|
| 1530 |
+
"learning_rate": 2.698032961190856e-05,
|
| 1531 |
+
"loss": 0.0024,
|
| 1532 |
+
"step": 2150
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 2.1167464114832537,
|
| 1536 |
+
"grad_norm": 0.014709273353219032,
|
| 1537 |
+
"learning_rate": 2.6847421584263693e-05,
|
| 1538 |
+
"loss": 0.0234,
|
| 1539 |
+
"step": 2160
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 2.1191387559808614,
|
| 1543 |
+
"grad_norm": 0.060059864073991776,
|
| 1544 |
+
"learning_rate": 2.671451355661882e-05,
|
| 1545 |
+
"loss": 0.0406,
|
| 1546 |
+
"step": 2170
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 2.1215311004784687,
|
| 1550 |
+
"grad_norm": 0.025863196700811386,
|
| 1551 |
+
"learning_rate": 2.6581605528973953e-05,
|
| 1552 |
+
"loss": 0.1082,
|
| 1553 |
+
"step": 2180
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 2.1239234449760764,
|
| 1557 |
+
"grad_norm": 0.011012042872607708,
|
| 1558 |
+
"learning_rate": 2.6448697501329082e-05,
|
| 1559 |
+
"loss": 0.1153,
|
| 1560 |
+
"step": 2190
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 2.126315789473684,
|
| 1564 |
+
"grad_norm": 0.009336967021226883,
|
| 1565 |
+
"learning_rate": 2.6315789473684212e-05,
|
| 1566 |
+
"loss": 0.1899,
|
| 1567 |
+
"step": 2200
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 2.128708133971292,
|
| 1571 |
+
"grad_norm": 0.022180695086717606,
|
| 1572 |
+
"learning_rate": 2.618288144603934e-05,
|
| 1573 |
+
"loss": 0.0947,
|
| 1574 |
+
"step": 2210
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 2.1311004784688996,
|
| 1578 |
+
"grad_norm": 0.06437038630247116,
|
| 1579 |
+
"learning_rate": 2.6049973418394475e-05,
|
| 1580 |
+
"loss": 0.046,
|
| 1581 |
+
"step": 2220
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 2.1334928229665073,
|
| 1585 |
+
"grad_norm": 0.023490458726882935,
|
| 1586 |
+
"learning_rate": 2.59170653907496e-05,
|
| 1587 |
+
"loss": 0.0028,
|
| 1588 |
+
"step": 2230
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 2.1358851674641146,
|
| 1592 |
+
"grad_norm": 0.010341249406337738,
|
| 1593 |
+
"learning_rate": 2.5784157363104734e-05,
|
| 1594 |
+
"loss": 0.0011,
|
| 1595 |
+
"step": 2240
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 2.1382775119617223,
|
| 1599 |
+
"grad_norm": 0.019244810566306114,
|
| 1600 |
+
"learning_rate": 2.5651249335459864e-05,
|
| 1601 |
+
"loss": 0.001,
|
| 1602 |
+
"step": 2250
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 2.14066985645933,
|
| 1606 |
+
"grad_norm": 2.1345505714416504,
|
| 1607 |
+
"learning_rate": 2.551834130781499e-05,
|
| 1608 |
+
"loss": 0.1981,
|
| 1609 |
+
"step": 2260
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 2.143062200956938,
|
| 1613 |
+
"grad_norm": 0.06913293898105621,
|
| 1614 |
+
"learning_rate": 2.5385433280170123e-05,
|
| 1615 |
+
"loss": 0.0117,
|
| 1616 |
+
"step": 2270
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 2.1454545454545455,
|
| 1620 |
+
"grad_norm": 17.53659439086914,
|
| 1621 |
+
"learning_rate": 2.5252525252525256e-05,
|
| 1622 |
+
"loss": 0.0048,
|
| 1623 |
+
"step": 2280
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 2.1478468899521532,
|
| 1627 |
+
"grad_norm": 1.950246810913086,
|
| 1628 |
+
"learning_rate": 2.5119617224880382e-05,
|
| 1629 |
+
"loss": 0.0045,
|
| 1630 |
+
"step": 2290
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 2.150239234449761,
|
| 1634 |
+
"grad_norm": 0.15338365733623505,
|
| 1635 |
+
"learning_rate": 2.4986709197235515e-05,
|
| 1636 |
+
"loss": 0.0287,
|
| 1637 |
+
"step": 2300
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 2.1526315789473682,
|
| 1641 |
+
"grad_norm": 0.08566514402627945,
|
| 1642 |
+
"learning_rate": 2.485380116959064e-05,
|
| 1643 |
+
"loss": 0.0009,
|
| 1644 |
+
"step": 2310
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 2.155023923444976,
|
| 1648 |
+
"grad_norm": 0.005700881592929363,
|
| 1649 |
+
"learning_rate": 2.4720893141945774e-05,
|
| 1650 |
+
"loss": 0.0013,
|
| 1651 |
+
"step": 2320
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 2.1574162679425837,
|
| 1655 |
+
"grad_norm": 0.010471828281879425,
|
| 1656 |
+
"learning_rate": 2.4587985114300904e-05,
|
| 1657 |
+
"loss": 0.04,
|
| 1658 |
+
"step": 2330
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 2.1598086124401914,
|
| 1662 |
+
"grad_norm": 0.15442727506160736,
|
| 1663 |
+
"learning_rate": 2.4455077086656037e-05,
|
| 1664 |
+
"loss": 0.0032,
|
| 1665 |
+
"step": 2340
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 2.162200956937799,
|
| 1669 |
+
"grad_norm": 0.013687992468476295,
|
| 1670 |
+
"learning_rate": 2.4322169059011167e-05,
|
| 1671 |
+
"loss": 0.0006,
|
| 1672 |
+
"step": 2350
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 2.164593301435407,
|
| 1676 |
+
"grad_norm": 15.276226043701172,
|
| 1677 |
+
"learning_rate": 2.4189261031366296e-05,
|
| 1678 |
+
"loss": 0.1538,
|
| 1679 |
+
"step": 2360
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 2.166985645933014,
|
| 1683 |
+
"grad_norm": 0.011922297067940235,
|
| 1684 |
+
"learning_rate": 2.4056353003721426e-05,
|
| 1685 |
+
"loss": 0.004,
|
| 1686 |
+
"step": 2370
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 2.169377990430622,
|
| 1690 |
+
"grad_norm": 0.012630579993128777,
|
| 1691 |
+
"learning_rate": 2.3923444976076556e-05,
|
| 1692 |
+
"loss": 0.0029,
|
| 1693 |
+
"step": 2380
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 2.1717703349282296,
|
| 1697 |
+
"grad_norm": 2.478867530822754,
|
| 1698 |
+
"learning_rate": 2.3790536948431685e-05,
|
| 1699 |
+
"loss": 0.0054,
|
| 1700 |
+
"step": 2390
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 2.1741626794258373,
|
| 1704 |
+
"grad_norm": 0.032478444278240204,
|
| 1705 |
+
"learning_rate": 2.365762892078682e-05,
|
| 1706 |
+
"loss": 0.0009,
|
| 1707 |
+
"step": 2400
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 2.176555023923445,
|
| 1711 |
+
"grad_norm": 1.2129992246627808,
|
| 1712 |
+
"learning_rate": 2.3524720893141948e-05,
|
| 1713 |
+
"loss": 0.0015,
|
| 1714 |
+
"step": 2410
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 2.1789473684210527,
|
| 1718 |
+
"grad_norm": 68.73790740966797,
|
| 1719 |
+
"learning_rate": 2.3391812865497074e-05,
|
| 1720 |
+
"loss": 0.0066,
|
| 1721 |
+
"step": 2420
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 2.1813397129186605,
|
| 1725 |
+
"grad_norm": 0.03823688626289368,
|
| 1726 |
+
"learning_rate": 2.3258904837852207e-05,
|
| 1727 |
+
"loss": 0.0034,
|
| 1728 |
+
"step": 2430
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 2.1837320574162677,
|
| 1732 |
+
"grad_norm": 0.010348823852837086,
|
| 1733 |
+
"learning_rate": 2.3125996810207337e-05,
|
| 1734 |
+
"loss": 0.0034,
|
| 1735 |
+
"step": 2440
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 2.1861244019138755,
|
| 1739 |
+
"grad_norm": 0.03327915072441101,
|
| 1740 |
+
"learning_rate": 2.299308878256247e-05,
|
| 1741 |
+
"loss": 0.0044,
|
| 1742 |
+
"step": 2450
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 2.188516746411483,
|
| 1746 |
+
"grad_norm": 0.0061185345984995365,
|
| 1747 |
+
"learning_rate": 2.28601807549176e-05,
|
| 1748 |
+
"loss": 0.001,
|
| 1749 |
+
"step": 2460
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 2.190909090909091,
|
| 1753 |
+
"grad_norm": 0.019545989111065865,
|
| 1754 |
+
"learning_rate": 2.272727272727273e-05,
|
| 1755 |
+
"loss": 0.0545,
|
| 1756 |
+
"step": 2470
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 2.1933014354066986,
|
| 1760 |
+
"grad_norm": 0.027101656422019005,
|
| 1761 |
+
"learning_rate": 2.259436469962786e-05,
|
| 1762 |
+
"loss": 0.0021,
|
| 1763 |
+
"step": 2480
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 2.1956937799043064,
|
| 1767 |
+
"grad_norm": 99.50630950927734,
|
| 1768 |
+
"learning_rate": 2.246145667198299e-05,
|
| 1769 |
+
"loss": 0.0218,
|
| 1770 |
+
"step": 2490
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 2.1980861244019136,
|
| 1774 |
+
"grad_norm": 0.009054621681571007,
|
| 1775 |
+
"learning_rate": 2.2328548644338118e-05,
|
| 1776 |
+
"loss": 0.0008,
|
| 1777 |
+
"step": 2500
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 2.2,
|
| 1781 |
+
"eval_accuracy": 0.9797395079594791,
|
| 1782 |
+
"eval_loss": 0.09082657098770142,
|
| 1783 |
+
"eval_runtime": 794.713,
|
| 1784 |
+
"eval_samples_per_second": 0.869,
|
| 1785 |
+
"eval_steps_per_second": 0.218,
|
| 1786 |
+
"step": 2508
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 3.0004784688995216,
|
| 1790 |
+
"grad_norm": 0.006668519228696823,
|
| 1791 |
+
"learning_rate": 2.219564061669325e-05,
|
| 1792 |
+
"loss": 0.2009,
|
| 1793 |
+
"step": 2510
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 3.0028708133971294,
|
| 1797 |
+
"grad_norm": 38.766822814941406,
|
| 1798 |
+
"learning_rate": 2.206273258904838e-05,
|
| 1799 |
+
"loss": 0.2188,
|
| 1800 |
+
"step": 2520
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 3.0052631578947366,
|
| 1804 |
+
"grad_norm": 0.01620459370315075,
|
| 1805 |
+
"learning_rate": 2.1929824561403507e-05,
|
| 1806 |
+
"loss": 0.0007,
|
| 1807 |
+
"step": 2530
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 3.0076555023923444,
|
| 1811 |
+
"grad_norm": 0.010571425780653954,
|
| 1812 |
+
"learning_rate": 2.179691653375864e-05,
|
| 1813 |
+
"loss": 0.0011,
|
| 1814 |
+
"step": 2540
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 3.010047846889952,
|
| 1818 |
+
"grad_norm": 0.007847270928323269,
|
| 1819 |
+
"learning_rate": 2.166400850611377e-05,
|
| 1820 |
+
"loss": 0.0019,
|
| 1821 |
+
"step": 2550
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 3.01244019138756,
|
| 1825 |
+
"grad_norm": 2.9406144618988037,
|
| 1826 |
+
"learning_rate": 2.1531100478468903e-05,
|
| 1827 |
+
"loss": 0.008,
|
| 1828 |
+
"step": 2560
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 3.0148325358851675,
|
| 1832 |
+
"grad_norm": 0.16233409941196442,
|
| 1833 |
+
"learning_rate": 2.1398192450824032e-05,
|
| 1834 |
+
"loss": 0.001,
|
| 1835 |
+
"step": 2570
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 3.0172248803827753,
|
| 1839 |
+
"grad_norm": 0.034926705062389374,
|
| 1840 |
+
"learning_rate": 2.126528442317916e-05,
|
| 1841 |
+
"loss": 0.0007,
|
| 1842 |
+
"step": 2580
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 3.019617224880383,
|
| 1846 |
+
"grad_norm": 0.007242594845592976,
|
| 1847 |
+
"learning_rate": 2.1132376395534292e-05,
|
| 1848 |
+
"loss": 0.2104,
|
| 1849 |
+
"step": 2590
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 3.0220095693779903,
|
| 1853 |
+
"grad_norm": 0.007837713696062565,
|
| 1854 |
+
"learning_rate": 2.099946836788942e-05,
|
| 1855 |
+
"loss": 0.2077,
|
| 1856 |
+
"step": 2600
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 3.024401913875598,
|
| 1860 |
+
"grad_norm": 0.006022488698363304,
|
| 1861 |
+
"learning_rate": 2.086656034024455e-05,
|
| 1862 |
+
"loss": 0.0011,
|
| 1863 |
+
"step": 2610
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 3.0267942583732057,
|
| 1867 |
+
"grad_norm": 0.08888078480958939,
|
| 1868 |
+
"learning_rate": 2.0733652312599684e-05,
|
| 1869 |
+
"loss": 0.0017,
|
| 1870 |
+
"step": 2620
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 3.0291866028708134,
|
| 1874 |
+
"grad_norm": 0.05022843927145004,
|
| 1875 |
+
"learning_rate": 2.0600744284954814e-05,
|
| 1876 |
+
"loss": 0.1537,
|
| 1877 |
+
"step": 2630
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 3.031578947368421,
|
| 1881 |
+
"grad_norm": 0.020734012126922607,
|
| 1882 |
+
"learning_rate": 2.046783625730994e-05,
|
| 1883 |
+
"loss": 0.0015,
|
| 1884 |
+
"step": 2640
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 3.033971291866029,
|
| 1888 |
+
"grad_norm": 97.93255615234375,
|
| 1889 |
+
"learning_rate": 2.0334928229665073e-05,
|
| 1890 |
+
"loss": 0.0167,
|
| 1891 |
+
"step": 2650
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 3.036363636363636,
|
| 1895 |
+
"grad_norm": 0.018566004931926727,
|
| 1896 |
+
"learning_rate": 2.0202020202020203e-05,
|
| 1897 |
+
"loss": 0.0012,
|
| 1898 |
+
"step": 2660
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 3.038755980861244,
|
| 1902 |
+
"grad_norm": 0.005064102355390787,
|
| 1903 |
+
"learning_rate": 2.0069112174375332e-05,
|
| 1904 |
+
"loss": 0.0005,
|
| 1905 |
+
"step": 2670
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 3.0411483253588516,
|
| 1909 |
+
"grad_norm": 0.019417481496930122,
|
| 1910 |
+
"learning_rate": 1.9936204146730465e-05,
|
| 1911 |
+
"loss": 0.0049,
|
| 1912 |
+
"step": 2680
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 3.0435406698564593,
|
| 1916 |
+
"grad_norm": 0.007396561559289694,
|
| 1917 |
+
"learning_rate": 1.980329611908559e-05,
|
| 1918 |
+
"loss": 0.0025,
|
| 1919 |
+
"step": 2690
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 3.045933014354067,
|
| 1923 |
+
"grad_norm": 61.74516677856445,
|
| 1924 |
+
"learning_rate": 1.9670388091440725e-05,
|
| 1925 |
+
"loss": 0.0991,
|
| 1926 |
+
"step": 2700
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 3.0483253588516748,
|
| 1930 |
+
"grad_norm": 0.021895471960306168,
|
| 1931 |
+
"learning_rate": 1.9537480063795854e-05,
|
| 1932 |
+
"loss": 0.0007,
|
| 1933 |
+
"step": 2710
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 3.0507177033492825,
|
| 1937 |
+
"grad_norm": 0.011149121448397636,
|
| 1938 |
+
"learning_rate": 1.9404572036150984e-05,
|
| 1939 |
+
"loss": 0.0006,
|
| 1940 |
+
"step": 2720
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 3.0531100478468898,
|
| 1944 |
+
"grad_norm": 0.004441165365278721,
|
| 1945 |
+
"learning_rate": 1.9271664008506117e-05,
|
| 1946 |
+
"loss": 0.1103,
|
| 1947 |
+
"step": 2730
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 3.0555023923444975,
|
| 1951 |
+
"grad_norm": 0.01931576617062092,
|
| 1952 |
+
"learning_rate": 1.9138755980861243e-05,
|
| 1953 |
+
"loss": 0.0029,
|
| 1954 |
+
"step": 2740
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 3.057894736842105,
|
| 1958 |
+
"grad_norm": 0.005823149345815182,
|
| 1959 |
+
"learning_rate": 1.9005847953216373e-05,
|
| 1960 |
+
"loss": 0.1682,
|
| 1961 |
+
"step": 2750
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 3.060287081339713,
|
| 1965 |
+
"grad_norm": 0.00678264070302248,
|
| 1966 |
+
"learning_rate": 1.8872939925571506e-05,
|
| 1967 |
+
"loss": 0.1651,
|
| 1968 |
+
"step": 2760
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 3.0626794258373207,
|
| 1972 |
+
"grad_norm": 0.010020887479186058,
|
| 1973 |
+
"learning_rate": 1.8740031897926636e-05,
|
| 1974 |
+
"loss": 0.0011,
|
| 1975 |
+
"step": 2770
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 3.0650717703349284,
|
| 1979 |
+
"grad_norm": 0.005996669642627239,
|
| 1980 |
+
"learning_rate": 1.8607123870281765e-05,
|
| 1981 |
+
"loss": 0.0018,
|
| 1982 |
+
"step": 2780
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 3.0674641148325357,
|
| 1986 |
+
"grad_norm": 0.09342297166585922,
|
| 1987 |
+
"learning_rate": 1.8474215842636898e-05,
|
| 1988 |
+
"loss": 0.0008,
|
| 1989 |
+
"step": 2790
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 3.0698564593301434,
|
| 1993 |
+
"grad_norm": 0.00897641945630312,
|
| 1994 |
+
"learning_rate": 1.8341307814992024e-05,
|
| 1995 |
+
"loss": 0.0108,
|
| 1996 |
+
"step": 2800
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 3.072248803827751,
|
| 2000 |
+
"grad_norm": 0.1280699372291565,
|
| 2001 |
+
"learning_rate": 1.8208399787347157e-05,
|
| 2002 |
+
"loss": 0.0066,
|
| 2003 |
+
"step": 2810
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 3.074641148325359,
|
| 2007 |
+
"grad_norm": 0.0044610267505049706,
|
| 2008 |
+
"learning_rate": 1.8075491759702287e-05,
|
| 2009 |
+
"loss": 0.0006,
|
| 2010 |
+
"step": 2820
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 3.0770334928229666,
|
| 2014 |
+
"grad_norm": 0.006019377615302801,
|
| 2015 |
+
"learning_rate": 1.7942583732057417e-05,
|
| 2016 |
+
"loss": 0.0005,
|
| 2017 |
+
"step": 2830
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 3.0794258373205743,
|
| 2021 |
+
"grad_norm": 0.0070389206521213055,
|
| 2022 |
+
"learning_rate": 1.780967570441255e-05,
|
| 2023 |
+
"loss": 0.006,
|
| 2024 |
+
"step": 2840
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 3.081818181818182,
|
| 2028 |
+
"grad_norm": 0.010742398910224438,
|
| 2029 |
+
"learning_rate": 1.7676767676767676e-05,
|
| 2030 |
+
"loss": 0.0012,
|
| 2031 |
+
"step": 2850
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 3.0842105263157893,
|
| 2035 |
+
"grad_norm": 0.019535677507519722,
|
| 2036 |
+
"learning_rate": 1.7543859649122806e-05,
|
| 2037 |
+
"loss": 0.0005,
|
| 2038 |
+
"step": 2860
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 3.086602870813397,
|
| 2042 |
+
"grad_norm": 0.004338996019214392,
|
| 2043 |
+
"learning_rate": 1.741095162147794e-05,
|
| 2044 |
+
"loss": 0.0006,
|
| 2045 |
+
"step": 2870
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 3.0889952153110047,
|
| 2049 |
+
"grad_norm": 0.004541581030935049,
|
| 2050 |
+
"learning_rate": 1.727804359383307e-05,
|
| 2051 |
+
"loss": 0.0873,
|
| 2052 |
+
"step": 2880
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 3.0913875598086125,
|
| 2056 |
+
"grad_norm": 0.006278311833739281,
|
| 2057 |
+
"learning_rate": 1.7145135566188198e-05,
|
| 2058 |
+
"loss": 0.0012,
|
| 2059 |
+
"step": 2890
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 3.09377990430622,
|
| 2063 |
+
"grad_norm": 1.023903250694275,
|
| 2064 |
+
"learning_rate": 1.701222753854333e-05,
|
| 2065 |
+
"loss": 0.0007,
|
| 2066 |
+
"step": 2900
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 3.096172248803828,
|
| 2070 |
+
"grad_norm": 0.005157362204045057,
|
| 2071 |
+
"learning_rate": 1.6879319510898457e-05,
|
| 2072 |
+
"loss": 0.0004,
|
| 2073 |
+
"step": 2910
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 3.098564593301435,
|
| 2077 |
+
"grad_norm": 0.006036865524947643,
|
| 2078 |
+
"learning_rate": 1.674641148325359e-05,
|
| 2079 |
+
"loss": 0.0041,
|
| 2080 |
+
"step": 2920
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 3.100956937799043,
|
| 2084 |
+
"grad_norm": 0.7861006259918213,
|
| 2085 |
+
"learning_rate": 1.661350345560872e-05,
|
| 2086 |
+
"loss": 0.003,
|
| 2087 |
+
"step": 2930
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 3.1033492822966506,
|
| 2091 |
+
"grad_norm": 0.4006275534629822,
|
| 2092 |
+
"learning_rate": 1.648059542796385e-05,
|
| 2093 |
+
"loss": 0.0253,
|
| 2094 |
+
"step": 2940
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 3.1057416267942584,
|
| 2098 |
+
"grad_norm": 0.13688111305236816,
|
| 2099 |
+
"learning_rate": 1.6347687400318983e-05,
|
| 2100 |
+
"loss": 0.0006,
|
| 2101 |
+
"step": 2950
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 3.108133971291866,
|
| 2105 |
+
"grad_norm": 0.007709797937422991,
|
| 2106 |
+
"learning_rate": 1.621477937267411e-05,
|
| 2107 |
+
"loss": 0.0006,
|
| 2108 |
+
"step": 2960
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 3.110526315789474,
|
| 2112 |
+
"grad_norm": 0.004617704544216394,
|
| 2113 |
+
"learning_rate": 1.608187134502924e-05,
|
| 2114 |
+
"loss": 0.0008,
|
| 2115 |
+
"step": 2970
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 3.1129186602870815,
|
| 2119 |
+
"grad_norm": 0.011701893992722034,
|
| 2120 |
+
"learning_rate": 1.594896331738437e-05,
|
| 2121 |
+
"loss": 0.0582,
|
| 2122 |
+
"step": 2980
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 3.115311004784689,
|
| 2126 |
+
"grad_norm": 0.20932549238204956,
|
| 2127 |
+
"learning_rate": 1.58160552897395e-05,
|
| 2128 |
+
"loss": 0.0012,
|
| 2129 |
+
"step": 2990
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 3.1177033492822965,
|
| 2133 |
+
"grad_norm": 0.008711748756468296,
|
| 2134 |
+
"learning_rate": 1.568314726209463e-05,
|
| 2135 |
+
"loss": 0.0938,
|
| 2136 |
+
"step": 3000
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 3.1200956937799043,
|
| 2140 |
+
"grad_norm": 0.004468224477022886,
|
| 2141 |
+
"learning_rate": 1.555023923444976e-05,
|
| 2142 |
+
"loss": 0.0006,
|
| 2143 |
+
"step": 3010
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 3.122488038277512,
|
| 2147 |
+
"grad_norm": 0.061067596077919006,
|
| 2148 |
+
"learning_rate": 1.541733120680489e-05,
|
| 2149 |
+
"loss": 0.0007,
|
| 2150 |
+
"step": 3020
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 3.1248803827751197,
|
| 2154 |
+
"grad_norm": 0.025924433022737503,
|
| 2155 |
+
"learning_rate": 1.528442317916002e-05,
|
| 2156 |
+
"loss": 0.0147,
|
| 2157 |
+
"step": 3030
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 3.1272727272727274,
|
| 2161 |
+
"grad_norm": 0.33229324221611023,
|
| 2162 |
+
"learning_rate": 1.5151515151515153e-05,
|
| 2163 |
+
"loss": 0.0006,
|
| 2164 |
+
"step": 3040
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 3.1296650717703347,
|
| 2168 |
+
"grad_norm": 0.019822843372821808,
|
| 2169 |
+
"learning_rate": 1.5018607123870283e-05,
|
| 2170 |
+
"loss": 0.0017,
|
| 2171 |
+
"step": 3050
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 3.1320574162679424,
|
| 2175 |
+
"grad_norm": 0.009158506989479065,
|
| 2176 |
+
"learning_rate": 1.4885699096225414e-05,
|
| 2177 |
+
"loss": 0.0179,
|
| 2178 |
+
"step": 3060
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 3.13444976076555,
|
| 2182 |
+
"grad_norm": 104.4133071899414,
|
| 2183 |
+
"learning_rate": 1.4752791068580543e-05,
|
| 2184 |
+
"loss": 0.1185,
|
| 2185 |
+
"step": 3070
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 3.136842105263158,
|
| 2189 |
+
"grad_norm": 0.013615977019071579,
|
| 2190 |
+
"learning_rate": 1.4619883040935673e-05,
|
| 2191 |
+
"loss": 0.0012,
|
| 2192 |
+
"step": 3080
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 3.1392344497607656,
|
| 2196 |
+
"grad_norm": 0.007452609483152628,
|
| 2197 |
+
"learning_rate": 1.4486975013290804e-05,
|
| 2198 |
+
"loss": 0.0006,
|
| 2199 |
+
"step": 3090
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 3.1416267942583733,
|
| 2203 |
+
"grad_norm": 0.017694944515824318,
|
| 2204 |
+
"learning_rate": 1.4354066985645934e-05,
|
| 2205 |
+
"loss": 0.0889,
|
| 2206 |
+
"step": 3100
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 3.144019138755981,
|
| 2210 |
+
"grad_norm": 0.0034980145283043385,
|
| 2211 |
+
"learning_rate": 1.4221158958001062e-05,
|
| 2212 |
+
"loss": 0.0137,
|
| 2213 |
+
"step": 3110
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 3.1464114832535883,
|
| 2217 |
+
"grad_norm": 0.006997366435825825,
|
| 2218 |
+
"learning_rate": 1.4088250930356195e-05,
|
| 2219 |
+
"loss": 0.1826,
|
| 2220 |
+
"step": 3120
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 3.148803827751196,
|
| 2224 |
+
"grad_norm": 0.20292145013809204,
|
| 2225 |
+
"learning_rate": 1.3955342902711325e-05,
|
| 2226 |
+
"loss": 0.0007,
|
| 2227 |
+
"step": 3130
|
| 2228 |
+
},
|
| 2229 |
+
{
|
| 2230 |
+
"epoch": 3.1511961722488038,
|
| 2231 |
+
"grad_norm": 0.009792889468371868,
|
| 2232 |
+
"learning_rate": 1.3822434875066453e-05,
|
| 2233 |
+
"loss": 0.0338,
|
| 2234 |
+
"step": 3140
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 3.1535885167464115,
|
| 2238 |
+
"grad_norm": 0.006899695377796888,
|
| 2239 |
+
"learning_rate": 1.3689526847421586e-05,
|
| 2240 |
+
"loss": 0.0006,
|
| 2241 |
+
"step": 3150
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 3.155980861244019,
|
| 2245 |
+
"grad_norm": 0.004146776627749205,
|
| 2246 |
+
"learning_rate": 1.3556618819776715e-05,
|
| 2247 |
+
"loss": 0.0019,
|
| 2248 |
+
"step": 3160
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 3.158373205741627,
|
| 2252 |
+
"grad_norm": 0.003076452063396573,
|
| 2253 |
+
"learning_rate": 1.3423710792131847e-05,
|
| 2254 |
+
"loss": 0.0004,
|
| 2255 |
+
"step": 3170
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 3.1607655502392342,
|
| 2259 |
+
"grad_norm": 0.022380296140909195,
|
| 2260 |
+
"learning_rate": 1.3290802764486976e-05,
|
| 2261 |
+
"loss": 0.0011,
|
| 2262 |
+
"step": 3180
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 3.163157894736842,
|
| 2266 |
+
"grad_norm": 0.005187951028347015,
|
| 2267 |
+
"learning_rate": 1.3157894736842106e-05,
|
| 2268 |
+
"loss": 0.0005,
|
| 2269 |
+
"step": 3190
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 3.1655502392344497,
|
| 2273 |
+
"grad_norm": 0.004998629912734032,
|
| 2274 |
+
"learning_rate": 1.3024986709197237e-05,
|
| 2275 |
+
"loss": 0.0006,
|
| 2276 |
+
"step": 3200
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 3.1679425837320574,
|
| 2280 |
+
"grad_norm": 0.016898494213819504,
|
| 2281 |
+
"learning_rate": 1.2892078681552367e-05,
|
| 2282 |
+
"loss": 0.0007,
|
| 2283 |
+
"step": 3210
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 3.170334928229665,
|
| 2287 |
+
"grad_norm": 0.43881699442863464,
|
| 2288 |
+
"learning_rate": 1.2759170653907495e-05,
|
| 2289 |
+
"loss": 0.0041,
|
| 2290 |
+
"step": 3220
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 3.172727272727273,
|
| 2294 |
+
"grad_norm": 0.10537323355674744,
|
| 2295 |
+
"learning_rate": 1.2626262626262628e-05,
|
| 2296 |
+
"loss": 0.0009,
|
| 2297 |
+
"step": 3230
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 3.1751196172248806,
|
| 2301 |
+
"grad_norm": 0.00541999377310276,
|
| 2302 |
+
"learning_rate": 1.2493354598617758e-05,
|
| 2303 |
+
"loss": 0.0004,
|
| 2304 |
+
"step": 3240
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 3.177511961722488,
|
| 2308 |
+
"grad_norm": 0.0035200186539441347,
|
| 2309 |
+
"learning_rate": 1.2360446570972887e-05,
|
| 2310 |
+
"loss": 0.0007,
|
| 2311 |
+
"step": 3250
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 3.1799043062200956,
|
| 2315 |
+
"grad_norm": 0.003413601079955697,
|
| 2316 |
+
"learning_rate": 1.2227538543328019e-05,
|
| 2317 |
+
"loss": 0.0004,
|
| 2318 |
+
"step": 3260
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 3.1822966507177033,
|
| 2322 |
+
"grad_norm": 0.003805672749876976,
|
| 2323 |
+
"learning_rate": 1.2094630515683148e-05,
|
| 2324 |
+
"loss": 0.0004,
|
| 2325 |
+
"step": 3270
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 3.184688995215311,
|
| 2329 |
+
"grad_norm": 0.008552059531211853,
|
| 2330 |
+
"learning_rate": 1.1961722488038278e-05,
|
| 2331 |
+
"loss": 0.0048,
|
| 2332 |
+
"step": 3280
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 3.1870813397129187,
|
| 2336 |
+
"grad_norm": 0.003063785843551159,
|
| 2337 |
+
"learning_rate": 1.182881446039341e-05,
|
| 2338 |
+
"loss": 0.0003,
|
| 2339 |
+
"step": 3290
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 3.1894736842105265,
|
| 2343 |
+
"grad_norm": 0.41841933131217957,
|
| 2344 |
+
"learning_rate": 1.1695906432748537e-05,
|
| 2345 |
+
"loss": 0.0005,
|
| 2346 |
+
"step": 3300
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 3.1918660287081337,
|
| 2350 |
+
"grad_norm": 0.003419842105358839,
|
| 2351 |
+
"learning_rate": 1.1562998405103668e-05,
|
| 2352 |
+
"loss": 0.0048,
|
| 2353 |
+
"step": 3310
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 3.1942583732057415,
|
| 2357 |
+
"grad_norm": 0.007587193511426449,
|
| 2358 |
+
"learning_rate": 1.14300903774588e-05,
|
| 2359 |
+
"loss": 0.0004,
|
| 2360 |
+
"step": 3320
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 3.196650717703349,
|
| 2364 |
+
"grad_norm": 0.014273949898779392,
|
| 2365 |
+
"learning_rate": 1.129718234981393e-05,
|
| 2366 |
+
"loss": 0.0004,
|
| 2367 |
+
"step": 3330
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 3.199043062200957,
|
| 2371 |
+
"grad_norm": 0.003971030004322529,
|
| 2372 |
+
"learning_rate": 1.1164274322169059e-05,
|
| 2373 |
+
"loss": 0.0004,
|
| 2374 |
+
"step": 3340
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 3.2,
|
| 2378 |
+
"eval_accuracy": 0.9956584659913169,
|
| 2379 |
+
"eval_loss": 0.011571008712053299,
|
| 2380 |
+
"eval_runtime": 802.1767,
|
| 2381 |
+
"eval_samples_per_second": 0.861,
|
| 2382 |
+
"eval_steps_per_second": 0.216,
|
| 2383 |
+
"step": 3344
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 4.0014354066985645,
|
| 2387 |
+
"grad_norm": 0.0065171197056770325,
|
| 2388 |
+
"learning_rate": 1.103136629452419e-05,
|
| 2389 |
+
"loss": 0.0004,
|
| 2390 |
+
"step": 3350
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 4.003827751196172,
|
| 2394 |
+
"grad_norm": 0.005604379344731569,
|
| 2395 |
+
"learning_rate": 1.089845826687932e-05,
|
| 2396 |
+
"loss": 0.0013,
|
| 2397 |
+
"step": 3360
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 4.00622009569378,
|
| 2401 |
+
"grad_norm": 0.009186330251395702,
|
| 2402 |
+
"learning_rate": 1.0765550239234451e-05,
|
| 2403 |
+
"loss": 0.0004,
|
| 2404 |
+
"step": 3370
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 4.008612440191388,
|
| 2408 |
+
"grad_norm": 0.004106089938431978,
|
| 2409 |
+
"learning_rate": 1.063264221158958e-05,
|
| 2410 |
+
"loss": 0.0013,
|
| 2411 |
+
"step": 3380
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 4.011004784688995,
|
| 2415 |
+
"grad_norm": 0.01519398856908083,
|
| 2416 |
+
"learning_rate": 1.049973418394471e-05,
|
| 2417 |
+
"loss": 0.0004,
|
| 2418 |
+
"step": 3390
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 4.013397129186603,
|
| 2422 |
+
"grad_norm": 0.025932341814041138,
|
| 2423 |
+
"learning_rate": 1.0366826156299842e-05,
|
| 2424 |
+
"loss": 0.0005,
|
| 2425 |
+
"step": 3400
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 4.015789473684211,
|
| 2429 |
+
"grad_norm": 0.003742667380720377,
|
| 2430 |
+
"learning_rate": 1.023391812865497e-05,
|
| 2431 |
+
"loss": 0.0015,
|
| 2432 |
+
"step": 3410
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 4.0181818181818185,
|
| 2436 |
+
"grad_norm": 0.011091819033026695,
|
| 2437 |
+
"learning_rate": 1.0101010101010101e-05,
|
| 2438 |
+
"loss": 0.0005,
|
| 2439 |
+
"step": 3420
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 4.020574162679426,
|
| 2443 |
+
"grad_norm": 0.14429621398448944,
|
| 2444 |
+
"learning_rate": 9.968102073365233e-06,
|
| 2445 |
+
"loss": 0.0008,
|
| 2446 |
+
"step": 3430
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 4.022966507177033,
|
| 2450 |
+
"grad_norm": 0.006444338243454695,
|
| 2451 |
+
"learning_rate": 9.835194045720362e-06,
|
| 2452 |
+
"loss": 0.0004,
|
| 2453 |
+
"step": 3440
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 4.025358851674641,
|
| 2457 |
+
"grad_norm": 0.003463227301836014,
|
| 2458 |
+
"learning_rate": 9.702286018075492e-06,
|
| 2459 |
+
"loss": 0.0008,
|
| 2460 |
+
"step": 3450
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 4.0277511961722485,
|
| 2464 |
+
"grad_norm": 0.003970850259065628,
|
| 2465 |
+
"learning_rate": 9.569377990430622e-06,
|
| 2466 |
+
"loss": 0.0003,
|
| 2467 |
+
"step": 3460
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 4.030143540669856,
|
| 2471 |
+
"grad_norm": 0.004473558627068996,
|
| 2472 |
+
"learning_rate": 9.436469962785753e-06,
|
| 2473 |
+
"loss": 0.0325,
|
| 2474 |
+
"step": 3470
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 4.032535885167464,
|
| 2478 |
+
"grad_norm": 0.0038171063642948866,
|
| 2479 |
+
"learning_rate": 9.303561935140883e-06,
|
| 2480 |
+
"loss": 0.0004,
|
| 2481 |
+
"step": 3480
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 4.034928229665072,
|
| 2485 |
+
"grad_norm": 0.0045017702504992485,
|
| 2486 |
+
"learning_rate": 9.170653907496012e-06,
|
| 2487 |
+
"loss": 0.0003,
|
| 2488 |
+
"step": 3490
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 4.037320574162679,
|
| 2492 |
+
"grad_norm": 0.0030059998389333487,
|
| 2493 |
+
"learning_rate": 9.037745879851144e-06,
|
| 2494 |
+
"loss": 0.0004,
|
| 2495 |
+
"step": 3500
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 4.039712918660287,
|
| 2499 |
+
"grad_norm": 0.03430689498782158,
|
| 2500 |
+
"learning_rate": 8.904837852206275e-06,
|
| 2501 |
+
"loss": 0.0005,
|
| 2502 |
+
"step": 3510
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 4.042105263157895,
|
| 2506 |
+
"grad_norm": 0.0038995530921965837,
|
| 2507 |
+
"learning_rate": 8.771929824561403e-06,
|
| 2508 |
+
"loss": 0.0004,
|
| 2509 |
+
"step": 3520
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 4.044497607655503,
|
| 2513 |
+
"grad_norm": 0.005899826996028423,
|
| 2514 |
+
"learning_rate": 8.639021796916534e-06,
|
| 2515 |
+
"loss": 0.0004,
|
| 2516 |
+
"step": 3530
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 4.04688995215311,
|
| 2520 |
+
"grad_norm": 0.01342977024614811,
|
| 2521 |
+
"learning_rate": 8.506113769271666e-06,
|
| 2522 |
+
"loss": 0.0003,
|
| 2523 |
+
"step": 3540
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 4.049282296650718,
|
| 2527 |
+
"grad_norm": 0.0030721963848918676,
|
| 2528 |
+
"learning_rate": 8.373205741626795e-06,
|
| 2529 |
+
"loss": 0.0005,
|
| 2530 |
+
"step": 3550
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 4.051674641148326,
|
| 2534 |
+
"grad_norm": 0.017302511259913445,
|
| 2535 |
+
"learning_rate": 8.240297713981925e-06,
|
| 2536 |
+
"loss": 0.0005,
|
| 2537 |
+
"step": 3560
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 4.054066985645933,
|
| 2541 |
+
"grad_norm": 0.018140502274036407,
|
| 2542 |
+
"learning_rate": 8.107389686337054e-06,
|
| 2543 |
+
"loss": 0.0003,
|
| 2544 |
+
"step": 3570
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 4.05645933014354,
|
| 2548 |
+
"grad_norm": 0.0029472303576767445,
|
| 2549 |
+
"learning_rate": 7.974481658692186e-06,
|
| 2550 |
+
"loss": 0.0003,
|
| 2551 |
+
"step": 3580
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 4.058851674641148,
|
| 2555 |
+
"grad_norm": 0.005908080376684666,
|
| 2556 |
+
"learning_rate": 7.841573631047315e-06,
|
| 2557 |
+
"loss": 0.0003,
|
| 2558 |
+
"step": 3590
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 4.061244019138756,
|
| 2562 |
+
"grad_norm": 0.004262011032551527,
|
| 2563 |
+
"learning_rate": 7.708665603402445e-06,
|
| 2564 |
+
"loss": 0.0005,
|
| 2565 |
+
"step": 3600
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 4.0636363636363635,
|
| 2569 |
+
"grad_norm": 0.004099706653505564,
|
| 2570 |
+
"learning_rate": 7.5757575757575764e-06,
|
| 2571 |
+
"loss": 0.0003,
|
| 2572 |
+
"step": 3610
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 4.066028708133971,
|
| 2576 |
+
"grad_norm": 0.00331715471111238,
|
| 2577 |
+
"learning_rate": 7.442849548112707e-06,
|
| 2578 |
+
"loss": 0.0003,
|
| 2579 |
+
"step": 3620
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 4.068421052631579,
|
| 2583 |
+
"grad_norm": 0.0059309545904397964,
|
| 2584 |
+
"learning_rate": 7.3099415204678366e-06,
|
| 2585 |
+
"loss": 0.0003,
|
| 2586 |
+
"step": 3630
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 4.070813397129187,
|
| 2590 |
+
"grad_norm": 0.004388094414025545,
|
| 2591 |
+
"learning_rate": 7.177033492822967e-06,
|
| 2592 |
+
"loss": 0.0015,
|
| 2593 |
+
"step": 3640
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 4.073205741626794,
|
| 2597 |
+
"grad_norm": 0.029676349833607674,
|
| 2598 |
+
"learning_rate": 7.0441254651780975e-06,
|
| 2599 |
+
"loss": 0.0004,
|
| 2600 |
+
"step": 3650
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 4.075598086124402,
|
| 2604 |
+
"grad_norm": 0.003281336510553956,
|
| 2605 |
+
"learning_rate": 6.911217437533226e-06,
|
| 2606 |
+
"loss": 0.0003,
|
| 2607 |
+
"step": 3660
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 4.07799043062201,
|
| 2611 |
+
"grad_norm": 0.02144763059914112,
|
| 2612 |
+
"learning_rate": 6.778309409888358e-06,
|
| 2613 |
+
"loss": 0.001,
|
| 2614 |
+
"step": 3670
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 4.0803827751196176,
|
| 2618 |
+
"grad_norm": 0.0047360495664179325,
|
| 2619 |
+
"learning_rate": 6.645401382243488e-06,
|
| 2620 |
+
"loss": 0.0004,
|
| 2621 |
+
"step": 3680
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 4.082775119617225,
|
| 2625 |
+
"grad_norm": 0.004007202573120594,
|
| 2626 |
+
"learning_rate": 6.512493354598619e-06,
|
| 2627 |
+
"loss": 0.0003,
|
| 2628 |
+
"step": 3690
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 4.085167464114832,
|
| 2632 |
+
"grad_norm": 0.01407600287348032,
|
| 2633 |
+
"learning_rate": 6.3795853269537475e-06,
|
| 2634 |
+
"loss": 0.083,
|
| 2635 |
+
"step": 3700
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 4.08755980861244,
|
| 2639 |
+
"grad_norm": 0.004998557735234499,
|
| 2640 |
+
"learning_rate": 6.246677299308879e-06,
|
| 2641 |
+
"loss": 0.0004,
|
| 2642 |
+
"step": 3710
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 4.089952153110048,
|
| 2646 |
+
"grad_norm": 0.003679374000057578,
|
| 2647 |
+
"learning_rate": 6.113769271664009e-06,
|
| 2648 |
+
"loss": 0.0003,
|
| 2649 |
+
"step": 3720
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 4.092344497607655,
|
| 2653 |
+
"grad_norm": 0.004649085458368063,
|
| 2654 |
+
"learning_rate": 5.980861244019139e-06,
|
| 2655 |
+
"loss": 0.0003,
|
| 2656 |
+
"step": 3730
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 4.094736842105263,
|
| 2660 |
+
"grad_norm": 0.009950079955160618,
|
| 2661 |
+
"learning_rate": 5.8479532163742686e-06,
|
| 2662 |
+
"loss": 0.0003,
|
| 2663 |
+
"step": 3740
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 4.097129186602871,
|
| 2667 |
+
"grad_norm": 0.002897342899814248,
|
| 2668 |
+
"learning_rate": 5.7150451887294e-06,
|
| 2669 |
+
"loss": 0.0004,
|
| 2670 |
+
"step": 3750
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 4.0995215311004785,
|
| 2674 |
+
"grad_norm": 0.002834454644471407,
|
| 2675 |
+
"learning_rate": 5.5821371610845296e-06,
|
| 2676 |
+
"loss": 0.0004,
|
| 2677 |
+
"step": 3760
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 4.101913875598086,
|
| 2681 |
+
"grad_norm": 0.0034763123840093613,
|
| 2682 |
+
"learning_rate": 5.44922913343966e-06,
|
| 2683 |
+
"loss": 0.0003,
|
| 2684 |
+
"step": 3770
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 4.104306220095694,
|
| 2688 |
+
"grad_norm": 0.01105444598942995,
|
| 2689 |
+
"learning_rate": 5.31632110579479e-06,
|
| 2690 |
+
"loss": 0.0004,
|
| 2691 |
+
"step": 3780
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 4.106698564593302,
|
| 2695 |
+
"grad_norm": 0.0030907553154975176,
|
| 2696 |
+
"learning_rate": 5.183413078149921e-06,
|
| 2697 |
+
"loss": 0.0019,
|
| 2698 |
+
"step": 3790
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 4.109090909090909,
|
| 2702 |
+
"grad_norm": 0.00505461310967803,
|
| 2703 |
+
"learning_rate": 5.050505050505051e-06,
|
| 2704 |
+
"loss": 0.0004,
|
| 2705 |
+
"step": 3800
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 4.111483253588517,
|
| 2709 |
+
"grad_norm": 0.0030872768256813288,
|
| 2710 |
+
"learning_rate": 4.917597022860181e-06,
|
| 2711 |
+
"loss": 0.0003,
|
| 2712 |
+
"step": 3810
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 4.113875598086125,
|
| 2716 |
+
"grad_norm": 0.0036562816239893436,
|
| 2717 |
+
"learning_rate": 4.784688995215311e-06,
|
| 2718 |
+
"loss": 0.0004,
|
| 2719 |
+
"step": 3820
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 4.116267942583732,
|
| 2723 |
+
"grad_norm": 0.00387106416746974,
|
| 2724 |
+
"learning_rate": 4.651780967570441e-06,
|
| 2725 |
+
"loss": 0.0004,
|
| 2726 |
+
"step": 3830
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 4.118660287081339,
|
| 2730 |
+
"grad_norm": 0.00350612192414701,
|
| 2731 |
+
"learning_rate": 4.518872939925572e-06,
|
| 2732 |
+
"loss": 0.0004,
|
| 2733 |
+
"step": 3840
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 4.121052631578947,
|
| 2737 |
+
"grad_norm": 0.015147331170737743,
|
| 2738 |
+
"learning_rate": 4.3859649122807014e-06,
|
| 2739 |
+
"loss": 0.0003,
|
| 2740 |
+
"step": 3850
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 4.123444976076555,
|
| 2744 |
+
"grad_norm": 0.003980391658842564,
|
| 2745 |
+
"learning_rate": 4.253056884635833e-06,
|
| 2746 |
+
"loss": 0.0003,
|
| 2747 |
+
"step": 3860
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 4.1258373205741625,
|
| 2751 |
+
"grad_norm": 0.00292576034553349,
|
| 2752 |
+
"learning_rate": 4.120148856990962e-06,
|
| 2753 |
+
"loss": 0.0003,
|
| 2754 |
+
"step": 3870
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 4.12822966507177,
|
| 2758 |
+
"grad_norm": 0.003524401690810919,
|
| 2759 |
+
"learning_rate": 3.987240829346093e-06,
|
| 2760 |
+
"loss": 0.0003,
|
| 2761 |
+
"step": 3880
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 4.130622009569378,
|
| 2765 |
+
"grad_norm": 0.003340751165524125,
|
| 2766 |
+
"learning_rate": 3.8543328017012225e-06,
|
| 2767 |
+
"loss": 0.0003,
|
| 2768 |
+
"step": 3890
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 4.133014354066986,
|
| 2772 |
+
"grad_norm": 0.00597594678401947,
|
| 2773 |
+
"learning_rate": 3.7214247740563535e-06,
|
| 2774 |
+
"loss": 0.0003,
|
| 2775 |
+
"step": 3900
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 4.135406698564593,
|
| 2779 |
+
"grad_norm": 0.40342918038368225,
|
| 2780 |
+
"learning_rate": 3.5885167464114835e-06,
|
| 2781 |
+
"loss": 0.0004,
|
| 2782 |
+
"step": 3910
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 4.137799043062201,
|
| 2786 |
+
"grad_norm": 0.0027185771614313126,
|
| 2787 |
+
"learning_rate": 3.455608718766613e-06,
|
| 2788 |
+
"loss": 0.0003,
|
| 2789 |
+
"step": 3920
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 4.140191387559809,
|
| 2793 |
+
"grad_norm": 0.003001193981617689,
|
| 2794 |
+
"learning_rate": 3.322700691121744e-06,
|
| 2795 |
+
"loss": 0.0003,
|
| 2796 |
+
"step": 3930
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 4.142583732057417,
|
| 2800 |
+
"grad_norm": 0.011745316907763481,
|
| 2801 |
+
"learning_rate": 3.1897926634768737e-06,
|
| 2802 |
+
"loss": 0.0004,
|
| 2803 |
+
"step": 3940
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 4.144976076555024,
|
| 2807 |
+
"grad_norm": 0.004574726335704327,
|
| 2808 |
+
"learning_rate": 3.0568846358320046e-06,
|
| 2809 |
+
"loss": 0.0003,
|
| 2810 |
+
"step": 3950
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 4.147368421052631,
|
| 2814 |
+
"grad_norm": 0.00624841870740056,
|
| 2815 |
+
"learning_rate": 2.9239766081871343e-06,
|
| 2816 |
+
"loss": 0.0003,
|
| 2817 |
+
"step": 3960
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 4.149760765550239,
|
| 2821 |
+
"grad_norm": 0.005558318924158812,
|
| 2822 |
+
"learning_rate": 2.7910685805422648e-06,
|
| 2823 |
+
"loss": 0.0003,
|
| 2824 |
+
"step": 3970
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 4.152153110047847,
|
| 2828 |
+
"grad_norm": 0.0035703121684491634,
|
| 2829 |
+
"learning_rate": 2.658160552897395e-06,
|
| 2830 |
+
"loss": 0.0003,
|
| 2831 |
+
"step": 3980
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 4.154545454545454,
|
| 2835 |
+
"grad_norm": 0.0034527035895735025,
|
| 2836 |
+
"learning_rate": 2.5252525252525253e-06,
|
| 2837 |
+
"loss": 0.0003,
|
| 2838 |
+
"step": 3990
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 4.156937799043062,
|
| 2842 |
+
"grad_norm": 0.0036090209614485502,
|
| 2843 |
+
"learning_rate": 2.3923444976076554e-06,
|
| 2844 |
+
"loss": 0.0003,
|
| 2845 |
+
"step": 4000
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 4.15933014354067,
|
| 2849 |
+
"grad_norm": 0.0026591310743242502,
|
| 2850 |
+
"learning_rate": 2.259436469962786e-06,
|
| 2851 |
+
"loss": 0.004,
|
| 2852 |
+
"step": 4010
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 4.1617224880382775,
|
| 2856 |
+
"grad_norm": 0.0031338431872427464,
|
| 2857 |
+
"learning_rate": 2.1265284423179164e-06,
|
| 2858 |
+
"loss": 0.0003,
|
| 2859 |
+
"step": 4020
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 4.164114832535885,
|
| 2863 |
+
"grad_norm": 0.006377640645951033,
|
| 2864 |
+
"learning_rate": 1.9936204146730465e-06,
|
| 2865 |
+
"loss": 0.0003,
|
| 2866 |
+
"step": 4030
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 4.166507177033493,
|
| 2870 |
+
"grad_norm": 0.003080443711951375,
|
| 2871 |
+
"learning_rate": 1.8607123870281767e-06,
|
| 2872 |
+
"loss": 0.0003,
|
| 2873 |
+
"step": 4040
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 4.168899521531101,
|
| 2877 |
+
"grad_norm": 0.1149815171957016,
|
| 2878 |
+
"learning_rate": 1.7278043593833066e-06,
|
| 2879 |
+
"loss": 0.0004,
|
| 2880 |
+
"step": 4050
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 4.171291866028708,
|
| 2884 |
+
"grad_norm": 0.0024256748147308826,
|
| 2885 |
+
"learning_rate": 1.5948963317384369e-06,
|
| 2886 |
+
"loss": 0.1869,
|
| 2887 |
+
"step": 4060
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 4.173684210526316,
|
| 2891 |
+
"grad_norm": 0.007068783510476351,
|
| 2892 |
+
"learning_rate": 1.4619883040935671e-06,
|
| 2893 |
+
"loss": 0.0003,
|
| 2894 |
+
"step": 4070
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 4.176076555023924,
|
| 2898 |
+
"grad_norm": 0.034253306686878204,
|
| 2899 |
+
"learning_rate": 1.3290802764486974e-06,
|
| 2900 |
+
"loss": 0.0004,
|
| 2901 |
+
"step": 4080
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 4.1784688995215316,
|
| 2905 |
+
"grad_norm": 0.003701072186231613,
|
| 2906 |
+
"learning_rate": 1.1961722488038277e-06,
|
| 2907 |
+
"loss": 0.0003,
|
| 2908 |
+
"step": 4090
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 4.180861244019138,
|
| 2912 |
+
"grad_norm": 0.0036447285674512386,
|
| 2913 |
+
"learning_rate": 1.0632642211589582e-06,
|
| 2914 |
+
"loss": 0.0003,
|
| 2915 |
+
"step": 4100
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 4.183253588516746,
|
| 2919 |
+
"grad_norm": 0.011764708906412125,
|
| 2920 |
+
"learning_rate": 9.303561935140884e-07,
|
| 2921 |
+
"loss": 0.0003,
|
| 2922 |
+
"step": 4110
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 4.185645933014354,
|
| 2926 |
+
"grad_norm": 0.005143681075423956,
|
| 2927 |
+
"learning_rate": 7.974481658692184e-07,
|
| 2928 |
+
"loss": 0.0025,
|
| 2929 |
+
"step": 4120
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 4.188038277511962,
|
| 2933 |
+
"grad_norm": 0.005259453784674406,
|
| 2934 |
+
"learning_rate": 6.645401382243487e-07,
|
| 2935 |
+
"loss": 0.0003,
|
| 2936 |
+
"step": 4130
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 4.190430622009569,
|
| 2940 |
+
"grad_norm": 0.00764945475384593,
|
| 2941 |
+
"learning_rate": 5.316321105794791e-07,
|
| 2942 |
+
"loss": 0.0003,
|
| 2943 |
+
"step": 4140
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 4.192822966507177,
|
| 2947 |
+
"grad_norm": 0.004675131756812334,
|
| 2948 |
+
"learning_rate": 3.987240829346092e-07,
|
| 2949 |
+
"loss": 0.0003,
|
| 2950 |
+
"step": 4150
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 4.195215311004785,
|
| 2954 |
+
"grad_norm": 0.003127172589302063,
|
| 2955 |
+
"learning_rate": 2.6581605528973955e-07,
|
| 2956 |
+
"loss": 0.0003,
|
| 2957 |
+
"step": 4160
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 4.1976076555023925,
|
| 2961 |
+
"grad_norm": 0.003473317250609398,
|
| 2962 |
+
"learning_rate": 1.3290802764486977e-07,
|
| 2963 |
+
"loss": 0.0009,
|
| 2964 |
+
"step": 4170
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 4.2,
|
| 2968 |
+
"grad_norm": 0.006573684047907591,
|
| 2969 |
+
"learning_rate": 0.0,
|
| 2970 |
+
"loss": 0.0004,
|
| 2971 |
+
"step": 4180
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 4.2,
|
| 2975 |
+
"eval_accuracy": 0.9985528219971056,
|
| 2976 |
+
"eval_loss": 0.004572847858071327,
|
| 2977 |
+
"eval_runtime": 792.1261,
|
| 2978 |
+
"eval_samples_per_second": 0.872,
|
| 2979 |
+
"eval_steps_per_second": 0.218,
|
| 2980 |
+
"step": 4180
|
| 2981 |
+
},
|
| 2982 |
+
{
|
| 2983 |
+
"epoch": 4.2,
|
| 2984 |
+
"step": 4180,
|
| 2985 |
+
"total_flos": 2.0835872156866314e+19,
|
| 2986 |
+
"train_loss": 0.2077333271433265,
|
| 2987 |
+
"train_runtime": 25357.3809,
|
| 2988 |
+
"train_samples_per_second": 0.659,
|
| 2989 |
+
"train_steps_per_second": 0.165
|
| 2990 |
+
},
|
| 2991 |
+
{
|
| 2992 |
+
"epoch": 4.2,
|
| 2993 |
+
"eval_accuracy": 0.9929988331388565,
|
| 2994 |
+
"eval_loss": 0.025693388655781746,
|
| 2995 |
+
"eval_runtime": 985.7172,
|
| 2996 |
+
"eval_samples_per_second": 0.869,
|
| 2997 |
+
"eval_steps_per_second": 0.218,
|
| 2998 |
+
"step": 4180
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 4.2,
|
| 3002 |
+
"eval_accuracy": 0.9929988331388565,
|
| 3003 |
+
"eval_loss": 0.025693388655781746,
|
| 3004 |
+
"eval_runtime": 981.9678,
|
| 3005 |
+
"eval_samples_per_second": 0.873,
|
| 3006 |
+
"eval_steps_per_second": 0.219,
|
| 3007 |
+
"step": 4180
|
| 3008 |
+
}
|
| 3009 |
+
],
|
| 3010 |
+
"logging_steps": 10,
|
| 3011 |
+
"max_steps": 4180,
|
| 3012 |
+
"num_input_tokens_seen": 0,
|
| 3013 |
+
"num_train_epochs": 9223372036854775807,
|
| 3014 |
+
"save_steps": 500,
|
| 3015 |
+
"total_flos": 2.0835872156866314e+19,
|
| 3016 |
+
"train_batch_size": 4,
|
| 3017 |
+
"trial_name": null,
|
| 3018 |
+
"trial_params": null
|
| 3019 |
+
}
|