Commit
•
cbea244
1
Parent(s):
037b8db
End of training
Browse files- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2049 -0
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 1.6016486531331686e+18,
|
| 4 |
+
"train_loss": 0.9856665741236864,
|
| 5 |
+
"train_runtime": 4486.694,
|
| 6 |
+
"train_samples_per_second": 20.16,
|
| 7 |
+
"train_steps_per_second": 0.63
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 1.6016486531331686e+18,
|
| 4 |
+
"train_loss": 0.9856665741236864,
|
| 5 |
+
"train_runtime": 4486.694,
|
| 6 |
+
"train_samples_per_second": 20.16,
|
| 7 |
+
"train_steps_per_second": 0.63
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2049 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9866785079928952,
|
| 3 |
+
"best_model_checkpoint": "cvt-13-finetuned-ibird/checkpoint-1696",
|
| 4 |
+
"epoch": 4.995579133510168,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2825,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"grad_norm": 3.22133731842041,
|
| 14 |
+
"learning_rate": 1.76678445229682e-06,
|
| 15 |
+
"loss": 3.2244,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.04,
|
| 20 |
+
"grad_norm": 2.790149688720703,
|
| 21 |
+
"learning_rate": 3.53356890459364e-06,
|
| 22 |
+
"loss": 3.2169,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05,
|
| 27 |
+
"grad_norm": 2.872779369354248,
|
| 28 |
+
"learning_rate": 5.30035335689046e-06,
|
| 29 |
+
"loss": 3.2236,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07,
|
| 34 |
+
"grad_norm": 2.464066982269287,
|
| 35 |
+
"learning_rate": 7.06713780918728e-06,
|
| 36 |
+
"loss": 3.2237,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09,
|
| 41 |
+
"grad_norm": 3.8840532302856445,
|
| 42 |
+
"learning_rate": 8.8339222614841e-06,
|
| 43 |
+
"loss": 3.2091,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.11,
|
| 48 |
+
"grad_norm": 2.897418975830078,
|
| 49 |
+
"learning_rate": 1.060070671378092e-05,
|
| 50 |
+
"loss": 3.2041,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12,
|
| 55 |
+
"grad_norm": 2.91304874420166,
|
| 56 |
+
"learning_rate": 1.236749116607774e-05,
|
| 57 |
+
"loss": 3.1712,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14,
|
| 62 |
+
"grad_norm": 3.0000202655792236,
|
| 63 |
+
"learning_rate": 1.413427561837456e-05,
|
| 64 |
+
"loss": 3.1854,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16,
|
| 69 |
+
"grad_norm": 2.865732192993164,
|
| 70 |
+
"learning_rate": 1.5901060070671377e-05,
|
| 71 |
+
"loss": 3.1636,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.18,
|
| 76 |
+
"grad_norm": 3.4034979343414307,
|
| 77 |
+
"learning_rate": 1.76678445229682e-05,
|
| 78 |
+
"loss": 3.128,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.19,
|
| 83 |
+
"grad_norm": 3.288928747177124,
|
| 84 |
+
"learning_rate": 1.9434628975265016e-05,
|
| 85 |
+
"loss": 3.1292,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21,
|
| 90 |
+
"grad_norm": 5.1810078620910645,
|
| 91 |
+
"learning_rate": 2.120141342756184e-05,
|
| 92 |
+
"loss": 3.0734,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23,
|
| 97 |
+
"grad_norm": 2.9931440353393555,
|
| 98 |
+
"learning_rate": 2.296819787985866e-05,
|
| 99 |
+
"loss": 3.0704,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.25,
|
| 104 |
+
"grad_norm": 3.1638436317443848,
|
| 105 |
+
"learning_rate": 2.473498233215548e-05,
|
| 106 |
+
"loss": 3.0107,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.27,
|
| 111 |
+
"grad_norm": 2.532331705093384,
|
| 112 |
+
"learning_rate": 2.6501766784452298e-05,
|
| 113 |
+
"loss": 2.9914,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.28,
|
| 118 |
+
"grad_norm": 3.235107421875,
|
| 119 |
+
"learning_rate": 2.826855123674912e-05,
|
| 120 |
+
"loss": 2.9416,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3,
|
| 125 |
+
"grad_norm": 3.254007339477539,
|
| 126 |
+
"learning_rate": 3.003533568904594e-05,
|
| 127 |
+
"loss": 2.9103,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.32,
|
| 132 |
+
"grad_norm": 3.413905143737793,
|
| 133 |
+
"learning_rate": 3.1802120141342755e-05,
|
| 134 |
+
"loss": 2.8261,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.34,
|
| 139 |
+
"grad_norm": 3.3824758529663086,
|
| 140 |
+
"learning_rate": 3.356890459363958e-05,
|
| 141 |
+
"loss": 2.7711,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.35,
|
| 146 |
+
"grad_norm": 3.0212182998657227,
|
| 147 |
+
"learning_rate": 3.53356890459364e-05,
|
| 148 |
+
"loss": 2.7134,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.37,
|
| 153 |
+
"grad_norm": 3.3175628185272217,
|
| 154 |
+
"learning_rate": 3.710247349823322e-05,
|
| 155 |
+
"loss": 2.5992,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.39,
|
| 160 |
+
"grad_norm": 3.2799651622772217,
|
| 161 |
+
"learning_rate": 3.886925795053003e-05,
|
| 162 |
+
"loss": 2.565,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.41,
|
| 167 |
+
"grad_norm": 5.323554039001465,
|
| 168 |
+
"learning_rate": 4.063604240282686e-05,
|
| 169 |
+
"loss": 2.4975,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.42,
|
| 174 |
+
"grad_norm": 3.1259026527404785,
|
| 175 |
+
"learning_rate": 4.240282685512368e-05,
|
| 176 |
+
"loss": 2.3784,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.44,
|
| 181 |
+
"grad_norm": 3.178640842437744,
|
| 182 |
+
"learning_rate": 4.416961130742049e-05,
|
| 183 |
+
"loss": 2.3291,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.46,
|
| 188 |
+
"grad_norm": 3.33782696723938,
|
| 189 |
+
"learning_rate": 4.593639575971732e-05,
|
| 190 |
+
"loss": 2.2431,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.48,
|
| 195 |
+
"grad_norm": 3.880281686782837,
|
| 196 |
+
"learning_rate": 4.7703180212014135e-05,
|
| 197 |
+
"loss": 2.135,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5,
|
| 202 |
+
"grad_norm": 3.968682050704956,
|
| 203 |
+
"learning_rate": 4.946996466431096e-05,
|
| 204 |
+
"loss": 2.0128,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.51,
|
| 209 |
+
"grad_norm": 4.611078262329102,
|
| 210 |
+
"learning_rate": 4.9862313139260423e-05,
|
| 211 |
+
"loss": 2.0385,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.53,
|
| 216 |
+
"grad_norm": 3.0588765144348145,
|
| 217 |
+
"learning_rate": 4.9665617623918175e-05,
|
| 218 |
+
"loss": 1.8919,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.55,
|
| 223 |
+
"grad_norm": 3.524935722351074,
|
| 224 |
+
"learning_rate": 4.9468922108575926e-05,
|
| 225 |
+
"loss": 1.8712,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.57,
|
| 230 |
+
"grad_norm": 3.9682562351226807,
|
| 231 |
+
"learning_rate": 4.927222659323368e-05,
|
| 232 |
+
"loss": 1.7009,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.58,
|
| 237 |
+
"grad_norm": 3.5031826496124268,
|
| 238 |
+
"learning_rate": 4.907553107789143e-05,
|
| 239 |
+
"loss": 1.5617,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6,
|
| 244 |
+
"grad_norm": 3.6093549728393555,
|
| 245 |
+
"learning_rate": 4.887883556254917e-05,
|
| 246 |
+
"loss": 1.6847,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.62,
|
| 251 |
+
"grad_norm": 4.248983383178711,
|
| 252 |
+
"learning_rate": 4.8682140047206924e-05,
|
| 253 |
+
"loss": 1.5254,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.64,
|
| 258 |
+
"grad_norm": 3.7051937580108643,
|
| 259 |
+
"learning_rate": 4.8485444531864675e-05,
|
| 260 |
+
"loss": 1.4551,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.65,
|
| 265 |
+
"grad_norm": 3.0025603771209717,
|
| 266 |
+
"learning_rate": 4.8288749016522426e-05,
|
| 267 |
+
"loss": 1.5533,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.67,
|
| 272 |
+
"grad_norm": 3.4976930618286133,
|
| 273 |
+
"learning_rate": 4.809205350118017e-05,
|
| 274 |
+
"loss": 1.3265,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.69,
|
| 279 |
+
"grad_norm": 3.950221061706543,
|
| 280 |
+
"learning_rate": 4.789535798583792e-05,
|
| 281 |
+
"loss": 1.2725,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.71,
|
| 286 |
+
"grad_norm": 4.968952178955078,
|
| 287 |
+
"learning_rate": 4.769866247049567e-05,
|
| 288 |
+
"loss": 1.2879,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.73,
|
| 293 |
+
"grad_norm": 3.3156518936157227,
|
| 294 |
+
"learning_rate": 4.7501966955153424e-05,
|
| 295 |
+
"loss": 1.1826,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.74,
|
| 300 |
+
"grad_norm": 3.190901756286621,
|
| 301 |
+
"learning_rate": 4.7305271439811175e-05,
|
| 302 |
+
"loss": 1.1947,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.76,
|
| 307 |
+
"grad_norm": 2.0758602619171143,
|
| 308 |
+
"learning_rate": 4.7108575924468926e-05,
|
| 309 |
+
"loss": 1.2037,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.78,
|
| 314 |
+
"grad_norm": 2.5758004188537598,
|
| 315 |
+
"learning_rate": 4.691188040912668e-05,
|
| 316 |
+
"loss": 1.1067,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8,
|
| 321 |
+
"grad_norm": 3.4717776775360107,
|
| 322 |
+
"learning_rate": 4.671518489378442e-05,
|
| 323 |
+
"loss": 1.2233,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.81,
|
| 328 |
+
"grad_norm": 2.4771013259887695,
|
| 329 |
+
"learning_rate": 4.651848937844217e-05,
|
| 330 |
+
"loss": 1.0819,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.83,
|
| 335 |
+
"grad_norm": 2.409405469894409,
|
| 336 |
+
"learning_rate": 4.6321793863099924e-05,
|
| 337 |
+
"loss": 1.2001,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.85,
|
| 342 |
+
"grad_norm": 2.6338610649108887,
|
| 343 |
+
"learning_rate": 4.6125098347757675e-05,
|
| 344 |
+
"loss": 1.0693,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.87,
|
| 349 |
+
"grad_norm": 3.0745251178741455,
|
| 350 |
+
"learning_rate": 4.5928402832415426e-05,
|
| 351 |
+
"loss": 1.1109,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.88,
|
| 356 |
+
"grad_norm": 2.047109842300415,
|
| 357 |
+
"learning_rate": 4.573170731707318e-05,
|
| 358 |
+
"loss": 1.0295,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9,
|
| 363 |
+
"grad_norm": 4.1728410720825195,
|
| 364 |
+
"learning_rate": 4.553501180173092e-05,
|
| 365 |
+
"loss": 1.0533,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.92,
|
| 370 |
+
"grad_norm": 3.705878973007202,
|
| 371 |
+
"learning_rate": 4.533831628638867e-05,
|
| 372 |
+
"loss": 1.0001,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.94,
|
| 377 |
+
"grad_norm": 2.719705820083618,
|
| 378 |
+
"learning_rate": 4.5141620771046424e-05,
|
| 379 |
+
"loss": 1.0657,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.95,
|
| 384 |
+
"grad_norm": 3.654970169067383,
|
| 385 |
+
"learning_rate": 4.4944925255704175e-05,
|
| 386 |
+
"loss": 0.9326,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.97,
|
| 391 |
+
"grad_norm": 5.275374889373779,
|
| 392 |
+
"learning_rate": 4.4748229740361926e-05,
|
| 393 |
+
"loss": 1.0006,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.99,
|
| 398 |
+
"grad_norm": 3.071272373199463,
|
| 399 |
+
"learning_rate": 4.455153422501967e-05,
|
| 400 |
+
"loss": 1.0089,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.0,
|
| 405 |
+
"eval_accuracy": 0.9649200710479574,
|
| 406 |
+
"eval_loss": 0.2877684533596039,
|
| 407 |
+
"eval_runtime": 86.5125,
|
| 408 |
+
"eval_samples_per_second": 26.031,
|
| 409 |
+
"eval_steps_per_second": 3.26,
|
| 410 |
+
"step": 565
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.01,
|
| 414 |
+
"grad_norm": 3.3852458000183105,
|
| 415 |
+
"learning_rate": 4.435483870967742e-05,
|
| 416 |
+
"loss": 0.865,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.03,
|
| 421 |
+
"grad_norm": 2.0667455196380615,
|
| 422 |
+
"learning_rate": 4.415814319433517e-05,
|
| 423 |
+
"loss": 0.9117,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.04,
|
| 428 |
+
"grad_norm": 1.9478442668914795,
|
| 429 |
+
"learning_rate": 4.3961447678992924e-05,
|
| 430 |
+
"loss": 0.8391,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.06,
|
| 435 |
+
"grad_norm": 0.9961579442024231,
|
| 436 |
+
"learning_rate": 4.376475216365067e-05,
|
| 437 |
+
"loss": 0.8528,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.08,
|
| 442 |
+
"grad_norm": 3.6760354042053223,
|
| 443 |
+
"learning_rate": 4.356805664830842e-05,
|
| 444 |
+
"loss": 0.8387,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.1,
|
| 449 |
+
"grad_norm": 3.119180917739868,
|
| 450 |
+
"learning_rate": 4.337136113296617e-05,
|
| 451 |
+
"loss": 0.7549,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.11,
|
| 456 |
+
"grad_norm": 3.2463724613189697,
|
| 457 |
+
"learning_rate": 4.317466561762392e-05,
|
| 458 |
+
"loss": 0.9329,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.13,
|
| 463 |
+
"grad_norm": 1.7437245845794678,
|
| 464 |
+
"learning_rate": 4.297797010228167e-05,
|
| 465 |
+
"loss": 0.824,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.15,
|
| 470 |
+
"grad_norm": 2.5226683616638184,
|
| 471 |
+
"learning_rate": 4.278127458693942e-05,
|
| 472 |
+
"loss": 0.8805,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.17,
|
| 477 |
+
"grad_norm": 2.30314302444458,
|
| 478 |
+
"learning_rate": 4.258457907159717e-05,
|
| 479 |
+
"loss": 0.9097,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.18,
|
| 484 |
+
"grad_norm": 2.0539016723632812,
|
| 485 |
+
"learning_rate": 4.238788355625492e-05,
|
| 486 |
+
"loss": 0.8936,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.2,
|
| 491 |
+
"grad_norm": 3.0018861293792725,
|
| 492 |
+
"learning_rate": 4.219118804091267e-05,
|
| 493 |
+
"loss": 0.9214,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.22,
|
| 498 |
+
"grad_norm": 2.0286073684692383,
|
| 499 |
+
"learning_rate": 4.1994492525570416e-05,
|
| 500 |
+
"loss": 0.7991,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.24,
|
| 505 |
+
"grad_norm": 0.8428072929382324,
|
| 506 |
+
"learning_rate": 4.179779701022817e-05,
|
| 507 |
+
"loss": 0.756,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.26,
|
| 512 |
+
"grad_norm": 1.7304598093032837,
|
| 513 |
+
"learning_rate": 4.160110149488592e-05,
|
| 514 |
+
"loss": 0.7738,
|
| 515 |
+
"step": 710
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.27,
|
| 519 |
+
"grad_norm": 3.970334768295288,
|
| 520 |
+
"learning_rate": 4.140440597954367e-05,
|
| 521 |
+
"loss": 0.7426,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.29,
|
| 526 |
+
"grad_norm": 2.3020598888397217,
|
| 527 |
+
"learning_rate": 4.1207710464201413e-05,
|
| 528 |
+
"loss": 0.7884,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.31,
|
| 533 |
+
"grad_norm": 1.7566787004470825,
|
| 534 |
+
"learning_rate": 4.1011014948859165e-05,
|
| 535 |
+
"loss": 0.7569,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.33,
|
| 540 |
+
"grad_norm": 3.269785165786743,
|
| 541 |
+
"learning_rate": 4.0814319433516916e-05,
|
| 542 |
+
"loss": 0.8191,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.34,
|
| 547 |
+
"grad_norm": 2.191699504852295,
|
| 548 |
+
"learning_rate": 4.061762391817467e-05,
|
| 549 |
+
"loss": 0.8077,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.36,
|
| 554 |
+
"grad_norm": 1.7478408813476562,
|
| 555 |
+
"learning_rate": 4.042092840283242e-05,
|
| 556 |
+
"loss": 0.7621,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.38,
|
| 561 |
+
"grad_norm": 2.7008702754974365,
|
| 562 |
+
"learning_rate": 4.022423288749016e-05,
|
| 563 |
+
"loss": 0.81,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.4,
|
| 568 |
+
"grad_norm": 1.9396083354949951,
|
| 569 |
+
"learning_rate": 4.0027537372147914e-05,
|
| 570 |
+
"loss": 0.7493,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.41,
|
| 575 |
+
"grad_norm": 2.55366587638855,
|
| 576 |
+
"learning_rate": 3.9830841856805665e-05,
|
| 577 |
+
"loss": 0.7471,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.43,
|
| 582 |
+
"grad_norm": 1.6567083597183228,
|
| 583 |
+
"learning_rate": 3.9634146341463416e-05,
|
| 584 |
+
"loss": 0.9089,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.45,
|
| 589 |
+
"grad_norm": 4.915744781494141,
|
| 590 |
+
"learning_rate": 3.943745082612117e-05,
|
| 591 |
+
"loss": 0.8211,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.47,
|
| 596 |
+
"grad_norm": 1.2864570617675781,
|
| 597 |
+
"learning_rate": 3.924075531077892e-05,
|
| 598 |
+
"loss": 0.7662,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.49,
|
| 603 |
+
"grad_norm": 1.4654157161712646,
|
| 604 |
+
"learning_rate": 3.904405979543666e-05,
|
| 605 |
+
"loss": 0.7942,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.5,
|
| 610 |
+
"grad_norm": 5.769187927246094,
|
| 611 |
+
"learning_rate": 3.8847364280094414e-05,
|
| 612 |
+
"loss": 0.73,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.52,
|
| 617 |
+
"grad_norm": 0.6875308156013489,
|
| 618 |
+
"learning_rate": 3.8650668764752165e-05,
|
| 619 |
+
"loss": 0.7394,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.54,
|
| 624 |
+
"grad_norm": 1.2160513401031494,
|
| 625 |
+
"learning_rate": 3.8453973249409916e-05,
|
| 626 |
+
"loss": 0.846,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.56,
|
| 631 |
+
"grad_norm": 1.4410216808319092,
|
| 632 |
+
"learning_rate": 3.825727773406767e-05,
|
| 633 |
+
"loss": 0.7366,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.57,
|
| 638 |
+
"grad_norm": 4.212517261505127,
|
| 639 |
+
"learning_rate": 3.806058221872542e-05,
|
| 640 |
+
"loss": 0.8275,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.59,
|
| 645 |
+
"grad_norm": 0.9933316111564636,
|
| 646 |
+
"learning_rate": 3.786388670338317e-05,
|
| 647 |
+
"loss": 0.7688,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.61,
|
| 652 |
+
"grad_norm": 2.2545719146728516,
|
| 653 |
+
"learning_rate": 3.7667191188040914e-05,
|
| 654 |
+
"loss": 0.734,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.63,
|
| 659 |
+
"grad_norm": 1.0783274173736572,
|
| 660 |
+
"learning_rate": 3.7470495672698665e-05,
|
| 661 |
+
"loss": 0.8099,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.64,
|
| 666 |
+
"grad_norm": 1.6741780042648315,
|
| 667 |
+
"learning_rate": 3.7273800157356416e-05,
|
| 668 |
+
"loss": 0.9695,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.66,
|
| 673 |
+
"grad_norm": 3.448066473007202,
|
| 674 |
+
"learning_rate": 3.707710464201417e-05,
|
| 675 |
+
"loss": 0.7506,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.68,
|
| 680 |
+
"grad_norm": 9.674800872802734,
|
| 681 |
+
"learning_rate": 3.688040912667191e-05,
|
| 682 |
+
"loss": 0.6566,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.7,
|
| 687 |
+
"grad_norm": 6.721800327301025,
|
| 688 |
+
"learning_rate": 3.668371361132966e-05,
|
| 689 |
+
"loss": 0.7075,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.72,
|
| 694 |
+
"grad_norm": 5.588386058807373,
|
| 695 |
+
"learning_rate": 3.6487018095987414e-05,
|
| 696 |
+
"loss": 0.8246,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.73,
|
| 701 |
+
"grad_norm": 1.4187864065170288,
|
| 702 |
+
"learning_rate": 3.6290322580645165e-05,
|
| 703 |
+
"loss": 0.7808,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.75,
|
| 708 |
+
"grad_norm": 0.7722711563110352,
|
| 709 |
+
"learning_rate": 3.6093627065302916e-05,
|
| 710 |
+
"loss": 0.8957,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.77,
|
| 715 |
+
"grad_norm": 1.5007630586624146,
|
| 716 |
+
"learning_rate": 3.589693154996066e-05,
|
| 717 |
+
"loss": 0.7536,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.79,
|
| 722 |
+
"grad_norm": 0.7198464870452881,
|
| 723 |
+
"learning_rate": 3.570023603461841e-05,
|
| 724 |
+
"loss": 0.6727,
|
| 725 |
+
"step": 1010
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.8,
|
| 729 |
+
"grad_norm": 6.015268802642822,
|
| 730 |
+
"learning_rate": 3.550354051927616e-05,
|
| 731 |
+
"loss": 0.737,
|
| 732 |
+
"step": 1020
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.82,
|
| 736 |
+
"grad_norm": 0.5802467465400696,
|
| 737 |
+
"learning_rate": 3.5306845003933914e-05,
|
| 738 |
+
"loss": 0.6929,
|
| 739 |
+
"step": 1030
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.84,
|
| 743 |
+
"grad_norm": 6.546850681304932,
|
| 744 |
+
"learning_rate": 3.511014948859166e-05,
|
| 745 |
+
"loss": 0.7529,
|
| 746 |
+
"step": 1040
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.86,
|
| 750 |
+
"grad_norm": 7.255277156829834,
|
| 751 |
+
"learning_rate": 3.491345397324941e-05,
|
| 752 |
+
"loss": 0.8639,
|
| 753 |
+
"step": 1050
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.87,
|
| 757 |
+
"grad_norm": 0.564946711063385,
|
| 758 |
+
"learning_rate": 3.471675845790716e-05,
|
| 759 |
+
"loss": 0.7234,
|
| 760 |
+
"step": 1060
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.89,
|
| 764 |
+
"grad_norm": 1.2637860774993896,
|
| 765 |
+
"learning_rate": 3.452006294256491e-05,
|
| 766 |
+
"loss": 0.9388,
|
| 767 |
+
"step": 1070
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.91,
|
| 771 |
+
"grad_norm": 1.095790147781372,
|
| 772 |
+
"learning_rate": 3.432336742722266e-05,
|
| 773 |
+
"loss": 0.7637,
|
| 774 |
+
"step": 1080
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.93,
|
| 778 |
+
"grad_norm": 4.213192462921143,
|
| 779 |
+
"learning_rate": 3.412667191188041e-05,
|
| 780 |
+
"loss": 0.7857,
|
| 781 |
+
"step": 1090
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.95,
|
| 785 |
+
"grad_norm": 9.215189933776855,
|
| 786 |
+
"learning_rate": 3.392997639653816e-05,
|
| 787 |
+
"loss": 0.6239,
|
| 788 |
+
"step": 1100
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.96,
|
| 792 |
+
"grad_norm": 2.098184585571289,
|
| 793 |
+
"learning_rate": 3.373328088119591e-05,
|
| 794 |
+
"loss": 0.5597,
|
| 795 |
+
"step": 1110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.98,
|
| 799 |
+
"grad_norm": 1.9442644119262695,
|
| 800 |
+
"learning_rate": 3.353658536585366e-05,
|
| 801 |
+
"loss": 0.7885,
|
| 802 |
+
"step": 1120
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.0,
|
| 806 |
+
"grad_norm": 6.179882526397705,
|
| 807 |
+
"learning_rate": 3.3339889850511406e-05,
|
| 808 |
+
"loss": 0.7855,
|
| 809 |
+
"step": 1130
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 2.0,
|
| 813 |
+
"eval_accuracy": 0.9791296625222025,
|
| 814 |
+
"eval_loss": 0.10786169767379761,
|
| 815 |
+
"eval_runtime": 86.2088,
|
| 816 |
+
"eval_samples_per_second": 26.123,
|
| 817 |
+
"eval_steps_per_second": 3.271,
|
| 818 |
+
"step": 1131
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.02,
|
| 822 |
+
"grad_norm": 0.36537787318229675,
|
| 823 |
+
"learning_rate": 3.314319433516916e-05,
|
| 824 |
+
"loss": 0.8073,
|
| 825 |
+
"step": 1140
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 2.03,
|
| 829 |
+
"grad_norm": 0.9752954840660095,
|
| 830 |
+
"learning_rate": 3.294649881982691e-05,
|
| 831 |
+
"loss": 0.586,
|
| 832 |
+
"step": 1150
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.05,
|
| 836 |
+
"grad_norm": 0.6317528486251831,
|
| 837 |
+
"learning_rate": 3.274980330448466e-05,
|
| 838 |
+
"loss": 0.7136,
|
| 839 |
+
"step": 1160
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 2.07,
|
| 843 |
+
"grad_norm": 0.4661918580532074,
|
| 844 |
+
"learning_rate": 3.255310778914241e-05,
|
| 845 |
+
"loss": 0.7598,
|
| 846 |
+
"step": 1170
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.09,
|
| 850 |
+
"grad_norm": 0.6691815853118896,
|
| 851 |
+
"learning_rate": 3.2356412273800155e-05,
|
| 852 |
+
"loss": 0.9078,
|
| 853 |
+
"step": 1180
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 2.1,
|
| 857 |
+
"grad_norm": 2.7598626613616943,
|
| 858 |
+
"learning_rate": 3.2159716758457906e-05,
|
| 859 |
+
"loss": 0.5849,
|
| 860 |
+
"step": 1190
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.12,
|
| 864 |
+
"grad_norm": 0.4353080093860626,
|
| 865 |
+
"learning_rate": 3.196302124311566e-05,
|
| 866 |
+
"loss": 0.8519,
|
| 867 |
+
"step": 1200
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 2.14,
|
| 871 |
+
"grad_norm": 0.5575789213180542,
|
| 872 |
+
"learning_rate": 3.176632572777341e-05,
|
| 873 |
+
"loss": 0.7124,
|
| 874 |
+
"step": 1210
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 2.16,
|
| 878 |
+
"grad_norm": 2.028132677078247,
|
| 879 |
+
"learning_rate": 3.156963021243116e-05,
|
| 880 |
+
"loss": 0.8845,
|
| 881 |
+
"step": 1220
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 2.18,
|
| 885 |
+
"grad_norm": 2.471712350845337,
|
| 886 |
+
"learning_rate": 3.137293469708891e-05,
|
| 887 |
+
"loss": 0.7683,
|
| 888 |
+
"step": 1230
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.19,
|
| 892 |
+
"grad_norm": 3.2386505603790283,
|
| 893 |
+
"learning_rate": 3.1176239181746655e-05,
|
| 894 |
+
"loss": 0.6573,
|
| 895 |
+
"step": 1240
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 2.21,
|
| 899 |
+
"grad_norm": 0.6279830932617188,
|
| 900 |
+
"learning_rate": 3.0979543666404406e-05,
|
| 901 |
+
"loss": 0.6261,
|
| 902 |
+
"step": 1250
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.23,
|
| 906 |
+
"grad_norm": 1.3760743141174316,
|
| 907 |
+
"learning_rate": 3.078284815106216e-05,
|
| 908 |
+
"loss": 0.8395,
|
| 909 |
+
"step": 1260
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 2.25,
|
| 913 |
+
"grad_norm": 0.6067408919334412,
|
| 914 |
+
"learning_rate": 3.058615263571991e-05,
|
| 915 |
+
"loss": 0.6631,
|
| 916 |
+
"step": 1270
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 2.26,
|
| 920 |
+
"grad_norm": 1.8222625255584717,
|
| 921 |
+
"learning_rate": 3.0389457120377656e-05,
|
| 922 |
+
"loss": 0.7618,
|
| 923 |
+
"step": 1280
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.28,
|
| 927 |
+
"grad_norm": 0.4705863296985626,
|
| 928 |
+
"learning_rate": 3.0192761605035407e-05,
|
| 929 |
+
"loss": 0.657,
|
| 930 |
+
"step": 1290
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 2.3,
|
| 934 |
+
"grad_norm": 0.885718047618866,
|
| 935 |
+
"learning_rate": 2.999606608969316e-05,
|
| 936 |
+
"loss": 0.6618,
|
| 937 |
+
"step": 1300
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 2.32,
|
| 941 |
+
"grad_norm": 0.5552624464035034,
|
| 942 |
+
"learning_rate": 2.9799370574350903e-05,
|
| 943 |
+
"loss": 0.7135,
|
| 944 |
+
"step": 1310
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.33,
|
| 948 |
+
"grad_norm": 0.8557870388031006,
|
| 949 |
+
"learning_rate": 2.9602675059008654e-05,
|
| 950 |
+
"loss": 0.6492,
|
| 951 |
+
"step": 1320
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 2.35,
|
| 955 |
+
"grad_norm": 1.226236343383789,
|
| 956 |
+
"learning_rate": 2.9405979543666405e-05,
|
| 957 |
+
"loss": 0.5645,
|
| 958 |
+
"step": 1330
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 2.37,
|
| 962 |
+
"grad_norm": 0.4508123993873596,
|
| 963 |
+
"learning_rate": 2.9209284028324156e-05,
|
| 964 |
+
"loss": 0.653,
|
| 965 |
+
"step": 1340
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 2.39,
|
| 969 |
+
"grad_norm": 5.647830963134766,
|
| 970 |
+
"learning_rate": 2.9012588512981904e-05,
|
| 971 |
+
"loss": 0.6795,
|
| 972 |
+
"step": 1350
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.4,
|
| 976 |
+
"grad_norm": 0.6091474294662476,
|
| 977 |
+
"learning_rate": 2.8815892997639655e-05,
|
| 978 |
+
"loss": 0.7905,
|
| 979 |
+
"step": 1360
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 2.42,
|
| 983 |
+
"grad_norm": 0.5606883764266968,
|
| 984 |
+
"learning_rate": 2.8619197482297406e-05,
|
| 985 |
+
"loss": 0.6878,
|
| 986 |
+
"step": 1370
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.44,
|
| 990 |
+
"grad_norm": 2.7051727771759033,
|
| 991 |
+
"learning_rate": 2.8422501966955157e-05,
|
| 992 |
+
"loss": 0.6996,
|
| 993 |
+
"step": 1380
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 2.46,
|
| 997 |
+
"grad_norm": 0.7471837401390076,
|
| 998 |
+
"learning_rate": 2.822580645161291e-05,
|
| 999 |
+
"loss": 0.706,
|
| 1000 |
+
"step": 1390
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.48,
|
| 1004 |
+
"grad_norm": 2.322110176086426,
|
| 1005 |
+
"learning_rate": 2.8029110936270653e-05,
|
| 1006 |
+
"loss": 0.9161,
|
| 1007 |
+
"step": 1400
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 2.49,
|
| 1011 |
+
"grad_norm": 0.7188151478767395,
|
| 1012 |
+
"learning_rate": 2.7832415420928404e-05,
|
| 1013 |
+
"loss": 0.6905,
|
| 1014 |
+
"step": 1410
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.51,
|
| 1018 |
+
"grad_norm": 10.193795204162598,
|
| 1019 |
+
"learning_rate": 2.7635719905586155e-05,
|
| 1020 |
+
"loss": 0.7407,
|
| 1021 |
+
"step": 1420
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.53,
|
| 1025 |
+
"grad_norm": 0.9392734169960022,
|
| 1026 |
+
"learning_rate": 2.7439024390243906e-05,
|
| 1027 |
+
"loss": 0.6731,
|
| 1028 |
+
"step": 1430
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.55,
|
| 1032 |
+
"grad_norm": 0.5180323719978333,
|
| 1033 |
+
"learning_rate": 2.724232887490165e-05,
|
| 1034 |
+
"loss": 0.5429,
|
| 1035 |
+
"step": 1440
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 2.56,
|
| 1039 |
+
"grad_norm": 0.4639306664466858,
|
| 1040 |
+
"learning_rate": 2.7045633359559402e-05,
|
| 1041 |
+
"loss": 0.7713,
|
| 1042 |
+
"step": 1450
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 2.58,
|
| 1046 |
+
"grad_norm": 3.5896573066711426,
|
| 1047 |
+
"learning_rate": 2.6848937844217153e-05,
|
| 1048 |
+
"loss": 0.7773,
|
| 1049 |
+
"step": 1460
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 2.6,
|
| 1053 |
+
"grad_norm": 0.7149112820625305,
|
| 1054 |
+
"learning_rate": 2.6652242328874904e-05,
|
| 1055 |
+
"loss": 0.6672,
|
| 1056 |
+
"step": 1470
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.62,
|
| 1060 |
+
"grad_norm": 0.8536508679389954,
|
| 1061 |
+
"learning_rate": 2.645554681353265e-05,
|
| 1062 |
+
"loss": 0.7086,
|
| 1063 |
+
"step": 1480
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.63,
|
| 1067 |
+
"grad_norm": 1.054091215133667,
|
| 1068 |
+
"learning_rate": 2.62588512981904e-05,
|
| 1069 |
+
"loss": 0.6519,
|
| 1070 |
+
"step": 1490
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.65,
|
| 1074 |
+
"grad_norm": 0.39519792795181274,
|
| 1075 |
+
"learning_rate": 2.606215578284815e-05,
|
| 1076 |
+
"loss": 0.6434,
|
| 1077 |
+
"step": 1500
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.67,
|
| 1081 |
+
"grad_norm": 0.4575929343700409,
|
| 1082 |
+
"learning_rate": 2.5865460267505902e-05,
|
| 1083 |
+
"loss": 0.6553,
|
| 1084 |
+
"step": 1510
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.69,
|
| 1088 |
+
"grad_norm": 0.534916877746582,
|
| 1089 |
+
"learning_rate": 2.5668764752163653e-05,
|
| 1090 |
+
"loss": 0.7915,
|
| 1091 |
+
"step": 1520
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.71,
|
| 1095 |
+
"grad_norm": 0.7122427821159363,
|
| 1096 |
+
"learning_rate": 2.54720692368214e-05,
|
| 1097 |
+
"loss": 0.6913,
|
| 1098 |
+
"step": 1530
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.72,
|
| 1102 |
+
"grad_norm": 0.49569156765937805,
|
| 1103 |
+
"learning_rate": 2.5275373721479152e-05,
|
| 1104 |
+
"loss": 0.6085,
|
| 1105 |
+
"step": 1540
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.74,
|
| 1109 |
+
"grad_norm": 0.466847687959671,
|
| 1110 |
+
"learning_rate": 2.5078678206136903e-05,
|
| 1111 |
+
"loss": 0.6222,
|
| 1112 |
+
"step": 1550
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.76,
|
| 1116 |
+
"grad_norm": 0.34097394347190857,
|
| 1117 |
+
"learning_rate": 2.488198269079465e-05,
|
| 1118 |
+
"loss": 0.6883,
|
| 1119 |
+
"step": 1560
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.78,
|
| 1123 |
+
"grad_norm": 1.601562261581421,
|
| 1124 |
+
"learning_rate": 2.4685287175452402e-05,
|
| 1125 |
+
"loss": 0.6904,
|
| 1126 |
+
"step": 1570
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.79,
|
| 1130 |
+
"grad_norm": 0.5125943422317505,
|
| 1131 |
+
"learning_rate": 2.448859166011015e-05,
|
| 1132 |
+
"loss": 0.6484,
|
| 1133 |
+
"step": 1580
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.81,
|
| 1137 |
+
"grad_norm": 0.4327985644340515,
|
| 1138 |
+
"learning_rate": 2.42918961447679e-05,
|
| 1139 |
+
"loss": 0.7579,
|
| 1140 |
+
"step": 1590
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.83,
|
| 1144 |
+
"grad_norm": 0.4168800115585327,
|
| 1145 |
+
"learning_rate": 2.4095200629425652e-05,
|
| 1146 |
+
"loss": 0.7012,
|
| 1147 |
+
"step": 1600
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 2.85,
|
| 1151 |
+
"grad_norm": 2.620213270187378,
|
| 1152 |
+
"learning_rate": 2.38985051140834e-05,
|
| 1153 |
+
"loss": 0.6107,
|
| 1154 |
+
"step": 1610
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.86,
|
| 1158 |
+
"grad_norm": 0.47252315282821655,
|
| 1159 |
+
"learning_rate": 2.370180959874115e-05,
|
| 1160 |
+
"loss": 0.63,
|
| 1161 |
+
"step": 1620
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.88,
|
| 1165 |
+
"grad_norm": 3.600597620010376,
|
| 1166 |
+
"learning_rate": 2.35051140833989e-05,
|
| 1167 |
+
"loss": 0.7014,
|
| 1168 |
+
"step": 1630
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.9,
|
| 1172 |
+
"grad_norm": 0.9168177843093872,
|
| 1173 |
+
"learning_rate": 2.330841856805665e-05,
|
| 1174 |
+
"loss": 0.6946,
|
| 1175 |
+
"step": 1640
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.92,
|
| 1179 |
+
"grad_norm": 2.111616611480713,
|
| 1180 |
+
"learning_rate": 2.3111723052714398e-05,
|
| 1181 |
+
"loss": 0.5557,
|
| 1182 |
+
"step": 1650
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.94,
|
| 1186 |
+
"grad_norm": 0.4891892671585083,
|
| 1187 |
+
"learning_rate": 2.291502753737215e-05,
|
| 1188 |
+
"loss": 0.6718,
|
| 1189 |
+
"step": 1660
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.95,
|
| 1193 |
+
"grad_norm": 1.4157081842422485,
|
| 1194 |
+
"learning_rate": 2.2718332022029897e-05,
|
| 1195 |
+
"loss": 0.7549,
|
| 1196 |
+
"step": 1670
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.97,
|
| 1200 |
+
"grad_norm": 0.3661864399909973,
|
| 1201 |
+
"learning_rate": 2.2521636506687648e-05,
|
| 1202 |
+
"loss": 0.8072,
|
| 1203 |
+
"step": 1680
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.99,
|
| 1207 |
+
"grad_norm": 2.6403448581695557,
|
| 1208 |
+
"learning_rate": 2.2324940991345396e-05,
|
| 1209 |
+
"loss": 0.6657,
|
| 1210 |
+
"step": 1690
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 3.0,
|
| 1214 |
+
"eval_accuracy": 0.9866785079928952,
|
| 1215 |
+
"eval_loss": 0.07058515399694443,
|
| 1216 |
+
"eval_runtime": 87.367,
|
| 1217 |
+
"eval_samples_per_second": 25.776,
|
| 1218 |
+
"eval_steps_per_second": 3.228,
|
| 1219 |
+
"step": 1696
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 3.01,
|
| 1223 |
+
"grad_norm": 0.5366094708442688,
|
| 1224 |
+
"learning_rate": 2.2128245476003147e-05,
|
| 1225 |
+
"loss": 0.7239,
|
| 1226 |
+
"step": 1700
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 3.02,
|
| 1230 |
+
"grad_norm": 0.5593940019607544,
|
| 1231 |
+
"learning_rate": 2.1931549960660898e-05,
|
| 1232 |
+
"loss": 0.8491,
|
| 1233 |
+
"step": 1710
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 3.04,
|
| 1237 |
+
"grad_norm": 0.2935149371623993,
|
| 1238 |
+
"learning_rate": 2.1734854445318646e-05,
|
| 1239 |
+
"loss": 0.6358,
|
| 1240 |
+
"step": 1720
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 3.06,
|
| 1244 |
+
"grad_norm": 0.3379581570625305,
|
| 1245 |
+
"learning_rate": 2.1538158929976397e-05,
|
| 1246 |
+
"loss": 0.6124,
|
| 1247 |
+
"step": 1730
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 3.08,
|
| 1251 |
+
"grad_norm": 0.6903249621391296,
|
| 1252 |
+
"learning_rate": 2.134146341463415e-05,
|
| 1253 |
+
"loss": 0.7021,
|
| 1254 |
+
"step": 1740
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 3.09,
|
| 1258 |
+
"grad_norm": 0.5648875832557678,
|
| 1259 |
+
"learning_rate": 2.11447678992919e-05,
|
| 1260 |
+
"loss": 0.7315,
|
| 1261 |
+
"step": 1750
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 3.11,
|
| 1265 |
+
"grad_norm": 3.177462100982666,
|
| 1266 |
+
"learning_rate": 2.0948072383949647e-05,
|
| 1267 |
+
"loss": 0.7836,
|
| 1268 |
+
"step": 1760
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 3.13,
|
| 1272 |
+
"grad_norm": 2.3751437664031982,
|
| 1273 |
+
"learning_rate": 2.07513768686074e-05,
|
| 1274 |
+
"loss": 0.5984,
|
| 1275 |
+
"step": 1770
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 3.15,
|
| 1279 |
+
"grad_norm": 0.39180535078048706,
|
| 1280 |
+
"learning_rate": 2.0554681353265146e-05,
|
| 1281 |
+
"loss": 0.6462,
|
| 1282 |
+
"step": 1780
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 3.17,
|
| 1286 |
+
"grad_norm": 0.7616192102432251,
|
| 1287 |
+
"learning_rate": 2.0357985837922897e-05,
|
| 1288 |
+
"loss": 0.648,
|
| 1289 |
+
"step": 1790
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 3.18,
|
| 1293 |
+
"grad_norm": 0.7866030931472778,
|
| 1294 |
+
"learning_rate": 2.0161290322580645e-05,
|
| 1295 |
+
"loss": 0.6992,
|
| 1296 |
+
"step": 1800
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 3.2,
|
| 1300 |
+
"grad_norm": 2.5894863605499268,
|
| 1301 |
+
"learning_rate": 1.9964594807238396e-05,
|
| 1302 |
+
"loss": 0.7294,
|
| 1303 |
+
"step": 1810
|
| 1304 |
+
},
|
| 1305 |
+
{
|
| 1306 |
+
"epoch": 3.22,
|
| 1307 |
+
"grad_norm": 0.33011573553085327,
|
| 1308 |
+
"learning_rate": 1.9767899291896147e-05,
|
| 1309 |
+
"loss": 0.6715,
|
| 1310 |
+
"step": 1820
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 3.24,
|
| 1314 |
+
"grad_norm": 1.21015465259552,
|
| 1315 |
+
"learning_rate": 1.9571203776553895e-05,
|
| 1316 |
+
"loss": 0.6617,
|
| 1317 |
+
"step": 1830
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 3.25,
|
| 1321 |
+
"grad_norm": 0.3521706461906433,
|
| 1322 |
+
"learning_rate": 1.9374508261211646e-05,
|
| 1323 |
+
"loss": 0.7361,
|
| 1324 |
+
"step": 1840
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 3.27,
|
| 1328 |
+
"grad_norm": 1.986220359802246,
|
| 1329 |
+
"learning_rate": 1.9177812745869394e-05,
|
| 1330 |
+
"loss": 0.782,
|
| 1331 |
+
"step": 1850
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 3.29,
|
| 1335 |
+
"grad_norm": 0.6578150391578674,
|
| 1336 |
+
"learning_rate": 1.8981117230527145e-05,
|
| 1337 |
+
"loss": 0.6532,
|
| 1338 |
+
"step": 1860
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 3.31,
|
| 1342 |
+
"grad_norm": 0.5501819252967834,
|
| 1343 |
+
"learning_rate": 1.8784421715184893e-05,
|
| 1344 |
+
"loss": 0.5756,
|
| 1345 |
+
"step": 1870
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 3.32,
|
| 1349 |
+
"grad_norm": 1.0883512496948242,
|
| 1350 |
+
"learning_rate": 1.8587726199842644e-05,
|
| 1351 |
+
"loss": 0.6532,
|
| 1352 |
+
"step": 1880
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.34,
|
| 1356 |
+
"grad_norm": 0.49157479405403137,
|
| 1357 |
+
"learning_rate": 1.8391030684500392e-05,
|
| 1358 |
+
"loss": 0.6951,
|
| 1359 |
+
"step": 1890
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 3.36,
|
| 1363 |
+
"grad_norm": 0.6763016581535339,
|
| 1364 |
+
"learning_rate": 1.8194335169158143e-05,
|
| 1365 |
+
"loss": 0.5232,
|
| 1366 |
+
"step": 1900
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 3.38,
|
| 1370 |
+
"grad_norm": 1.0173105001449585,
|
| 1371 |
+
"learning_rate": 1.799763965381589e-05,
|
| 1372 |
+
"loss": 0.6729,
|
| 1373 |
+
"step": 1910
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 3.4,
|
| 1377 |
+
"grad_norm": 0.49198317527770996,
|
| 1378 |
+
"learning_rate": 1.7800944138473642e-05,
|
| 1379 |
+
"loss": 0.6914,
|
| 1380 |
+
"step": 1920
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 3.41,
|
| 1384 |
+
"grad_norm": 0.7755669355392456,
|
| 1385 |
+
"learning_rate": 1.7604248623131393e-05,
|
| 1386 |
+
"loss": 0.6728,
|
| 1387 |
+
"step": 1930
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 3.43,
|
| 1391 |
+
"grad_norm": 0.4817189574241638,
|
| 1392 |
+
"learning_rate": 1.7407553107789144e-05,
|
| 1393 |
+
"loss": 0.7147,
|
| 1394 |
+
"step": 1940
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 3.45,
|
| 1398 |
+
"grad_norm": 1.0511811971664429,
|
| 1399 |
+
"learning_rate": 1.7210857592446896e-05,
|
| 1400 |
+
"loss": 0.7447,
|
| 1401 |
+
"step": 1950
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 3.47,
|
| 1405 |
+
"grad_norm": 0.4621807038784027,
|
| 1406 |
+
"learning_rate": 1.7014162077104643e-05,
|
| 1407 |
+
"loss": 0.7905,
|
| 1408 |
+
"step": 1960
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 3.48,
|
| 1412 |
+
"grad_norm": 0.42619049549102783,
|
| 1413 |
+
"learning_rate": 1.6817466561762395e-05,
|
| 1414 |
+
"loss": 0.6953,
|
| 1415 |
+
"step": 1970
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 3.5,
|
| 1419 |
+
"grad_norm": 0.4883289933204651,
|
| 1420 |
+
"learning_rate": 1.6620771046420142e-05,
|
| 1421 |
+
"loss": 0.7246,
|
| 1422 |
+
"step": 1980
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 3.52,
|
| 1426 |
+
"grad_norm": 5.140800476074219,
|
| 1427 |
+
"learning_rate": 1.6424075531077893e-05,
|
| 1428 |
+
"loss": 0.677,
|
| 1429 |
+
"step": 1990
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 3.54,
|
| 1433 |
+
"grad_norm": 0.7420614361763,
|
| 1434 |
+
"learning_rate": 1.622738001573564e-05,
|
| 1435 |
+
"loss": 0.6223,
|
| 1436 |
+
"step": 2000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 3.55,
|
| 1440 |
+
"grad_norm": 0.5618042945861816,
|
| 1441 |
+
"learning_rate": 1.6030684500393392e-05,
|
| 1442 |
+
"loss": 0.7706,
|
| 1443 |
+
"step": 2010
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 3.57,
|
| 1447 |
+
"grad_norm": 8.696101188659668,
|
| 1448 |
+
"learning_rate": 1.583398898505114e-05,
|
| 1449 |
+
"loss": 0.5982,
|
| 1450 |
+
"step": 2020
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 3.59,
|
| 1454 |
+
"grad_norm": 0.613085925579071,
|
| 1455 |
+
"learning_rate": 1.563729346970889e-05,
|
| 1456 |
+
"loss": 0.5607,
|
| 1457 |
+
"step": 2030
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 3.61,
|
| 1461 |
+
"grad_norm": 0.37136101722717285,
|
| 1462 |
+
"learning_rate": 1.5440597954366642e-05,
|
| 1463 |
+
"loss": 0.7066,
|
| 1464 |
+
"step": 2040
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 3.63,
|
| 1468 |
+
"grad_norm": 1.0370784997940063,
|
| 1469 |
+
"learning_rate": 1.524390243902439e-05,
|
| 1470 |
+
"loss": 0.6117,
|
| 1471 |
+
"step": 2050
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 3.64,
|
| 1475 |
+
"grad_norm": 3.805783748626709,
|
| 1476 |
+
"learning_rate": 1.5047206923682141e-05,
|
| 1477 |
+
"loss": 0.7247,
|
| 1478 |
+
"step": 2060
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 3.66,
|
| 1482 |
+
"grad_norm": 0.3479630947113037,
|
| 1483 |
+
"learning_rate": 1.485051140833989e-05,
|
| 1484 |
+
"loss": 0.7094,
|
| 1485 |
+
"step": 2070
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 3.68,
|
| 1489 |
+
"grad_norm": 0.27747422456741333,
|
| 1490 |
+
"learning_rate": 1.465381589299764e-05,
|
| 1491 |
+
"loss": 0.6548,
|
| 1492 |
+
"step": 2080
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 3.7,
|
| 1496 |
+
"grad_norm": 0.4813224971294403,
|
| 1497 |
+
"learning_rate": 1.445712037765539e-05,
|
| 1498 |
+
"loss": 0.5747,
|
| 1499 |
+
"step": 2090
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 3.71,
|
| 1503 |
+
"grad_norm": 1.6526601314544678,
|
| 1504 |
+
"learning_rate": 1.4260424862313141e-05,
|
| 1505 |
+
"loss": 0.8374,
|
| 1506 |
+
"step": 2100
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 3.73,
|
| 1510 |
+
"grad_norm": 0.6736083626747131,
|
| 1511 |
+
"learning_rate": 1.4063729346970889e-05,
|
| 1512 |
+
"loss": 0.7677,
|
| 1513 |
+
"step": 2110
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 3.75,
|
| 1517 |
+
"grad_norm": 1.8762106895446777,
|
| 1518 |
+
"learning_rate": 1.386703383162864e-05,
|
| 1519 |
+
"loss": 0.5897,
|
| 1520 |
+
"step": 2120
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 3.77,
|
| 1524 |
+
"grad_norm": 0.622199296951294,
|
| 1525 |
+
"learning_rate": 1.3670338316286388e-05,
|
| 1526 |
+
"loss": 0.7407,
|
| 1527 |
+
"step": 2130
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 3.78,
|
| 1531 |
+
"grad_norm": 0.3692738115787506,
|
| 1532 |
+
"learning_rate": 1.3473642800944139e-05,
|
| 1533 |
+
"loss": 0.8474,
|
| 1534 |
+
"step": 2140
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 3.8,
|
| 1538 |
+
"grad_norm": 0.5641698241233826,
|
| 1539 |
+
"learning_rate": 1.327694728560189e-05,
|
| 1540 |
+
"loss": 0.6674,
|
| 1541 |
+
"step": 2150
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 3.82,
|
| 1545 |
+
"grad_norm": 0.7227357029914856,
|
| 1546 |
+
"learning_rate": 1.3080251770259638e-05,
|
| 1547 |
+
"loss": 0.6852,
|
| 1548 |
+
"step": 2160
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 3.84,
|
| 1552 |
+
"grad_norm": 0.80736243724823,
|
| 1553 |
+
"learning_rate": 1.2883556254917389e-05,
|
| 1554 |
+
"loss": 0.7686,
|
| 1555 |
+
"step": 2170
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 3.85,
|
| 1559 |
+
"grad_norm": 0.7187584042549133,
|
| 1560 |
+
"learning_rate": 1.2686860739575138e-05,
|
| 1561 |
+
"loss": 0.7859,
|
| 1562 |
+
"step": 2180
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 3.87,
|
| 1566 |
+
"grad_norm": 0.6288447976112366,
|
| 1567 |
+
"learning_rate": 1.249016522423289e-05,
|
| 1568 |
+
"loss": 0.683,
|
| 1569 |
+
"step": 2190
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 3.89,
|
| 1573 |
+
"grad_norm": 0.6657484769821167,
|
| 1574 |
+
"learning_rate": 1.2293469708890639e-05,
|
| 1575 |
+
"loss": 0.6807,
|
| 1576 |
+
"step": 2200
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 3.91,
|
| 1580 |
+
"grad_norm": 0.3658302128314972,
|
| 1581 |
+
"learning_rate": 1.2096774193548388e-05,
|
| 1582 |
+
"loss": 0.6274,
|
| 1583 |
+
"step": 2210
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 3.93,
|
| 1587 |
+
"grad_norm": 0.7069531083106995,
|
| 1588 |
+
"learning_rate": 1.1900078678206138e-05,
|
| 1589 |
+
"loss": 0.7722,
|
| 1590 |
+
"step": 2220
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 3.94,
|
| 1594 |
+
"grad_norm": 0.2362523227930069,
|
| 1595 |
+
"learning_rate": 1.1703383162863887e-05,
|
| 1596 |
+
"loss": 0.7107,
|
| 1597 |
+
"step": 2230
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 3.96,
|
| 1601 |
+
"grad_norm": 0.49751946330070496,
|
| 1602 |
+
"learning_rate": 1.1506687647521637e-05,
|
| 1603 |
+
"loss": 0.6972,
|
| 1604 |
+
"step": 2240
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 3.98,
|
| 1608 |
+
"grad_norm": 0.4281612038612366,
|
| 1609 |
+
"learning_rate": 1.1309992132179386e-05,
|
| 1610 |
+
"loss": 0.6935,
|
| 1611 |
+
"step": 2250
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 4.0,
|
| 1615 |
+
"grad_norm": 0.47943803668022156,
|
| 1616 |
+
"learning_rate": 1.1113296616837136e-05,
|
| 1617 |
+
"loss": 0.7475,
|
| 1618 |
+
"step": 2260
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 4.0,
|
| 1622 |
+
"eval_accuracy": 0.9866785079928952,
|
| 1623 |
+
"eval_loss": 0.06677553802728653,
|
| 1624 |
+
"eval_runtime": 87.1331,
|
| 1625 |
+
"eval_samples_per_second": 25.846,
|
| 1626 |
+
"eval_steps_per_second": 3.236,
|
| 1627 |
+
"step": 2262
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.01,
|
| 1631 |
+
"grad_norm": 1.7420241832733154,
|
| 1632 |
+
"learning_rate": 1.0916601101494885e-05,
|
| 1633 |
+
"loss": 0.7495,
|
| 1634 |
+
"step": 2270
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 4.03,
|
| 1638 |
+
"grad_norm": 0.43095558881759644,
|
| 1639 |
+
"learning_rate": 1.0719905586152636e-05,
|
| 1640 |
+
"loss": 0.6875,
|
| 1641 |
+
"step": 2280
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 4.05,
|
| 1645 |
+
"grad_norm": 4.584779739379883,
|
| 1646 |
+
"learning_rate": 1.0523210070810386e-05,
|
| 1647 |
+
"loss": 0.8217,
|
| 1648 |
+
"step": 2290
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 4.07,
|
| 1652 |
+
"grad_norm": 0.3836701810359955,
|
| 1653 |
+
"learning_rate": 1.0326514555468137e-05,
|
| 1654 |
+
"loss": 0.5724,
|
| 1655 |
+
"step": 2300
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 4.08,
|
| 1659 |
+
"grad_norm": 0.8410959243774414,
|
| 1660 |
+
"learning_rate": 1.0129819040125886e-05,
|
| 1661 |
+
"loss": 0.6068,
|
| 1662 |
+
"step": 2310
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 4.1,
|
| 1666 |
+
"grad_norm": 0.2401944100856781,
|
| 1667 |
+
"learning_rate": 9.933123524783636e-06,
|
| 1668 |
+
"loss": 0.6172,
|
| 1669 |
+
"step": 2320
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 4.12,
|
| 1673 |
+
"grad_norm": 1.9553325176239014,
|
| 1674 |
+
"learning_rate": 9.736428009441385e-06,
|
| 1675 |
+
"loss": 0.5384,
|
| 1676 |
+
"step": 2330
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 4.14,
|
| 1680 |
+
"grad_norm": 0.7399584650993347,
|
| 1681 |
+
"learning_rate": 9.539732494099135e-06,
|
| 1682 |
+
"loss": 0.6562,
|
| 1683 |
+
"step": 2340
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 4.16,
|
| 1687 |
+
"grad_norm": 0.6674773097038269,
|
| 1688 |
+
"learning_rate": 9.343036978756884e-06,
|
| 1689 |
+
"loss": 0.736,
|
| 1690 |
+
"step": 2350
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 4.17,
|
| 1694 |
+
"grad_norm": 0.35315585136413574,
|
| 1695 |
+
"learning_rate": 9.146341463414634e-06,
|
| 1696 |
+
"loss": 0.6334,
|
| 1697 |
+
"step": 2360
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 4.19,
|
| 1701 |
+
"grad_norm": 0.8621203303337097,
|
| 1702 |
+
"learning_rate": 8.949645948072383e-06,
|
| 1703 |
+
"loss": 0.7214,
|
| 1704 |
+
"step": 2370
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.21,
|
| 1708 |
+
"grad_norm": 6.545629501342773,
|
| 1709 |
+
"learning_rate": 8.752950432730134e-06,
|
| 1710 |
+
"loss": 0.5391,
|
| 1711 |
+
"step": 2380
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 4.23,
|
| 1715 |
+
"grad_norm": 0.5393751859664917,
|
| 1716 |
+
"learning_rate": 8.556254917387884e-06,
|
| 1717 |
+
"loss": 0.6851,
|
| 1718 |
+
"step": 2390
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 4.24,
|
| 1722 |
+
"grad_norm": 0.8293077945709229,
|
| 1723 |
+
"learning_rate": 8.359559402045635e-06,
|
| 1724 |
+
"loss": 0.791,
|
| 1725 |
+
"step": 2400
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 4.26,
|
| 1729 |
+
"grad_norm": 0.44837647676467896,
|
| 1730 |
+
"learning_rate": 8.162863886703385e-06,
|
| 1731 |
+
"loss": 0.5622,
|
| 1732 |
+
"step": 2410
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 4.28,
|
| 1736 |
+
"grad_norm": 0.5060232281684875,
|
| 1737 |
+
"learning_rate": 7.966168371361134e-06,
|
| 1738 |
+
"loss": 0.7254,
|
| 1739 |
+
"step": 2420
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 4.3,
|
| 1743 |
+
"grad_norm": 0.647249698638916,
|
| 1744 |
+
"learning_rate": 7.769472856018883e-06,
|
| 1745 |
+
"loss": 0.5765,
|
| 1746 |
+
"step": 2430
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 4.31,
|
| 1750 |
+
"grad_norm": 0.7065703868865967,
|
| 1751 |
+
"learning_rate": 7.572777340676633e-06,
|
| 1752 |
+
"loss": 0.7337,
|
| 1753 |
+
"step": 2440
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 4.33,
|
| 1757 |
+
"grad_norm": 0.6291216015815735,
|
| 1758 |
+
"learning_rate": 7.376081825334382e-06,
|
| 1759 |
+
"loss": 0.649,
|
| 1760 |
+
"step": 2450
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 4.35,
|
| 1764 |
+
"grad_norm": 0.6098889708518982,
|
| 1765 |
+
"learning_rate": 7.179386309992133e-06,
|
| 1766 |
+
"loss": 0.6712,
|
| 1767 |
+
"step": 2460
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 4.37,
|
| 1771 |
+
"grad_norm": 0.49595367908477783,
|
| 1772 |
+
"learning_rate": 6.982690794649882e-06,
|
| 1773 |
+
"loss": 0.5932,
|
| 1774 |
+
"step": 2470
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 4.39,
|
| 1778 |
+
"grad_norm": 0.34834733605384827,
|
| 1779 |
+
"learning_rate": 6.785995279307632e-06,
|
| 1780 |
+
"loss": 0.5702,
|
| 1781 |
+
"step": 2480
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 4.4,
|
| 1785 |
+
"grad_norm": 0.30979278683662415,
|
| 1786 |
+
"learning_rate": 6.589299763965381e-06,
|
| 1787 |
+
"loss": 0.6423,
|
| 1788 |
+
"step": 2490
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 4.42,
|
| 1792 |
+
"grad_norm": 1.0691258907318115,
|
| 1793 |
+
"learning_rate": 6.392604248623131e-06,
|
| 1794 |
+
"loss": 0.6254,
|
| 1795 |
+
"step": 2500
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 4.44,
|
| 1799 |
+
"grad_norm": 0.6674053072929382,
|
| 1800 |
+
"learning_rate": 6.195908733280882e-06,
|
| 1801 |
+
"loss": 0.6869,
|
| 1802 |
+
"step": 2510
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 4.46,
|
| 1806 |
+
"grad_norm": 0.4127647280693054,
|
| 1807 |
+
"learning_rate": 5.999213217938631e-06,
|
| 1808 |
+
"loss": 0.6614,
|
| 1809 |
+
"step": 2520
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 4.47,
|
| 1813 |
+
"grad_norm": 1.5234615802764893,
|
| 1814 |
+
"learning_rate": 5.802517702596381e-06,
|
| 1815 |
+
"loss": 0.716,
|
| 1816 |
+
"step": 2530
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 4.49,
|
| 1820 |
+
"grad_norm": 0.3525974750518799,
|
| 1821 |
+
"learning_rate": 5.605822187254131e-06,
|
| 1822 |
+
"loss": 0.8085,
|
| 1823 |
+
"step": 2540
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 4.51,
|
| 1827 |
+
"grad_norm": 0.9571001529693604,
|
| 1828 |
+
"learning_rate": 5.40912667191188e-06,
|
| 1829 |
+
"loss": 0.6391,
|
| 1830 |
+
"step": 2550
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 4.53,
|
| 1834 |
+
"grad_norm": 0.5705153942108154,
|
| 1835 |
+
"learning_rate": 5.212431156569631e-06,
|
| 1836 |
+
"loss": 0.5403,
|
| 1837 |
+
"step": 2560
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 4.54,
|
| 1841 |
+
"grad_norm": 0.6829048991203308,
|
| 1842 |
+
"learning_rate": 5.01573564122738e-06,
|
| 1843 |
+
"loss": 0.6538,
|
| 1844 |
+
"step": 2570
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 4.56,
|
| 1848 |
+
"grad_norm": 1.4802197217941284,
|
| 1849 |
+
"learning_rate": 4.81904012588513e-06,
|
| 1850 |
+
"loss": 0.7467,
|
| 1851 |
+
"step": 2580
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 4.58,
|
| 1855 |
+
"grad_norm": 1.3378098011016846,
|
| 1856 |
+
"learning_rate": 4.62234461054288e-06,
|
| 1857 |
+
"loss": 0.7882,
|
| 1858 |
+
"step": 2590
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 4.6,
|
| 1862 |
+
"grad_norm": 0.3483523428440094,
|
| 1863 |
+
"learning_rate": 4.425649095200629e-06,
|
| 1864 |
+
"loss": 0.6253,
|
| 1865 |
+
"step": 2600
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 4.62,
|
| 1869 |
+
"grad_norm": 0.7070222496986389,
|
| 1870 |
+
"learning_rate": 4.22895357985838e-06,
|
| 1871 |
+
"loss": 0.8082,
|
| 1872 |
+
"step": 2610
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 4.63,
|
| 1876 |
+
"grad_norm": 0.2973116934299469,
|
| 1877 |
+
"learning_rate": 4.032258064516129e-06,
|
| 1878 |
+
"loss": 0.7245,
|
| 1879 |
+
"step": 2620
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 4.65,
|
| 1883 |
+
"grad_norm": 0.6432918310165405,
|
| 1884 |
+
"learning_rate": 3.835562549173879e-06,
|
| 1885 |
+
"loss": 0.5764,
|
| 1886 |
+
"step": 2630
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 4.67,
|
| 1890 |
+
"grad_norm": 0.8339099287986755,
|
| 1891 |
+
"learning_rate": 3.638867033831629e-06,
|
| 1892 |
+
"loss": 0.607,
|
| 1893 |
+
"step": 2640
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 4.69,
|
| 1897 |
+
"grad_norm": 0.5163792371749878,
|
| 1898 |
+
"learning_rate": 3.442171518489379e-06,
|
| 1899 |
+
"loss": 0.6602,
|
| 1900 |
+
"step": 2650
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 4.7,
|
| 1904 |
+
"grad_norm": 1.1758856773376465,
|
| 1905 |
+
"learning_rate": 3.2454760031471283e-06,
|
| 1906 |
+
"loss": 0.8349,
|
| 1907 |
+
"step": 2660
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 4.72,
|
| 1911 |
+
"grad_norm": 0.5800232887268066,
|
| 1912 |
+
"learning_rate": 3.0487804878048782e-06,
|
| 1913 |
+
"loss": 0.6944,
|
| 1914 |
+
"step": 2670
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 4.74,
|
| 1918 |
+
"grad_norm": 0.342181921005249,
|
| 1919 |
+
"learning_rate": 2.852084972462628e-06,
|
| 1920 |
+
"loss": 0.6855,
|
| 1921 |
+
"step": 2680
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 4.76,
|
| 1925 |
+
"grad_norm": 0.5758516788482666,
|
| 1926 |
+
"learning_rate": 2.655389457120378e-06,
|
| 1927 |
+
"loss": 0.5884,
|
| 1928 |
+
"step": 2690
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 4.77,
|
| 1932 |
+
"grad_norm": 0.42382463812828064,
|
| 1933 |
+
"learning_rate": 2.4586939417781275e-06,
|
| 1934 |
+
"loss": 0.7499,
|
| 1935 |
+
"step": 2700
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 4.79,
|
| 1939 |
+
"grad_norm": 0.5452685952186584,
|
| 1940 |
+
"learning_rate": 2.2619984264358773e-06,
|
| 1941 |
+
"loss": 0.6471,
|
| 1942 |
+
"step": 2710
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 4.81,
|
| 1946 |
+
"grad_norm": 1.8580130338668823,
|
| 1947 |
+
"learning_rate": 2.0653029110936272e-06,
|
| 1948 |
+
"loss": 0.6492,
|
| 1949 |
+
"step": 2720
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 4.83,
|
| 1953 |
+
"grad_norm": 1.0084635019302368,
|
| 1954 |
+
"learning_rate": 1.868607395751377e-06,
|
| 1955 |
+
"loss": 0.5958,
|
| 1956 |
+
"step": 2730
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 4.85,
|
| 1960 |
+
"grad_norm": 0.44602665305137634,
|
| 1961 |
+
"learning_rate": 1.6719118804091268e-06,
|
| 1962 |
+
"loss": 0.6962,
|
| 1963 |
+
"step": 2740
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 4.86,
|
| 1967 |
+
"grad_norm": 0.45221227407455444,
|
| 1968 |
+
"learning_rate": 1.4752163650668765e-06,
|
| 1969 |
+
"loss": 0.6997,
|
| 1970 |
+
"step": 2750
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 4.88,
|
| 1974 |
+
"grad_norm": 1.639831304550171,
|
| 1975 |
+
"learning_rate": 1.2785208497246264e-06,
|
| 1976 |
+
"loss": 0.6848,
|
| 1977 |
+
"step": 2760
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 4.9,
|
| 1981 |
+
"grad_norm": 0.6076949238777161,
|
| 1982 |
+
"learning_rate": 1.0818253343823763e-06,
|
| 1983 |
+
"loss": 0.6033,
|
| 1984 |
+
"step": 2770
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 4.92,
|
| 1988 |
+
"grad_norm": 0.46627792716026306,
|
| 1989 |
+
"learning_rate": 8.85129819040126e-07,
|
| 1990 |
+
"loss": 0.6447,
|
| 1991 |
+
"step": 2780
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 4.93,
|
| 1995 |
+
"grad_norm": 1.5518214702606201,
|
| 1996 |
+
"learning_rate": 6.884343036978757e-07,
|
| 1997 |
+
"loss": 0.7177,
|
| 1998 |
+
"step": 2790
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 4.95,
|
| 2002 |
+
"grad_norm": 0.4077186584472656,
|
| 2003 |
+
"learning_rate": 4.917387883556255e-07,
|
| 2004 |
+
"loss": 0.6174,
|
| 2005 |
+
"step": 2800
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 4.97,
|
| 2009 |
+
"grad_norm": 0.6033698916435242,
|
| 2010 |
+
"learning_rate": 2.9504327301337533e-07,
|
| 2011 |
+
"loss": 0.5902,
|
| 2012 |
+
"step": 2810
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 4.99,
|
| 2016 |
+
"grad_norm": 0.20630072057247162,
|
| 2017 |
+
"learning_rate": 9.834775767112511e-08,
|
| 2018 |
+
"loss": 0.6021,
|
| 2019 |
+
"step": 2820
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 5.0,
|
| 2023 |
+
"eval_accuracy": 0.9853463587921847,
|
| 2024 |
+
"eval_loss": 0.0637223869562149,
|
| 2025 |
+
"eval_runtime": 86.6358,
|
| 2026 |
+
"eval_samples_per_second": 25.994,
|
| 2027 |
+
"eval_steps_per_second": 3.255,
|
| 2028 |
+
"step": 2825
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 5.0,
|
| 2032 |
+
"step": 2825,
|
| 2033 |
+
"total_flos": 1.6016486531331686e+18,
|
| 2034 |
+
"train_loss": 0.9856665741236864,
|
| 2035 |
+
"train_runtime": 4486.694,
|
| 2036 |
+
"train_samples_per_second": 20.16,
|
| 2037 |
+
"train_steps_per_second": 0.63
|
| 2038 |
+
}
|
| 2039 |
+
],
|
| 2040 |
+
"logging_steps": 10,
|
| 2041 |
+
"max_steps": 2825,
|
| 2042 |
+
"num_input_tokens_seen": 0,
|
| 2043 |
+
"num_train_epochs": 5,
|
| 2044 |
+
"save_steps": 500,
|
| 2045 |
+
"total_flos": 1.6016486531331686e+18,
|
| 2046 |
+
"train_batch_size": 8,
|
| 2047 |
+
"trial_name": null,
|
| 2048 |
+
"trial_params": null
|
| 2049 |
+
}
|