Commit
•
3fea9c6
1
Parent(s):
2e8b09f
End of training
Browse files- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +2842 -0
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9987515605493134,
|
| 3 |
+
"total_flos": 2.847507751108608e+17,
|
| 4 |
+
"train_loss": 2.1486885741353037,
|
| 5 |
+
"train_runtime": 92894.5901,
|
| 6 |
+
"train_samples": 22417,
|
| 7 |
+
"train_samples_per_second": 0.241,
|
| 8 |
+
"train_steps_per_second": 0.004
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9987515605493134,
|
| 3 |
+
"total_flos": 2.847507751108608e+17,
|
| 4 |
+
"train_loss": 2.1486885741353037,
|
| 5 |
+
"train_runtime": 92894.5901,
|
| 6 |
+
"train_samples": 22417,
|
| 7 |
+
"train_samples_per_second": 0.241,
|
| 8 |
+
"train_steps_per_second": 0.004
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2842 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9987515605493134,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 400,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0024968789013732834,
|
| 13 |
+
"grad_norm": NaN,
|
| 14 |
+
"learning_rate": 5e-05,
|
| 15 |
+
"loss": 2.5487,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.004993757802746567,
|
| 20 |
+
"grad_norm": NaN,
|
| 21 |
+
"learning_rate": 5e-05,
|
| 22 |
+
"loss": 2.5726,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.00749063670411985,
|
| 27 |
+
"grad_norm": NaN,
|
| 28 |
+
"learning_rate": 5e-05,
|
| 29 |
+
"loss": 2.4761,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.009987515605493134,
|
| 34 |
+
"grad_norm": NaN,
|
| 35 |
+
"learning_rate": 5e-05,
|
| 36 |
+
"loss": 2.5582,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.012484394506866416,
|
| 41 |
+
"grad_norm": NaN,
|
| 42 |
+
"learning_rate": 5e-05,
|
| 43 |
+
"loss": 2.5577,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0149812734082397,
|
| 48 |
+
"grad_norm": NaN,
|
| 49 |
+
"learning_rate": 5e-05,
|
| 50 |
+
"loss": 2.5281,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.017478152309612985,
|
| 55 |
+
"grad_norm": NaN,
|
| 56 |
+
"learning_rate": 5e-05,
|
| 57 |
+
"loss": 2.5385,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.019975031210986267,
|
| 62 |
+
"grad_norm": NaN,
|
| 63 |
+
"learning_rate": 5e-05,
|
| 64 |
+
"loss": 2.4239,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.02247191011235955,
|
| 69 |
+
"grad_norm": NaN,
|
| 70 |
+
"learning_rate": 5e-05,
|
| 71 |
+
"loss": 2.558,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.024968789013732832,
|
| 76 |
+
"grad_norm": NaN,
|
| 77 |
+
"learning_rate": 5e-05,
|
| 78 |
+
"loss": 2.4404,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.02746566791510612,
|
| 83 |
+
"grad_norm": NaN,
|
| 84 |
+
"learning_rate": 5e-05,
|
| 85 |
+
"loss": 2.5149,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.0299625468164794,
|
| 90 |
+
"grad_norm": NaN,
|
| 91 |
+
"learning_rate": 5e-05,
|
| 92 |
+
"loss": 2.5407,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.03245942571785269,
|
| 97 |
+
"grad_norm": NaN,
|
| 98 |
+
"learning_rate": 5e-05,
|
| 99 |
+
"loss": 2.5754,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.03495630461922597,
|
| 104 |
+
"grad_norm": NaN,
|
| 105 |
+
"learning_rate": 5e-05,
|
| 106 |
+
"loss": 2.5222,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.03745318352059925,
|
| 111 |
+
"grad_norm": NaN,
|
| 112 |
+
"learning_rate": 5e-05,
|
| 113 |
+
"loss": 2.5261,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.039950062421972535,
|
| 118 |
+
"grad_norm": NaN,
|
| 119 |
+
"learning_rate": 5e-05,
|
| 120 |
+
"loss": 2.5199,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.04244694132334582,
|
| 125 |
+
"grad_norm": NaN,
|
| 126 |
+
"learning_rate": 5e-05,
|
| 127 |
+
"loss": 2.5327,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.0449438202247191,
|
| 132 |
+
"grad_norm": NaN,
|
| 133 |
+
"learning_rate": 5e-05,
|
| 134 |
+
"loss": 2.4925,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.04744069912609238,
|
| 139 |
+
"grad_norm": NaN,
|
| 140 |
+
"learning_rate": 5e-05,
|
| 141 |
+
"loss": 2.5452,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.049937578027465665,
|
| 146 |
+
"grad_norm": NaN,
|
| 147 |
+
"learning_rate": 5e-05,
|
| 148 |
+
"loss": 2.4986,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.052434456928838954,
|
| 153 |
+
"grad_norm": NaN,
|
| 154 |
+
"learning_rate": 5e-05,
|
| 155 |
+
"loss": 2.5148,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.05493133583021224,
|
| 160 |
+
"grad_norm": NaN,
|
| 161 |
+
"learning_rate": 5e-05,
|
| 162 |
+
"loss": 2.6131,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.05742821473158552,
|
| 167 |
+
"grad_norm": NaN,
|
| 168 |
+
"learning_rate": 5e-05,
|
| 169 |
+
"loss": 2.534,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.0599250936329588,
|
| 174 |
+
"grad_norm": 3.498404026031494,
|
| 175 |
+
"learning_rate": 4.9875000000000006e-05,
|
| 176 |
+
"loss": 2.5259,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.062421972534332085,
|
| 181 |
+
"grad_norm": 3.7281525135040283,
|
| 182 |
+
"learning_rate": 4.975e-05,
|
| 183 |
+
"loss": 2.5641,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.06491885143570537,
|
| 188 |
+
"grad_norm": 3.4878628253936768,
|
| 189 |
+
"learning_rate": 4.962500000000001e-05,
|
| 190 |
+
"loss": 2.5001,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.06741573033707865,
|
| 195 |
+
"grad_norm": 3.498518228530884,
|
| 196 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 197 |
+
"loss": 2.4923,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.06991260923845194,
|
| 202 |
+
"grad_norm": 3.5697684288024902,
|
| 203 |
+
"learning_rate": 4.937500000000001e-05,
|
| 204 |
+
"loss": 2.4743,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.07240948813982521,
|
| 209 |
+
"grad_norm": 3.58793044090271,
|
| 210 |
+
"learning_rate": 4.9250000000000004e-05,
|
| 211 |
+
"loss": 2.4403,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.0749063670411985,
|
| 216 |
+
"grad_norm": 3.395378828048706,
|
| 217 |
+
"learning_rate": 4.9125e-05,
|
| 218 |
+
"loss": 2.4839,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.07740324594257178,
|
| 223 |
+
"grad_norm": 3.4821157455444336,
|
| 224 |
+
"learning_rate": 4.9e-05,
|
| 225 |
+
"loss": 2.4865,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.07990012484394507,
|
| 230 |
+
"grad_norm": 3.5381717681884766,
|
| 231 |
+
"learning_rate": 4.8875e-05,
|
| 232 |
+
"loss": 2.3919,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.08239700374531835,
|
| 237 |
+
"grad_norm": 3.7879738807678223,
|
| 238 |
+
"learning_rate": 4.875e-05,
|
| 239 |
+
"loss": 2.5253,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.08489388264669163,
|
| 244 |
+
"grad_norm": 3.9536428451538086,
|
| 245 |
+
"learning_rate": 4.8625e-05,
|
| 246 |
+
"loss": 2.557,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.08739076154806492,
|
| 251 |
+
"grad_norm": 3.5281002521514893,
|
| 252 |
+
"learning_rate": 4.85e-05,
|
| 253 |
+
"loss": 2.419,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.0898876404494382,
|
| 258 |
+
"grad_norm": 3.8065526485443115,
|
| 259 |
+
"learning_rate": 4.8375000000000004e-05,
|
| 260 |
+
"loss": 2.4839,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.09238451935081149,
|
| 265 |
+
"grad_norm": 3.48492431640625,
|
| 266 |
+
"learning_rate": 4.825e-05,
|
| 267 |
+
"loss": 2.3443,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.09488139825218476,
|
| 272 |
+
"grad_norm": 3.5769503116607666,
|
| 273 |
+
"learning_rate": 4.8125000000000004e-05,
|
| 274 |
+
"loss": 2.5158,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.09737827715355805,
|
| 279 |
+
"grad_norm": 3.5777034759521484,
|
| 280 |
+
"learning_rate": 4.8e-05,
|
| 281 |
+
"loss": 2.403,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.09987515605493133,
|
| 286 |
+
"grad_norm": 3.471339225769043,
|
| 287 |
+
"learning_rate": 4.7875000000000005e-05,
|
| 288 |
+
"loss": 2.3998,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.10237203495630462,
|
| 293 |
+
"grad_norm": 3.6594278812408447,
|
| 294 |
+
"learning_rate": 4.775e-05,
|
| 295 |
+
"loss": 2.4084,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.10486891385767791,
|
| 300 |
+
"grad_norm": 3.5339417457580566,
|
| 301 |
+
"learning_rate": 4.7625000000000006e-05,
|
| 302 |
+
"loss": 2.4,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.10736579275905118,
|
| 307 |
+
"grad_norm": 3.4090983867645264,
|
| 308 |
+
"learning_rate": 4.75e-05,
|
| 309 |
+
"loss": 2.367,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.10986267166042447,
|
| 314 |
+
"grad_norm": 3.769968032836914,
|
| 315 |
+
"learning_rate": 4.7375e-05,
|
| 316 |
+
"loss": 2.3624,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.11235955056179775,
|
| 321 |
+
"grad_norm": 3.788815975189209,
|
| 322 |
+
"learning_rate": 4.7249999999999997e-05,
|
| 323 |
+
"loss": 2.4006,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.11485642946317104,
|
| 328 |
+
"grad_norm": 3.7179384231567383,
|
| 329 |
+
"learning_rate": 4.7125e-05,
|
| 330 |
+
"loss": 2.376,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.11735330836454431,
|
| 335 |
+
"grad_norm": 3.4289679527282715,
|
| 336 |
+
"learning_rate": 4.7e-05,
|
| 337 |
+
"loss": 2.3354,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.1198501872659176,
|
| 342 |
+
"grad_norm": 3.5412650108337402,
|
| 343 |
+
"learning_rate": 4.6875e-05,
|
| 344 |
+
"loss": 2.3317,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.12234706616729088,
|
| 349 |
+
"grad_norm": 3.5559499263763428,
|
| 350 |
+
"learning_rate": 4.6750000000000005e-05,
|
| 351 |
+
"loss": 2.3663,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.12484394506866417,
|
| 356 |
+
"grad_norm": 3.6865787506103516,
|
| 357 |
+
"learning_rate": 4.6625e-05,
|
| 358 |
+
"loss": 2.4062,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.12734082397003746,
|
| 363 |
+
"grad_norm": 3.417897939682007,
|
| 364 |
+
"learning_rate": 4.6500000000000005e-05,
|
| 365 |
+
"loss": 2.3314,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.12983770287141075,
|
| 370 |
+
"grad_norm": 3.5215134620666504,
|
| 371 |
+
"learning_rate": 4.6375e-05,
|
| 372 |
+
"loss": 2.367,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.132334581772784,
|
| 377 |
+
"grad_norm": 3.478131055831909,
|
| 378 |
+
"learning_rate": 4.6250000000000006e-05,
|
| 379 |
+
"loss": 2.285,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.1348314606741573,
|
| 384 |
+
"grad_norm": 3.8263564109802246,
|
| 385 |
+
"learning_rate": 4.6125e-05,
|
| 386 |
+
"loss": 2.402,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.1373283395755306,
|
| 391 |
+
"grad_norm": 3.4977402687072754,
|
| 392 |
+
"learning_rate": 4.600000000000001e-05,
|
| 393 |
+
"loss": 2.317,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.13982521847690388,
|
| 398 |
+
"grad_norm": 3.5797126293182373,
|
| 399 |
+
"learning_rate": 4.5875000000000004e-05,
|
| 400 |
+
"loss": 2.3097,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.14232209737827714,
|
| 405 |
+
"grad_norm": 3.6681716442108154,
|
| 406 |
+
"learning_rate": 4.575e-05,
|
| 407 |
+
"loss": 2.3766,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.14481897627965043,
|
| 412 |
+
"grad_norm": 3.6064581871032715,
|
| 413 |
+
"learning_rate": 4.5625e-05,
|
| 414 |
+
"loss": 2.2548,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.14731585518102372,
|
| 419 |
+
"grad_norm": 3.680893659591675,
|
| 420 |
+
"learning_rate": 4.55e-05,
|
| 421 |
+
"loss": 2.4249,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.149812734082397,
|
| 426 |
+
"grad_norm": 3.5588841438293457,
|
| 427 |
+
"learning_rate": 4.5375e-05,
|
| 428 |
+
"loss": 2.3109,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.1523096129837703,
|
| 433 |
+
"grad_norm": 3.4901130199432373,
|
| 434 |
+
"learning_rate": 4.525e-05,
|
| 435 |
+
"loss": 2.2421,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.15480649188514356,
|
| 440 |
+
"grad_norm": 3.5400235652923584,
|
| 441 |
+
"learning_rate": 4.5125e-05,
|
| 442 |
+
"loss": 2.3226,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.15730337078651685,
|
| 447 |
+
"grad_norm": 3.578420877456665,
|
| 448 |
+
"learning_rate": 4.5e-05,
|
| 449 |
+
"loss": 2.2904,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.15980024968789014,
|
| 454 |
+
"grad_norm": 3.583136558532715,
|
| 455 |
+
"learning_rate": 4.4875e-05,
|
| 456 |
+
"loss": 2.3255,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.16229712858926343,
|
| 461 |
+
"grad_norm": 3.35066819190979,
|
| 462 |
+
"learning_rate": 4.4750000000000004e-05,
|
| 463 |
+
"loss": 2.2183,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.1647940074906367,
|
| 468 |
+
"grad_norm": 3.4735934734344482,
|
| 469 |
+
"learning_rate": 4.4625e-05,
|
| 470 |
+
"loss": 2.2754,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.16729088639200998,
|
| 475 |
+
"grad_norm": 3.3307180404663086,
|
| 476 |
+
"learning_rate": 4.4500000000000004e-05,
|
| 477 |
+
"loss": 2.2964,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.16978776529338327,
|
| 482 |
+
"grad_norm": 3.516148567199707,
|
| 483 |
+
"learning_rate": 4.4375e-05,
|
| 484 |
+
"loss": 2.3107,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.17228464419475656,
|
| 489 |
+
"grad_norm": 3.4239909648895264,
|
| 490 |
+
"learning_rate": 4.4250000000000005e-05,
|
| 491 |
+
"loss": 2.2945,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.17478152309612985,
|
| 496 |
+
"grad_norm": 3.507624626159668,
|
| 497 |
+
"learning_rate": 4.4125e-05,
|
| 498 |
+
"loss": 2.2859,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.1772784019975031,
|
| 503 |
+
"grad_norm": 3.6821298599243164,
|
| 504 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 505 |
+
"loss": 2.3248,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.1797752808988764,
|
| 510 |
+
"grad_norm": 3.3753108978271484,
|
| 511 |
+
"learning_rate": 4.3875e-05,
|
| 512 |
+
"loss": 2.2001,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.1822721598002497,
|
| 517 |
+
"grad_norm": 3.4027931690216064,
|
| 518 |
+
"learning_rate": 4.375e-05,
|
| 519 |
+
"loss": 2.2668,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.18476903870162298,
|
| 524 |
+
"grad_norm": 3.524763584136963,
|
| 525 |
+
"learning_rate": 4.3625e-05,
|
| 526 |
+
"loss": 2.2686,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.18726591760299627,
|
| 531 |
+
"grad_norm": 3.5769248008728027,
|
| 532 |
+
"learning_rate": 4.35e-05,
|
| 533 |
+
"loss": 2.2757,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.18976279650436953,
|
| 538 |
+
"grad_norm": 3.5016684532165527,
|
| 539 |
+
"learning_rate": 4.3375000000000004e-05,
|
| 540 |
+
"loss": 2.2218,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.19225967540574282,
|
| 545 |
+
"grad_norm": 3.45440411567688,
|
| 546 |
+
"learning_rate": 4.325e-05,
|
| 547 |
+
"loss": 2.2405,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.1947565543071161,
|
| 552 |
+
"grad_norm": 3.5838842391967773,
|
| 553 |
+
"learning_rate": 4.3125000000000005e-05,
|
| 554 |
+
"loss": 2.2426,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.1972534332084894,
|
| 559 |
+
"grad_norm": 3.716230869293213,
|
| 560 |
+
"learning_rate": 4.3e-05,
|
| 561 |
+
"loss": 2.2819,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.19975031210986266,
|
| 566 |
+
"grad_norm": 3.258978843688965,
|
| 567 |
+
"learning_rate": 4.2875000000000005e-05,
|
| 568 |
+
"loss": 2.2049,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.20224719101123595,
|
| 573 |
+
"grad_norm": 3.575076103210449,
|
| 574 |
+
"learning_rate": 4.275e-05,
|
| 575 |
+
"loss": 2.2412,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.20474406991260924,
|
| 580 |
+
"grad_norm": 3.5371599197387695,
|
| 581 |
+
"learning_rate": 4.2625000000000006e-05,
|
| 582 |
+
"loss": 2.2206,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.20724094881398253,
|
| 587 |
+
"grad_norm": 3.6981263160705566,
|
| 588 |
+
"learning_rate": 4.25e-05,
|
| 589 |
+
"loss": 2.266,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.20973782771535582,
|
| 594 |
+
"grad_norm": 3.451237678527832,
|
| 595 |
+
"learning_rate": 4.237500000000001e-05,
|
| 596 |
+
"loss": 2.2564,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.21223470661672908,
|
| 601 |
+
"grad_norm": 3.4362707138061523,
|
| 602 |
+
"learning_rate": 4.2250000000000004e-05,
|
| 603 |
+
"loss": 2.1987,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.21473158551810237,
|
| 608 |
+
"grad_norm": 3.45649790763855,
|
| 609 |
+
"learning_rate": 4.2125e-05,
|
| 610 |
+
"loss": 2.2001,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.21722846441947566,
|
| 615 |
+
"grad_norm": 3.476231575012207,
|
| 616 |
+
"learning_rate": 4.2e-05,
|
| 617 |
+
"loss": 2.1378,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.21972534332084895,
|
| 622 |
+
"grad_norm": 3.5050230026245117,
|
| 623 |
+
"learning_rate": 4.1875e-05,
|
| 624 |
+
"loss": 2.1478,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.2222222222222222,
|
| 629 |
+
"grad_norm": 3.455080509185791,
|
| 630 |
+
"learning_rate": 4.175e-05,
|
| 631 |
+
"loss": 2.2395,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.2247191011235955,
|
| 636 |
+
"grad_norm": 3.500718832015991,
|
| 637 |
+
"learning_rate": 4.1625e-05,
|
| 638 |
+
"loss": 2.1794,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.2272159800249688,
|
| 643 |
+
"grad_norm": 3.53946852684021,
|
| 644 |
+
"learning_rate": 4.15e-05,
|
| 645 |
+
"loss": 2.2581,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.22971285892634208,
|
| 650 |
+
"grad_norm": 3.506479024887085,
|
| 651 |
+
"learning_rate": 4.1375e-05,
|
| 652 |
+
"loss": 2.2028,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.23220973782771537,
|
| 657 |
+
"grad_norm": 3.4707953929901123,
|
| 658 |
+
"learning_rate": 4.125e-05,
|
| 659 |
+
"loss": 2.1435,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.23470661672908863,
|
| 664 |
+
"grad_norm": 3.4014856815338135,
|
| 665 |
+
"learning_rate": 4.1125000000000004e-05,
|
| 666 |
+
"loss": 2.2146,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.23720349563046192,
|
| 671 |
+
"grad_norm": 3.42364764213562,
|
| 672 |
+
"learning_rate": 4.1e-05,
|
| 673 |
+
"loss": 2.1824,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.2397003745318352,
|
| 678 |
+
"grad_norm": 3.480703115463257,
|
| 679 |
+
"learning_rate": 4.0875000000000004e-05,
|
| 680 |
+
"loss": 2.2208,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.2421972534332085,
|
| 685 |
+
"grad_norm": 3.6021103858947754,
|
| 686 |
+
"learning_rate": 4.075e-05,
|
| 687 |
+
"loss": 2.1939,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.24469413233458176,
|
| 692 |
+
"grad_norm": 3.673717498779297,
|
| 693 |
+
"learning_rate": 4.0625000000000005e-05,
|
| 694 |
+
"loss": 2.234,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.24719101123595505,
|
| 699 |
+
"grad_norm": 3.511613368988037,
|
| 700 |
+
"learning_rate": 4.05e-05,
|
| 701 |
+
"loss": 2.2176,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.24968789013732834,
|
| 706 |
+
"grad_norm": 3.5998849868774414,
|
| 707 |
+
"learning_rate": 4.0375e-05,
|
| 708 |
+
"loss": 2.2259,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.25218476903870163,
|
| 713 |
+
"grad_norm": 3.4577770233154297,
|
| 714 |
+
"learning_rate": 4.025e-05,
|
| 715 |
+
"loss": 2.2238,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.2546816479400749,
|
| 720 |
+
"grad_norm": 3.612593173980713,
|
| 721 |
+
"learning_rate": 4.0125e-05,
|
| 722 |
+
"loss": 2.161,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.2571785268414482,
|
| 727 |
+
"grad_norm": 3.477313280105591,
|
| 728 |
+
"learning_rate": 4e-05,
|
| 729 |
+
"loss": 2.1839,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.2596754057428215,
|
| 734 |
+
"grad_norm": 3.553819417953491,
|
| 735 |
+
"learning_rate": 3.9875e-05,
|
| 736 |
+
"loss": 2.2249,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.26217228464419473,
|
| 741 |
+
"grad_norm": 3.412799596786499,
|
| 742 |
+
"learning_rate": 3.9750000000000004e-05,
|
| 743 |
+
"loss": 2.1834,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.264669163545568,
|
| 748 |
+
"grad_norm": 3.4956414699554443,
|
| 749 |
+
"learning_rate": 3.9625e-05,
|
| 750 |
+
"loss": 2.1456,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.2671660424469413,
|
| 755 |
+
"grad_norm": 3.6447088718414307,
|
| 756 |
+
"learning_rate": 3.9500000000000005e-05,
|
| 757 |
+
"loss": 2.166,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.2696629213483146,
|
| 762 |
+
"grad_norm": 3.6700491905212402,
|
| 763 |
+
"learning_rate": 3.9375e-05,
|
| 764 |
+
"loss": 2.2626,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.2721598002496879,
|
| 769 |
+
"grad_norm": 3.570895195007324,
|
| 770 |
+
"learning_rate": 3.9250000000000005e-05,
|
| 771 |
+
"loss": 2.1324,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.2746566791510612,
|
| 776 |
+
"grad_norm": 3.456331968307495,
|
| 777 |
+
"learning_rate": 3.9125e-05,
|
| 778 |
+
"loss": 2.1244,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.27715355805243447,
|
| 783 |
+
"grad_norm": 3.4046988487243652,
|
| 784 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 785 |
+
"loss": 2.1525,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.27965043695380776,
|
| 790 |
+
"grad_norm": 3.423374652862549,
|
| 791 |
+
"learning_rate": 3.8875e-05,
|
| 792 |
+
"loss": 2.1114,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.28214731585518105,
|
| 797 |
+
"grad_norm": 3.3426084518432617,
|
| 798 |
+
"learning_rate": 3.875e-05,
|
| 799 |
+
"loss": 2.1863,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.2846441947565543,
|
| 804 |
+
"grad_norm": 3.2898874282836914,
|
| 805 |
+
"learning_rate": 3.8625e-05,
|
| 806 |
+
"loss": 2.2121,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.28714107365792757,
|
| 811 |
+
"grad_norm": 3.395939826965332,
|
| 812 |
+
"learning_rate": 3.85e-05,
|
| 813 |
+
"loss": 2.1401,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.28963795255930086,
|
| 818 |
+
"grad_norm": 3.3747150897979736,
|
| 819 |
+
"learning_rate": 3.8375e-05,
|
| 820 |
+
"loss": 2.0959,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.29213483146067415,
|
| 825 |
+
"grad_norm": 3.5609467029571533,
|
| 826 |
+
"learning_rate": 3.825e-05,
|
| 827 |
+
"loss": 2.1268,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.29463171036204744,
|
| 832 |
+
"grad_norm": 3.6404829025268555,
|
| 833 |
+
"learning_rate": 3.8125e-05,
|
| 834 |
+
"loss": 2.184,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.29712858926342073,
|
| 839 |
+
"grad_norm": 3.50461745262146,
|
| 840 |
+
"learning_rate": 3.8e-05,
|
| 841 |
+
"loss": 2.168,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.299625468164794,
|
| 846 |
+
"grad_norm": 3.4648733139038086,
|
| 847 |
+
"learning_rate": 3.7875e-05,
|
| 848 |
+
"loss": 2.1526,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.3021223470661673,
|
| 853 |
+
"grad_norm": 3.696610450744629,
|
| 854 |
+
"learning_rate": 3.775e-05,
|
| 855 |
+
"loss": 2.204,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.3046192259675406,
|
| 860 |
+
"grad_norm": 3.4781978130340576,
|
| 861 |
+
"learning_rate": 3.7625e-05,
|
| 862 |
+
"loss": 2.1915,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.30711610486891383,
|
| 867 |
+
"grad_norm": 3.490053176879883,
|
| 868 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 869 |
+
"loss": 2.1437,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.3096129837702871,
|
| 874 |
+
"grad_norm": 3.452174186706543,
|
| 875 |
+
"learning_rate": 3.737500000000001e-05,
|
| 876 |
+
"loss": 2.1535,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.3121098626716604,
|
| 881 |
+
"grad_norm": 3.5188889503479004,
|
| 882 |
+
"learning_rate": 3.7250000000000004e-05,
|
| 883 |
+
"loss": 2.1696,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.3146067415730337,
|
| 888 |
+
"grad_norm": 3.452965021133423,
|
| 889 |
+
"learning_rate": 3.7125e-05,
|
| 890 |
+
"loss": 2.1077,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.317103620474407,
|
| 895 |
+
"grad_norm": 3.4781739711761475,
|
| 896 |
+
"learning_rate": 3.7e-05,
|
| 897 |
+
"loss": 2.1442,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.3196004993757803,
|
| 902 |
+
"grad_norm": 3.455982208251953,
|
| 903 |
+
"learning_rate": 3.6875e-05,
|
| 904 |
+
"loss": 2.2422,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.32209737827715357,
|
| 909 |
+
"grad_norm": 3.735812187194824,
|
| 910 |
+
"learning_rate": 3.675e-05,
|
| 911 |
+
"loss": 2.211,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.32459425717852686,
|
| 916 |
+
"grad_norm": 3.7122321128845215,
|
| 917 |
+
"learning_rate": 3.6625e-05,
|
| 918 |
+
"loss": 2.1534,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.32709113607990015,
|
| 923 |
+
"grad_norm": 3.407907724380493,
|
| 924 |
+
"learning_rate": 3.65e-05,
|
| 925 |
+
"loss": 2.135,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.3295880149812734,
|
| 930 |
+
"grad_norm": 3.610645055770874,
|
| 931 |
+
"learning_rate": 3.6375e-05,
|
| 932 |
+
"loss": 2.1437,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.33208489388264667,
|
| 937 |
+
"grad_norm": 3.316847085952759,
|
| 938 |
+
"learning_rate": 3.625e-05,
|
| 939 |
+
"loss": 2.1594,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.33458177278401996,
|
| 944 |
+
"grad_norm": 3.504835367202759,
|
| 945 |
+
"learning_rate": 3.6125000000000004e-05,
|
| 946 |
+
"loss": 2.1369,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.33707865168539325,
|
| 951 |
+
"grad_norm": 3.4886667728424072,
|
| 952 |
+
"learning_rate": 3.6e-05,
|
| 953 |
+
"loss": 2.1854,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.33957553058676654,
|
| 958 |
+
"grad_norm": 3.4116599559783936,
|
| 959 |
+
"learning_rate": 3.5875000000000005e-05,
|
| 960 |
+
"loss": 2.0741,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.34207240948813983,
|
| 965 |
+
"grad_norm": 3.4653735160827637,
|
| 966 |
+
"learning_rate": 3.575e-05,
|
| 967 |
+
"loss": 2.1779,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.3445692883895131,
|
| 972 |
+
"grad_norm": 3.496469497680664,
|
| 973 |
+
"learning_rate": 3.5625000000000005e-05,
|
| 974 |
+
"loss": 2.1337,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.3470661672908864,
|
| 979 |
+
"grad_norm": 3.4037115573883057,
|
| 980 |
+
"learning_rate": 3.55e-05,
|
| 981 |
+
"loss": 2.1346,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.3495630461922597,
|
| 986 |
+
"grad_norm": 3.6134746074676514,
|
| 987 |
+
"learning_rate": 3.5375e-05,
|
| 988 |
+
"loss": 2.212,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.352059925093633,
|
| 993 |
+
"grad_norm": 3.444700002670288,
|
| 994 |
+
"learning_rate": 3.525e-05,
|
| 995 |
+
"loss": 2.1676,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.3545568039950062,
|
| 1000 |
+
"grad_norm": 3.404010534286499,
|
| 1001 |
+
"learning_rate": 3.5125e-05,
|
| 1002 |
+
"loss": 2.0873,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.3570536828963795,
|
| 1007 |
+
"grad_norm": 3.3055148124694824,
|
| 1008 |
+
"learning_rate": 3.5e-05,
|
| 1009 |
+
"loss": 2.1762,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.3595505617977528,
|
| 1014 |
+
"grad_norm": 3.5183520317077637,
|
| 1015 |
+
"learning_rate": 3.4875e-05,
|
| 1016 |
+
"loss": 2.0908,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.3620474406991261,
|
| 1021 |
+
"grad_norm": 3.661907911300659,
|
| 1022 |
+
"learning_rate": 3.475e-05,
|
| 1023 |
+
"loss": 2.0899,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.3645443196004994,
|
| 1028 |
+
"grad_norm": 3.574948787689209,
|
| 1029 |
+
"learning_rate": 3.4625e-05,
|
| 1030 |
+
"loss": 2.1531,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.36704119850187267,
|
| 1035 |
+
"grad_norm": 3.4745309352874756,
|
| 1036 |
+
"learning_rate": 3.45e-05,
|
| 1037 |
+
"loss": 2.1238,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.36953807740324596,
|
| 1042 |
+
"grad_norm": 3.6377320289611816,
|
| 1043 |
+
"learning_rate": 3.4375e-05,
|
| 1044 |
+
"loss": 2.0556,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.37203495630461925,
|
| 1049 |
+
"grad_norm": 3.3080575466156006,
|
| 1050 |
+
"learning_rate": 3.4250000000000006e-05,
|
| 1051 |
+
"loss": 2.1218,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.37453183520599254,
|
| 1056 |
+
"grad_norm": 3.2884228229522705,
|
| 1057 |
+
"learning_rate": 3.4125e-05,
|
| 1058 |
+
"loss": 2.0843,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.37702871410736577,
|
| 1063 |
+
"grad_norm": 3.4235734939575195,
|
| 1064 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 1065 |
+
"loss": 2.1042,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.37952559300873906,
|
| 1070 |
+
"grad_norm": 3.284189462661743,
|
| 1071 |
+
"learning_rate": 3.3875000000000003e-05,
|
| 1072 |
+
"loss": 2.0937,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.38202247191011235,
|
| 1077 |
+
"grad_norm": 3.311265468597412,
|
| 1078 |
+
"learning_rate": 3.375000000000001e-05,
|
| 1079 |
+
"loss": 2.0621,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.38451935081148564,
|
| 1084 |
+
"grad_norm": 3.5135247707366943,
|
| 1085 |
+
"learning_rate": 3.3625000000000004e-05,
|
| 1086 |
+
"loss": 2.1317,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.38701622971285893,
|
| 1091 |
+
"grad_norm": 3.428800582885742,
|
| 1092 |
+
"learning_rate": 3.35e-05,
|
| 1093 |
+
"loss": 2.1447,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.3895131086142322,
|
| 1098 |
+
"grad_norm": 3.4077048301696777,
|
| 1099 |
+
"learning_rate": 3.3375e-05,
|
| 1100 |
+
"loss": 2.1757,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.3920099875156055,
|
| 1105 |
+
"grad_norm": 3.572735071182251,
|
| 1106 |
+
"learning_rate": 3.325e-05,
|
| 1107 |
+
"loss": 2.0373,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.3945068664169788,
|
| 1112 |
+
"grad_norm": 3.465317487716675,
|
| 1113 |
+
"learning_rate": 3.3125e-05,
|
| 1114 |
+
"loss": 2.0783,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.3970037453183521,
|
| 1119 |
+
"grad_norm": 3.5976178646087646,
|
| 1120 |
+
"learning_rate": 3.3e-05,
|
| 1121 |
+
"loss": 2.2142,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.3995006242197253,
|
| 1126 |
+
"grad_norm": 3.576270341873169,
|
| 1127 |
+
"learning_rate": 3.2875e-05,
|
| 1128 |
+
"loss": 2.134,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.4019975031210986,
|
| 1133 |
+
"grad_norm": 3.536389112472534,
|
| 1134 |
+
"learning_rate": 3.275e-05,
|
| 1135 |
+
"loss": 2.1068,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.4044943820224719,
|
| 1140 |
+
"grad_norm": 3.45173716545105,
|
| 1141 |
+
"learning_rate": 3.2625e-05,
|
| 1142 |
+
"loss": 2.0449,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.4069912609238452,
|
| 1147 |
+
"grad_norm": 3.498258352279663,
|
| 1148 |
+
"learning_rate": 3.2500000000000004e-05,
|
| 1149 |
+
"loss": 2.1712,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.4094881398252185,
|
| 1154 |
+
"grad_norm": 3.2850704193115234,
|
| 1155 |
+
"learning_rate": 3.2375e-05,
|
| 1156 |
+
"loss": 2.0633,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.41198501872659177,
|
| 1161 |
+
"grad_norm": 3.3497467041015625,
|
| 1162 |
+
"learning_rate": 3.2250000000000005e-05,
|
| 1163 |
+
"loss": 2.1995,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.41448189762796506,
|
| 1168 |
+
"grad_norm": 3.4341518878936768,
|
| 1169 |
+
"learning_rate": 3.2125e-05,
|
| 1170 |
+
"loss": 2.1297,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.41697877652933835,
|
| 1175 |
+
"grad_norm": 3.3831026554107666,
|
| 1176 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 1177 |
+
"loss": 2.198,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.41947565543071164,
|
| 1182 |
+
"grad_norm": 3.450352191925049,
|
| 1183 |
+
"learning_rate": 3.1875e-05,
|
| 1184 |
+
"loss": 2.092,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.42197253433208487,
|
| 1189 |
+
"grad_norm": 3.3830325603485107,
|
| 1190 |
+
"learning_rate": 3.175e-05,
|
| 1191 |
+
"loss": 2.1015,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.42446941323345816,
|
| 1196 |
+
"grad_norm": 3.548405885696411,
|
| 1197 |
+
"learning_rate": 3.1624999999999996e-05,
|
| 1198 |
+
"loss": 2.0894,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.42696629213483145,
|
| 1203 |
+
"grad_norm": 3.675614595413208,
|
| 1204 |
+
"learning_rate": 3.15e-05,
|
| 1205 |
+
"loss": 2.1759,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.42946317103620474,
|
| 1210 |
+
"grad_norm": 3.520315408706665,
|
| 1211 |
+
"learning_rate": 3.1375e-05,
|
| 1212 |
+
"loss": 2.1231,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.43196004993757803,
|
| 1217 |
+
"grad_norm": 3.604353666305542,
|
| 1218 |
+
"learning_rate": 3.125e-05,
|
| 1219 |
+
"loss": 2.048,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.4344569288389513,
|
| 1224 |
+
"grad_norm": 3.609004259109497,
|
| 1225 |
+
"learning_rate": 3.1125000000000004e-05,
|
| 1226 |
+
"loss": 2.1223,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.4369538077403246,
|
| 1231 |
+
"grad_norm": 3.4398860931396484,
|
| 1232 |
+
"learning_rate": 3.1e-05,
|
| 1233 |
+
"loss": 2.0657,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.4394506866416979,
|
| 1238 |
+
"grad_norm": 3.4411044120788574,
|
| 1239 |
+
"learning_rate": 3.0875000000000005e-05,
|
| 1240 |
+
"loss": 2.0593,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.4419475655430712,
|
| 1245 |
+
"grad_norm": 3.5043869018554688,
|
| 1246 |
+
"learning_rate": 3.075e-05,
|
| 1247 |
+
"loss": 2.157,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.4444444444444444,
|
| 1252 |
+
"grad_norm": 3.3451061248779297,
|
| 1253 |
+
"learning_rate": 3.0625000000000006e-05,
|
| 1254 |
+
"loss": 1.996,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.4469413233458177,
|
| 1259 |
+
"grad_norm": 3.6858675479888916,
|
| 1260 |
+
"learning_rate": 3.05e-05,
|
| 1261 |
+
"loss": 2.1263,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.449438202247191,
|
| 1266 |
+
"grad_norm": 3.5967984199523926,
|
| 1267 |
+
"learning_rate": 3.0375000000000003e-05,
|
| 1268 |
+
"loss": 2.0165,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.4519350811485643,
|
| 1273 |
+
"grad_norm": 3.5284011363983154,
|
| 1274 |
+
"learning_rate": 3.025e-05,
|
| 1275 |
+
"loss": 2.1286,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.4544319600499376,
|
| 1280 |
+
"grad_norm": 3.548715353012085,
|
| 1281 |
+
"learning_rate": 3.0125000000000004e-05,
|
| 1282 |
+
"loss": 2.0539,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.45692883895131087,
|
| 1287 |
+
"grad_norm": 3.52622127532959,
|
| 1288 |
+
"learning_rate": 3e-05,
|
| 1289 |
+
"loss": 2.1254,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.45942571785268416,
|
| 1294 |
+
"grad_norm": 3.3532912731170654,
|
| 1295 |
+
"learning_rate": 2.9875000000000004e-05,
|
| 1296 |
+
"loss": 2.1257,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.46192259675405745,
|
| 1301 |
+
"grad_norm": 3.4587035179138184,
|
| 1302 |
+
"learning_rate": 2.975e-05,
|
| 1303 |
+
"loss": 2.1171,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.46441947565543074,
|
| 1308 |
+
"grad_norm": 3.5046117305755615,
|
| 1309 |
+
"learning_rate": 2.9625000000000002e-05,
|
| 1310 |
+
"loss": 2.1057,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.46691635455680397,
|
| 1315 |
+
"grad_norm": 3.33771014213562,
|
| 1316 |
+
"learning_rate": 2.95e-05,
|
| 1317 |
+
"loss": 2.0668,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.46941323345817726,
|
| 1322 |
+
"grad_norm": 3.402855157852173,
|
| 1323 |
+
"learning_rate": 2.9375000000000003e-05,
|
| 1324 |
+
"loss": 2.087,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.47191011235955055,
|
| 1329 |
+
"grad_norm": 3.417536735534668,
|
| 1330 |
+
"learning_rate": 2.925e-05,
|
| 1331 |
+
"loss": 1.9994,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.47440699126092384,
|
| 1336 |
+
"grad_norm": 3.627957582473755,
|
| 1337 |
+
"learning_rate": 2.9125000000000003e-05,
|
| 1338 |
+
"loss": 2.1349,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.4769038701622971,
|
| 1343 |
+
"grad_norm": 3.5531859397888184,
|
| 1344 |
+
"learning_rate": 2.9e-05,
|
| 1345 |
+
"loss": 2.1836,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.4794007490636704,
|
| 1350 |
+
"grad_norm": 3.5431745052337646,
|
| 1351 |
+
"learning_rate": 2.8875e-05,
|
| 1352 |
+
"loss": 2.0185,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.4818976279650437,
|
| 1357 |
+
"grad_norm": 3.592146158218384,
|
| 1358 |
+
"learning_rate": 2.8749999999999997e-05,
|
| 1359 |
+
"loss": 2.0973,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.484394506866417,
|
| 1364 |
+
"grad_norm": 3.4691829681396484,
|
| 1365 |
+
"learning_rate": 2.8625e-05,
|
| 1366 |
+
"loss": 2.1199,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.4868913857677903,
|
| 1371 |
+
"grad_norm": 3.599349021911621,
|
| 1372 |
+
"learning_rate": 2.8499999999999998e-05,
|
| 1373 |
+
"loss": 2.0849,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.4893882646691635,
|
| 1378 |
+
"grad_norm": 3.5513803958892822,
|
| 1379 |
+
"learning_rate": 2.8375000000000002e-05,
|
| 1380 |
+
"loss": 2.0859,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.4918851435705368,
|
| 1385 |
+
"grad_norm": 3.4182136058807373,
|
| 1386 |
+
"learning_rate": 2.825e-05,
|
| 1387 |
+
"loss": 2.0096,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.4943820224719101,
|
| 1392 |
+
"grad_norm": 3.347532272338867,
|
| 1393 |
+
"learning_rate": 2.8125000000000003e-05,
|
| 1394 |
+
"loss": 2.1039,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.4968789013732834,
|
| 1399 |
+
"grad_norm": 3.5202083587646484,
|
| 1400 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 1401 |
+
"loss": 2.0436,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.4993757802746567,
|
| 1406 |
+
"grad_norm": 3.5688531398773193,
|
| 1407 |
+
"learning_rate": 2.7875e-05,
|
| 1408 |
+
"loss": 2.0736,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.50187265917603,
|
| 1413 |
+
"grad_norm": 3.531306743621826,
|
| 1414 |
+
"learning_rate": 2.7750000000000004e-05,
|
| 1415 |
+
"loss": 2.0244,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.5043695380774033,
|
| 1420 |
+
"grad_norm": 3.595841884613037,
|
| 1421 |
+
"learning_rate": 2.7625e-05,
|
| 1422 |
+
"loss": 2.0808,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.5068664169787765,
|
| 1427 |
+
"grad_norm": 3.610381603240967,
|
| 1428 |
+
"learning_rate": 2.7500000000000004e-05,
|
| 1429 |
+
"loss": 2.0863,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.5093632958801498,
|
| 1434 |
+
"grad_norm": 3.2456705570220947,
|
| 1435 |
+
"learning_rate": 2.7375e-05,
|
| 1436 |
+
"loss": 2.1102,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.5118601747815231,
|
| 1441 |
+
"grad_norm": 3.5384883880615234,
|
| 1442 |
+
"learning_rate": 2.725e-05,
|
| 1443 |
+
"loss": 2.0491,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.5143570536828964,
|
| 1448 |
+
"grad_norm": 3.4325191974639893,
|
| 1449 |
+
"learning_rate": 2.7125000000000002e-05,
|
| 1450 |
+
"loss": 2.0562,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.5168539325842697,
|
| 1455 |
+
"grad_norm": 3.428877115249634,
|
| 1456 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 1457 |
+
"loss": 2.0329,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.519350811485643,
|
| 1462 |
+
"grad_norm": 3.451172351837158,
|
| 1463 |
+
"learning_rate": 2.6875e-05,
|
| 1464 |
+
"loss": 2.1357,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.5218476903870163,
|
| 1469 |
+
"grad_norm": 3.5145444869995117,
|
| 1470 |
+
"learning_rate": 2.6750000000000003e-05,
|
| 1471 |
+
"loss": 2.133,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.5243445692883895,
|
| 1476 |
+
"grad_norm": 3.452249050140381,
|
| 1477 |
+
"learning_rate": 2.6625e-05,
|
| 1478 |
+
"loss": 2.033,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.5268414481897628,
|
| 1483 |
+
"grad_norm": 3.493157148361206,
|
| 1484 |
+
"learning_rate": 2.6500000000000004e-05,
|
| 1485 |
+
"loss": 2.0387,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.529338327091136,
|
| 1490 |
+
"grad_norm": 3.4883360862731934,
|
| 1491 |
+
"learning_rate": 2.6375e-05,
|
| 1492 |
+
"loss": 2.0182,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.5318352059925093,
|
| 1497 |
+
"grad_norm": 3.377523183822632,
|
| 1498 |
+
"learning_rate": 2.625e-05,
|
| 1499 |
+
"loss": 2.0963,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.5343320848938826,
|
| 1504 |
+
"grad_norm": 3.4430289268493652,
|
| 1505 |
+
"learning_rate": 2.6124999999999998e-05,
|
| 1506 |
+
"loss": 2.0959,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.5368289637952559,
|
| 1511 |
+
"grad_norm": 3.380902051925659,
|
| 1512 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 1513 |
+
"loss": 2.0883,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.5393258426966292,
|
| 1518 |
+
"grad_norm": 3.437469005584717,
|
| 1519 |
+
"learning_rate": 2.5875e-05,
|
| 1520 |
+
"loss": 2.0548,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.5418227215980025,
|
| 1525 |
+
"grad_norm": 3.628032684326172,
|
| 1526 |
+
"learning_rate": 2.5750000000000002e-05,
|
| 1527 |
+
"loss": 2.0784,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.5443196004993758,
|
| 1532 |
+
"grad_norm": 3.2870864868164062,
|
| 1533 |
+
"learning_rate": 2.5625e-05,
|
| 1534 |
+
"loss": 2.0463,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.5468164794007491,
|
| 1539 |
+
"grad_norm": 3.5962319374084473,
|
| 1540 |
+
"learning_rate": 2.5500000000000003e-05,
|
| 1541 |
+
"loss": 2.0625,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.5493133583021224,
|
| 1546 |
+
"grad_norm": 3.3193917274475098,
|
| 1547 |
+
"learning_rate": 2.5375e-05,
|
| 1548 |
+
"loss": 2.0249,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.5518102372034956,
|
| 1553 |
+
"grad_norm": 3.4375598430633545,
|
| 1554 |
+
"learning_rate": 2.525e-05,
|
| 1555 |
+
"loss": 2.0802,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.5543071161048689,
|
| 1560 |
+
"grad_norm": 3.500824451446533,
|
| 1561 |
+
"learning_rate": 2.5124999999999997e-05,
|
| 1562 |
+
"loss": 2.0309,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.5568039950062422,
|
| 1567 |
+
"grad_norm": 3.5241281986236572,
|
| 1568 |
+
"learning_rate": 2.5e-05,
|
| 1569 |
+
"loss": 2.1013,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.5593008739076155,
|
| 1574 |
+
"grad_norm": 3.4705541133880615,
|
| 1575 |
+
"learning_rate": 2.4875e-05,
|
| 1576 |
+
"loss": 2.0883,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.5617977528089888,
|
| 1581 |
+
"grad_norm": 3.4071896076202393,
|
| 1582 |
+
"learning_rate": 2.4750000000000002e-05,
|
| 1583 |
+
"loss": 2.0965,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.5642946317103621,
|
| 1588 |
+
"grad_norm": 3.315619707107544,
|
| 1589 |
+
"learning_rate": 2.4625000000000002e-05,
|
| 1590 |
+
"loss": 2.0697,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.5667915106117354,
|
| 1595 |
+
"grad_norm": 3.280471086502075,
|
| 1596 |
+
"learning_rate": 2.45e-05,
|
| 1597 |
+
"loss": 2.1179,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.5692883895131086,
|
| 1602 |
+
"grad_norm": 3.7654521465301514,
|
| 1603 |
+
"learning_rate": 2.4375e-05,
|
| 1604 |
+
"loss": 2.0855,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.5717852684144819,
|
| 1609 |
+
"grad_norm": 3.7220211029052734,
|
| 1610 |
+
"learning_rate": 2.425e-05,
|
| 1611 |
+
"loss": 2.0319,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.5742821473158551,
|
| 1616 |
+
"grad_norm": 3.419952154159546,
|
| 1617 |
+
"learning_rate": 2.4125e-05,
|
| 1618 |
+
"loss": 2.0362,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.5767790262172284,
|
| 1623 |
+
"grad_norm": 3.46474552154541,
|
| 1624 |
+
"learning_rate": 2.4e-05,
|
| 1625 |
+
"loss": 2.0325,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.5792759051186017,
|
| 1630 |
+
"grad_norm": 3.272773027420044,
|
| 1631 |
+
"learning_rate": 2.3875e-05,
|
| 1632 |
+
"loss": 2.0735,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.581772784019975,
|
| 1637 |
+
"grad_norm": 3.659705877304077,
|
| 1638 |
+
"learning_rate": 2.375e-05,
|
| 1639 |
+
"loss": 2.1283,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.5842696629213483,
|
| 1644 |
+
"grad_norm": 3.303420305252075,
|
| 1645 |
+
"learning_rate": 2.3624999999999998e-05,
|
| 1646 |
+
"loss": 2.1808,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.5867665418227216,
|
| 1651 |
+
"grad_norm": 3.622915744781494,
|
| 1652 |
+
"learning_rate": 2.35e-05,
|
| 1653 |
+
"loss": 2.0583,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.5892634207240949,
|
| 1658 |
+
"grad_norm": 3.3635072708129883,
|
| 1659 |
+
"learning_rate": 2.3375000000000002e-05,
|
| 1660 |
+
"loss": 2.0519,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.5917602996254682,
|
| 1665 |
+
"grad_norm": 3.5549280643463135,
|
| 1666 |
+
"learning_rate": 2.3250000000000003e-05,
|
| 1667 |
+
"loss": 2.0682,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.5942571785268415,
|
| 1672 |
+
"grad_norm": 3.578655958175659,
|
| 1673 |
+
"learning_rate": 2.3125000000000003e-05,
|
| 1674 |
+
"loss": 2.077,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.5967540574282147,
|
| 1679 |
+
"grad_norm": 3.611335277557373,
|
| 1680 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 1681 |
+
"loss": 1.9911,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.599250936329588,
|
| 1686 |
+
"grad_norm": 3.3821637630462646,
|
| 1687 |
+
"learning_rate": 2.2875e-05,
|
| 1688 |
+
"loss": 2.0082,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.6017478152309613,
|
| 1693 |
+
"grad_norm": 3.4852335453033447,
|
| 1694 |
+
"learning_rate": 2.275e-05,
|
| 1695 |
+
"loss": 2.0278,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.6042446941323346,
|
| 1700 |
+
"grad_norm": 3.525944948196411,
|
| 1701 |
+
"learning_rate": 2.2625e-05,
|
| 1702 |
+
"loss": 2.0914,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.6067415730337079,
|
| 1707 |
+
"grad_norm": 3.3227696418762207,
|
| 1708 |
+
"learning_rate": 2.25e-05,
|
| 1709 |
+
"loss": 2.107,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.6092384519350812,
|
| 1714 |
+
"grad_norm": 3.669677257537842,
|
| 1715 |
+
"learning_rate": 2.2375000000000002e-05,
|
| 1716 |
+
"loss": 1.9829,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.6117353308364545,
|
| 1721 |
+
"grad_norm": 3.625420570373535,
|
| 1722 |
+
"learning_rate": 2.2250000000000002e-05,
|
| 1723 |
+
"loss": 2.0123,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.6142322097378277,
|
| 1728 |
+
"grad_norm": 3.4239187240600586,
|
| 1729 |
+
"learning_rate": 2.2125000000000002e-05,
|
| 1730 |
+
"loss": 2.0562,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.616729088639201,
|
| 1735 |
+
"grad_norm": 3.5444185733795166,
|
| 1736 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 1737 |
+
"loss": 2.0755,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.6192259675405742,
|
| 1742 |
+
"grad_norm": 3.500282049179077,
|
| 1743 |
+
"learning_rate": 2.1875e-05,
|
| 1744 |
+
"loss": 2.0275,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.6217228464419475,
|
| 1749 |
+
"grad_norm": 3.4052162170410156,
|
| 1750 |
+
"learning_rate": 2.175e-05,
|
| 1751 |
+
"loss": 2.0654,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.6242197253433208,
|
| 1756 |
+
"grad_norm": 3.5831384658813477,
|
| 1757 |
+
"learning_rate": 2.1625e-05,
|
| 1758 |
+
"loss": 1.994,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.6267166042446941,
|
| 1763 |
+
"grad_norm": 3.322357654571533,
|
| 1764 |
+
"learning_rate": 2.15e-05,
|
| 1765 |
+
"loss": 2.0476,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.6292134831460674,
|
| 1770 |
+
"grad_norm": 3.543062448501587,
|
| 1771 |
+
"learning_rate": 2.1375e-05,
|
| 1772 |
+
"loss": 2.0585,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.6317103620474407,
|
| 1777 |
+
"grad_norm": 3.521042823791504,
|
| 1778 |
+
"learning_rate": 2.125e-05,
|
| 1779 |
+
"loss": 2.0239,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.634207240948814,
|
| 1784 |
+
"grad_norm": 3.395559549331665,
|
| 1785 |
+
"learning_rate": 2.1125000000000002e-05,
|
| 1786 |
+
"loss": 2.0957,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.6367041198501873,
|
| 1791 |
+
"grad_norm": 3.567340135574341,
|
| 1792 |
+
"learning_rate": 2.1e-05,
|
| 1793 |
+
"loss": 2.0649,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.6392009987515606,
|
| 1798 |
+
"grad_norm": 3.397399663925171,
|
| 1799 |
+
"learning_rate": 2.0875e-05,
|
| 1800 |
+
"loss": 2.1739,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.6416978776529338,
|
| 1805 |
+
"grad_norm": 3.4432034492492676,
|
| 1806 |
+
"learning_rate": 2.075e-05,
|
| 1807 |
+
"loss": 2.0686,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.6441947565543071,
|
| 1812 |
+
"grad_norm": 3.4990758895874023,
|
| 1813 |
+
"learning_rate": 2.0625e-05,
|
| 1814 |
+
"loss": 2.1194,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.6466916354556804,
|
| 1819 |
+
"grad_norm": 3.497478723526001,
|
| 1820 |
+
"learning_rate": 2.05e-05,
|
| 1821 |
+
"loss": 2.0429,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.6491885143570537,
|
| 1826 |
+
"grad_norm": 3.4457273483276367,
|
| 1827 |
+
"learning_rate": 2.0375e-05,
|
| 1828 |
+
"loss": 2.1038,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.651685393258427,
|
| 1833 |
+
"grad_norm": 3.3803672790527344,
|
| 1834 |
+
"learning_rate": 2.025e-05,
|
| 1835 |
+
"loss": 2.1119,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.6541822721598003,
|
| 1840 |
+
"grad_norm": 3.474200963973999,
|
| 1841 |
+
"learning_rate": 2.0125e-05,
|
| 1842 |
+
"loss": 2.0346,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.6566791510611736,
|
| 1847 |
+
"grad_norm": 3.4010913372039795,
|
| 1848 |
+
"learning_rate": 2e-05,
|
| 1849 |
+
"loss": 2.0361,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.6591760299625468,
|
| 1854 |
+
"grad_norm": 3.391678810119629,
|
| 1855 |
+
"learning_rate": 1.9875000000000002e-05,
|
| 1856 |
+
"loss": 2.0216,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.66167290886392,
|
| 1861 |
+
"grad_norm": 3.484393835067749,
|
| 1862 |
+
"learning_rate": 1.9750000000000002e-05,
|
| 1863 |
+
"loss": 2.1634,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.6641697877652933,
|
| 1868 |
+
"grad_norm": 3.491408586502075,
|
| 1869 |
+
"learning_rate": 1.9625000000000003e-05,
|
| 1870 |
+
"loss": 2.0471,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.6666666666666666,
|
| 1875 |
+
"grad_norm": 3.562328815460205,
|
| 1876 |
+
"learning_rate": 1.9500000000000003e-05,
|
| 1877 |
+
"loss": 2.0928,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.6691635455680399,
|
| 1882 |
+
"grad_norm": 3.6033616065979004,
|
| 1883 |
+
"learning_rate": 1.9375e-05,
|
| 1884 |
+
"loss": 2.0641,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.6716604244694132,
|
| 1889 |
+
"grad_norm": 3.441002130508423,
|
| 1890 |
+
"learning_rate": 1.925e-05,
|
| 1891 |
+
"loss": 2.0484,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.6741573033707865,
|
| 1896 |
+
"grad_norm": 3.4448623657226562,
|
| 1897 |
+
"learning_rate": 1.9125e-05,
|
| 1898 |
+
"loss": 2.0558,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.6766541822721598,
|
| 1903 |
+
"grad_norm": 3.481809377670288,
|
| 1904 |
+
"learning_rate": 1.9e-05,
|
| 1905 |
+
"loss": 2.0449,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.6791510611735331,
|
| 1910 |
+
"grad_norm": 3.2882351875305176,
|
| 1911 |
+
"learning_rate": 1.8875e-05,
|
| 1912 |
+
"loss": 1.9441,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.6816479400749064,
|
| 1917 |
+
"grad_norm": 3.5338022708892822,
|
| 1918 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 1919 |
+
"loss": 2.1015,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.6841448189762797,
|
| 1924 |
+
"grad_norm": 3.349963665008545,
|
| 1925 |
+
"learning_rate": 1.8625000000000002e-05,
|
| 1926 |
+
"loss": 2.0868,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.686641697877653,
|
| 1931 |
+
"grad_norm": 3.562603712081909,
|
| 1932 |
+
"learning_rate": 1.85e-05,
|
| 1933 |
+
"loss": 2.0308,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.6891385767790262,
|
| 1938 |
+
"grad_norm": 3.484811782836914,
|
| 1939 |
+
"learning_rate": 1.8375e-05,
|
| 1940 |
+
"loss": 2.0972,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.6916354556803995,
|
| 1945 |
+
"grad_norm": 3.4616386890411377,
|
| 1946 |
+
"learning_rate": 1.825e-05,
|
| 1947 |
+
"loss": 2.0084,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.6941323345817728,
|
| 1952 |
+
"grad_norm": 3.5441646575927734,
|
| 1953 |
+
"learning_rate": 1.8125e-05,
|
| 1954 |
+
"loss": 2.0395,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.6966292134831461,
|
| 1959 |
+
"grad_norm": 3.5133986473083496,
|
| 1960 |
+
"learning_rate": 1.8e-05,
|
| 1961 |
+
"loss": 2.0973,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.6991260923845194,
|
| 1966 |
+
"grad_norm": 3.6960537433624268,
|
| 1967 |
+
"learning_rate": 1.7875e-05,
|
| 1968 |
+
"loss": 1.9632,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.7016229712858927,
|
| 1973 |
+
"grad_norm": 3.499337911605835,
|
| 1974 |
+
"learning_rate": 1.775e-05,
|
| 1975 |
+
"loss": 2.0925,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.704119850187266,
|
| 1980 |
+
"grad_norm": 3.451720952987671,
|
| 1981 |
+
"learning_rate": 1.7625e-05,
|
| 1982 |
+
"loss": 2.0816,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.7066167290886392,
|
| 1987 |
+
"grad_norm": 3.540565252304077,
|
| 1988 |
+
"learning_rate": 1.75e-05,
|
| 1989 |
+
"loss": 2.0853,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.7091136079900124,
|
| 1994 |
+
"grad_norm": 3.5862390995025635,
|
| 1995 |
+
"learning_rate": 1.7375e-05,
|
| 1996 |
+
"loss": 1.9933,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.7116104868913857,
|
| 2001 |
+
"grad_norm": 3.436612844467163,
|
| 2002 |
+
"learning_rate": 1.725e-05,
|
| 2003 |
+
"loss": 2.0519,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.714107365792759,
|
| 2008 |
+
"grad_norm": 3.5170860290527344,
|
| 2009 |
+
"learning_rate": 1.7125000000000003e-05,
|
| 2010 |
+
"loss": 2.0087,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.7166042446941323,
|
| 2015 |
+
"grad_norm": 3.613799571990967,
|
| 2016 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 2017 |
+
"loss": 2.0785,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.7191011235955056,
|
| 2022 |
+
"grad_norm": 3.4286880493164062,
|
| 2023 |
+
"learning_rate": 1.6875000000000004e-05,
|
| 2024 |
+
"loss": 2.0805,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.7215980024968789,
|
| 2029 |
+
"grad_norm": 3.4460251331329346,
|
| 2030 |
+
"learning_rate": 1.675e-05,
|
| 2031 |
+
"loss": 2.0529,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.7240948813982522,
|
| 2036 |
+
"grad_norm": 3.5601372718811035,
|
| 2037 |
+
"learning_rate": 1.6625e-05,
|
| 2038 |
+
"loss": 2.0795,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.7265917602996255,
|
| 2043 |
+
"grad_norm": 3.3969712257385254,
|
| 2044 |
+
"learning_rate": 1.65e-05,
|
| 2045 |
+
"loss": 2.0261,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.7290886392009988,
|
| 2050 |
+
"grad_norm": 3.285691738128662,
|
| 2051 |
+
"learning_rate": 1.6375e-05,
|
| 2052 |
+
"loss": 2.0902,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.731585518102372,
|
| 2057 |
+
"grad_norm": 3.482919454574585,
|
| 2058 |
+
"learning_rate": 1.6250000000000002e-05,
|
| 2059 |
+
"loss": 2.0239,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.7340823970037453,
|
| 2064 |
+
"grad_norm": 3.489551544189453,
|
| 2065 |
+
"learning_rate": 1.6125000000000002e-05,
|
| 2066 |
+
"loss": 2.0235,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.7365792759051186,
|
| 2071 |
+
"grad_norm": 3.461838960647583,
|
| 2072 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 2073 |
+
"loss": 2.0562,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.7390761548064919,
|
| 2078 |
+
"grad_norm": 3.3764636516571045,
|
| 2079 |
+
"learning_rate": 1.5875e-05,
|
| 2080 |
+
"loss": 2.0649,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.7415730337078652,
|
| 2085 |
+
"grad_norm": 3.4460861682891846,
|
| 2086 |
+
"learning_rate": 1.575e-05,
|
| 2087 |
+
"loss": 2.0674,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.7440699126092385,
|
| 2092 |
+
"grad_norm": 3.5665292739868164,
|
| 2093 |
+
"learning_rate": 1.5625e-05,
|
| 2094 |
+
"loss": 2.0219,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.7465667915106118,
|
| 2099 |
+
"grad_norm": 3.505133867263794,
|
| 2100 |
+
"learning_rate": 1.55e-05,
|
| 2101 |
+
"loss": 2.0278,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.7490636704119851,
|
| 2106 |
+
"grad_norm": 3.615962266921997,
|
| 2107 |
+
"learning_rate": 1.5375e-05,
|
| 2108 |
+
"loss": 2.0006,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.7515605493133583,
|
| 2113 |
+
"grad_norm": 3.462989330291748,
|
| 2114 |
+
"learning_rate": 1.525e-05,
|
| 2115 |
+
"loss": 2.0423,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.7540574282147315,
|
| 2120 |
+
"grad_norm": 3.351733446121216,
|
| 2121 |
+
"learning_rate": 1.5125e-05,
|
| 2122 |
+
"loss": 2.0353,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.7565543071161048,
|
| 2127 |
+
"grad_norm": 3.3703553676605225,
|
| 2128 |
+
"learning_rate": 1.5e-05,
|
| 2129 |
+
"loss": 2.0352,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.7590511860174781,
|
| 2134 |
+
"grad_norm": 3.578253984451294,
|
| 2135 |
+
"learning_rate": 1.4875e-05,
|
| 2136 |
+
"loss": 2.027,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.7615480649188514,
|
| 2141 |
+
"grad_norm": 3.304739236831665,
|
| 2142 |
+
"learning_rate": 1.475e-05,
|
| 2143 |
+
"loss": 2.0538,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.7640449438202247,
|
| 2148 |
+
"grad_norm": 3.5209975242614746,
|
| 2149 |
+
"learning_rate": 1.4625e-05,
|
| 2150 |
+
"loss": 1.9996,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.766541822721598,
|
| 2155 |
+
"grad_norm": 3.5905096530914307,
|
| 2156 |
+
"learning_rate": 1.45e-05,
|
| 2157 |
+
"loss": 2.0141,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.7690387016229713,
|
| 2162 |
+
"grad_norm": 3.383382558822632,
|
| 2163 |
+
"learning_rate": 1.4374999999999999e-05,
|
| 2164 |
+
"loss": 1.9648,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.7715355805243446,
|
| 2169 |
+
"grad_norm": 3.4596011638641357,
|
| 2170 |
+
"learning_rate": 1.4249999999999999e-05,
|
| 2171 |
+
"loss": 2.0336,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.7740324594257179,
|
| 2176 |
+
"grad_norm": 3.4215636253356934,
|
| 2177 |
+
"learning_rate": 1.4125e-05,
|
| 2178 |
+
"loss": 2.0386,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.7765293383270911,
|
| 2183 |
+
"grad_norm": 3.341522693634033,
|
| 2184 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 2185 |
+
"loss": 2.1078,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.7790262172284644,
|
| 2190 |
+
"grad_norm": 3.366393566131592,
|
| 2191 |
+
"learning_rate": 1.3875000000000002e-05,
|
| 2192 |
+
"loss": 2.079,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.7815230961298377,
|
| 2197 |
+
"grad_norm": 3.497201919555664,
|
| 2198 |
+
"learning_rate": 1.3750000000000002e-05,
|
| 2199 |
+
"loss": 2.0535,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.784019975031211,
|
| 2204 |
+
"grad_norm": 3.563197135925293,
|
| 2205 |
+
"learning_rate": 1.3625e-05,
|
| 2206 |
+
"loss": 2.0283,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.7865168539325843,
|
| 2211 |
+
"grad_norm": 3.4403908252716064,
|
| 2212 |
+
"learning_rate": 1.3500000000000001e-05,
|
| 2213 |
+
"loss": 2.0832,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.7890137328339576,
|
| 2218 |
+
"grad_norm": 3.4520885944366455,
|
| 2219 |
+
"learning_rate": 1.3375000000000002e-05,
|
| 2220 |
+
"loss": 2.0146,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.7915106117353309,
|
| 2225 |
+
"grad_norm": 3.5180275440216064,
|
| 2226 |
+
"learning_rate": 1.3250000000000002e-05,
|
| 2227 |
+
"loss": 2.004,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.7940074906367042,
|
| 2232 |
+
"grad_norm": 3.153761863708496,
|
| 2233 |
+
"learning_rate": 1.3125e-05,
|
| 2234 |
+
"loss": 2.0935,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.7965043695380774,
|
| 2239 |
+
"grad_norm": 3.468968629837036,
|
| 2240 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 2241 |
+
"loss": 2.066,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.7990012484394506,
|
| 2246 |
+
"grad_norm": 3.5003297328948975,
|
| 2247 |
+
"learning_rate": 1.2875000000000001e-05,
|
| 2248 |
+
"loss": 1.9856,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.8014981273408239,
|
| 2253 |
+
"grad_norm": 3.5374064445495605,
|
| 2254 |
+
"learning_rate": 1.2750000000000002e-05,
|
| 2255 |
+
"loss": 2.0671,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.8039950062421972,
|
| 2260 |
+
"grad_norm": 3.659308433532715,
|
| 2261 |
+
"learning_rate": 1.2625e-05,
|
| 2262 |
+
"loss": 1.9864,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.8064918851435705,
|
| 2267 |
+
"grad_norm": 3.5156655311584473,
|
| 2268 |
+
"learning_rate": 1.25e-05,
|
| 2269 |
+
"loss": 1.9394,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.8089887640449438,
|
| 2274 |
+
"grad_norm": 3.326158285140991,
|
| 2275 |
+
"learning_rate": 1.2375000000000001e-05,
|
| 2276 |
+
"loss": 1.961,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.8114856429463171,
|
| 2281 |
+
"grad_norm": 3.4294004440307617,
|
| 2282 |
+
"learning_rate": 1.225e-05,
|
| 2283 |
+
"loss": 2.0097,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.8139825218476904,
|
| 2288 |
+
"grad_norm": 3.4218974113464355,
|
| 2289 |
+
"learning_rate": 1.2125e-05,
|
| 2290 |
+
"loss": 2.0647,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.8164794007490637,
|
| 2295 |
+
"grad_norm": 3.438875198364258,
|
| 2296 |
+
"learning_rate": 1.2e-05,
|
| 2297 |
+
"loss": 2.0179,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.818976279650437,
|
| 2302 |
+
"grad_norm": 3.523358106613159,
|
| 2303 |
+
"learning_rate": 1.1875e-05,
|
| 2304 |
+
"loss": 2.0494,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.8214731585518102,
|
| 2309 |
+
"grad_norm": 3.506168842315674,
|
| 2310 |
+
"learning_rate": 1.175e-05,
|
| 2311 |
+
"loss": 2.1027,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.8239700374531835,
|
| 2316 |
+
"grad_norm": 3.548112154006958,
|
| 2317 |
+
"learning_rate": 1.1625000000000001e-05,
|
| 2318 |
+
"loss": 2.0389,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.8264669163545568,
|
| 2323 |
+
"grad_norm": 3.5158162117004395,
|
| 2324 |
+
"learning_rate": 1.1500000000000002e-05,
|
| 2325 |
+
"loss": 1.978,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.8289637952559301,
|
| 2330 |
+
"grad_norm": 3.4753963947296143,
|
| 2331 |
+
"learning_rate": 1.1375e-05,
|
| 2332 |
+
"loss": 2.0421,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.8314606741573034,
|
| 2337 |
+
"grad_norm": 3.5032706260681152,
|
| 2338 |
+
"learning_rate": 1.125e-05,
|
| 2339 |
+
"loss": 1.9888,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.8339575530586767,
|
| 2344 |
+
"grad_norm": 3.2991061210632324,
|
| 2345 |
+
"learning_rate": 1.1125000000000001e-05,
|
| 2346 |
+
"loss": 2.0151,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.83645443196005,
|
| 2351 |
+
"grad_norm": 3.425863027572632,
|
| 2352 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 2353 |
+
"loss": 2.0503,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.8389513108614233,
|
| 2358 |
+
"grad_norm": 3.469403028488159,
|
| 2359 |
+
"learning_rate": 1.0875e-05,
|
| 2360 |
+
"loss": 1.9425,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.8414481897627965,
|
| 2365 |
+
"grad_norm": 3.3070104122161865,
|
| 2366 |
+
"learning_rate": 1.075e-05,
|
| 2367 |
+
"loss": 2.0603,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.8439450686641697,
|
| 2372 |
+
"grad_norm": 3.418203115463257,
|
| 2373 |
+
"learning_rate": 1.0625e-05,
|
| 2374 |
+
"loss": 2.0018,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.846441947565543,
|
| 2379 |
+
"grad_norm": 3.5066540241241455,
|
| 2380 |
+
"learning_rate": 1.05e-05,
|
| 2381 |
+
"loss": 2.0122,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.8489388264669163,
|
| 2386 |
+
"grad_norm": 3.4425551891326904,
|
| 2387 |
+
"learning_rate": 1.0375e-05,
|
| 2388 |
+
"loss": 2.0225,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.8514357053682896,
|
| 2393 |
+
"grad_norm": 3.343202829360962,
|
| 2394 |
+
"learning_rate": 1.025e-05,
|
| 2395 |
+
"loss": 2.0755,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.8539325842696629,
|
| 2400 |
+
"grad_norm": 3.28155255317688,
|
| 2401 |
+
"learning_rate": 1.0125e-05,
|
| 2402 |
+
"loss": 1.9847,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.8564294631710362,
|
| 2407 |
+
"grad_norm": 3.3546323776245117,
|
| 2408 |
+
"learning_rate": 1e-05,
|
| 2409 |
+
"loss": 1.991,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.8589263420724095,
|
| 2414 |
+
"grad_norm": 3.420783758163452,
|
| 2415 |
+
"learning_rate": 9.875000000000001e-06,
|
| 2416 |
+
"loss": 2.0471,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.8614232209737828,
|
| 2421 |
+
"grad_norm": 3.347703456878662,
|
| 2422 |
+
"learning_rate": 9.750000000000002e-06,
|
| 2423 |
+
"loss": 2.0444,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.8639200998751561,
|
| 2428 |
+
"grad_norm": 3.4276270866394043,
|
| 2429 |
+
"learning_rate": 9.625e-06,
|
| 2430 |
+
"loss": 2.0548,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.8664169787765293,
|
| 2435 |
+
"grad_norm": 3.3684396743774414,
|
| 2436 |
+
"learning_rate": 9.5e-06,
|
| 2437 |
+
"loss": 2.0664,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.8689138576779026,
|
| 2442 |
+
"grad_norm": 3.5308725833892822,
|
| 2443 |
+
"learning_rate": 9.375000000000001e-06,
|
| 2444 |
+
"loss": 2.0263,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.8714107365792759,
|
| 2449 |
+
"grad_norm": 3.5037009716033936,
|
| 2450 |
+
"learning_rate": 9.25e-06,
|
| 2451 |
+
"loss": 2.0629,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.8739076154806492,
|
| 2456 |
+
"grad_norm": 3.3732502460479736,
|
| 2457 |
+
"learning_rate": 9.125e-06,
|
| 2458 |
+
"loss": 2.0872,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.8764044943820225,
|
| 2463 |
+
"grad_norm": 3.379492998123169,
|
| 2464 |
+
"learning_rate": 9e-06,
|
| 2465 |
+
"loss": 2.0494,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.8789013732833958,
|
| 2470 |
+
"grad_norm": 3.356729030609131,
|
| 2471 |
+
"learning_rate": 8.875e-06,
|
| 2472 |
+
"loss": 2.0578,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.8813982521847691,
|
| 2477 |
+
"grad_norm": 3.505227565765381,
|
| 2478 |
+
"learning_rate": 8.75e-06,
|
| 2479 |
+
"loss": 2.0547,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.8838951310861424,
|
| 2484 |
+
"grad_norm": 3.325188159942627,
|
| 2485 |
+
"learning_rate": 8.625e-06,
|
| 2486 |
+
"loss": 2.0479,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.8863920099875156,
|
| 2491 |
+
"grad_norm": 3.220914363861084,
|
| 2492 |
+
"learning_rate": 8.500000000000002e-06,
|
| 2493 |
+
"loss": 2.1145,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.8888888888888888,
|
| 2498 |
+
"grad_norm": 3.2536563873291016,
|
| 2499 |
+
"learning_rate": 8.375e-06,
|
| 2500 |
+
"loss": 2.1226,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.8913857677902621,
|
| 2505 |
+
"grad_norm": 3.4424164295196533,
|
| 2506 |
+
"learning_rate": 8.25e-06,
|
| 2507 |
+
"loss": 1.9732,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.8938826466916354,
|
| 2512 |
+
"grad_norm": 3.210689067840576,
|
| 2513 |
+
"learning_rate": 8.125000000000001e-06,
|
| 2514 |
+
"loss": 2.0682,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.8963795255930087,
|
| 2519 |
+
"grad_norm": 3.6276254653930664,
|
| 2520 |
+
"learning_rate": 8.000000000000001e-06,
|
| 2521 |
+
"loss": 1.9996,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.898876404494382,
|
| 2526 |
+
"grad_norm": 3.284266471862793,
|
| 2527 |
+
"learning_rate": 7.875e-06,
|
| 2528 |
+
"loss": 1.9988,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.9013732833957553,
|
| 2533 |
+
"grad_norm": 3.3102824687957764,
|
| 2534 |
+
"learning_rate": 7.75e-06,
|
| 2535 |
+
"loss": 2.0986,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.9038701622971286,
|
| 2540 |
+
"grad_norm": 3.4718074798583984,
|
| 2541 |
+
"learning_rate": 7.625e-06,
|
| 2542 |
+
"loss": 2.0236,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.9063670411985019,
|
| 2547 |
+
"grad_norm": 3.356231689453125,
|
| 2548 |
+
"learning_rate": 7.5e-06,
|
| 2549 |
+
"loss": 1.9603,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.9088639200998752,
|
| 2554 |
+
"grad_norm": 3.488931655883789,
|
| 2555 |
+
"learning_rate": 7.375e-06,
|
| 2556 |
+
"loss": 2.0287,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.9113607990012484,
|
| 2561 |
+
"grad_norm": 3.370278835296631,
|
| 2562 |
+
"learning_rate": 7.25e-06,
|
| 2563 |
+
"loss": 2.055,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.9138576779026217,
|
| 2568 |
+
"grad_norm": 3.357985019683838,
|
| 2569 |
+
"learning_rate": 7.1249999999999995e-06,
|
| 2570 |
+
"loss": 2.0225,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.916354556803995,
|
| 2575 |
+
"grad_norm": 3.3029778003692627,
|
| 2576 |
+
"learning_rate": 7.000000000000001e-06,
|
| 2577 |
+
"loss": 1.9724,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.9188514357053683,
|
| 2582 |
+
"grad_norm": 3.6066527366638184,
|
| 2583 |
+
"learning_rate": 6.875000000000001e-06,
|
| 2584 |
+
"loss": 2.0172,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.9213483146067416,
|
| 2589 |
+
"grad_norm": 3.4643936157226562,
|
| 2590 |
+
"learning_rate": 6.750000000000001e-06,
|
| 2591 |
+
"loss": 2.0598,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.9238451935081149,
|
| 2596 |
+
"grad_norm": 3.2209830284118652,
|
| 2597 |
+
"learning_rate": 6.625000000000001e-06,
|
| 2598 |
+
"loss": 2.0395,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.9263420724094882,
|
| 2603 |
+
"grad_norm": 3.308652639389038,
|
| 2604 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 2605 |
+
"loss": 2.0576,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.9288389513108615,
|
| 2610 |
+
"grad_norm": 3.378284215927124,
|
| 2611 |
+
"learning_rate": 6.375000000000001e-06,
|
| 2612 |
+
"loss": 2.1037,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.9313358302122348,
|
| 2617 |
+
"grad_norm": 3.408998489379883,
|
| 2618 |
+
"learning_rate": 6.25e-06,
|
| 2619 |
+
"loss": 2.0508,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.9338327091136079,
|
| 2624 |
+
"grad_norm": 3.4758336544036865,
|
| 2625 |
+
"learning_rate": 6.125e-06,
|
| 2626 |
+
"loss": 2.0642,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.9363295880149812,
|
| 2631 |
+
"grad_norm": 3.3644943237304688,
|
| 2632 |
+
"learning_rate": 6e-06,
|
| 2633 |
+
"loss": 2.104,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.9388264669163545,
|
| 2638 |
+
"grad_norm": 3.3621973991394043,
|
| 2639 |
+
"learning_rate": 5.875e-06,
|
| 2640 |
+
"loss": 2.05,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.9413233458177278,
|
| 2645 |
+
"grad_norm": 3.3936052322387695,
|
| 2646 |
+
"learning_rate": 5.750000000000001e-06,
|
| 2647 |
+
"loss": 1.9648,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.9438202247191011,
|
| 2652 |
+
"grad_norm": 3.28177809715271,
|
| 2653 |
+
"learning_rate": 5.625e-06,
|
| 2654 |
+
"loss": 2.0129,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.9463171036204744,
|
| 2659 |
+
"grad_norm": 3.1901121139526367,
|
| 2660 |
+
"learning_rate": 5.500000000000001e-06,
|
| 2661 |
+
"loss": 2.0564,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.9488139825218477,
|
| 2666 |
+
"grad_norm": 3.4150338172912598,
|
| 2667 |
+
"learning_rate": 5.375e-06,
|
| 2668 |
+
"loss": 2.0344,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.951310861423221,
|
| 2673 |
+
"grad_norm": 3.3532180786132812,
|
| 2674 |
+
"learning_rate": 5.25e-06,
|
| 2675 |
+
"loss": 2.012,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.9538077403245943,
|
| 2680 |
+
"grad_norm": 3.434931993484497,
|
| 2681 |
+
"learning_rate": 5.125e-06,
|
| 2682 |
+
"loss": 2.073,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.9563046192259675,
|
| 2687 |
+
"grad_norm": 3.34334135055542,
|
| 2688 |
+
"learning_rate": 5e-06,
|
| 2689 |
+
"loss": 2.0716,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.9588014981273408,
|
| 2694 |
+
"grad_norm": 3.3420920372009277,
|
| 2695 |
+
"learning_rate": 4.875000000000001e-06,
|
| 2696 |
+
"loss": 2.0691,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.9612983770287141,
|
| 2701 |
+
"grad_norm": 3.648798942565918,
|
| 2702 |
+
"learning_rate": 4.75e-06,
|
| 2703 |
+
"loss": 2.0198,
|
| 2704 |
+
"step": 385
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.9637952559300874,
|
| 2708 |
+
"grad_norm": 3.2924463748931885,
|
| 2709 |
+
"learning_rate": 4.625e-06,
|
| 2710 |
+
"loss": 2.015,
|
| 2711 |
+
"step": 386
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.9662921348314607,
|
| 2715 |
+
"grad_norm": 3.2124977111816406,
|
| 2716 |
+
"learning_rate": 4.5e-06,
|
| 2717 |
+
"loss": 2.1013,
|
| 2718 |
+
"step": 387
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.968789013732834,
|
| 2722 |
+
"grad_norm": 3.4171626567840576,
|
| 2723 |
+
"learning_rate": 4.375e-06,
|
| 2724 |
+
"loss": 2.0419,
|
| 2725 |
+
"step": 388
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.9712858926342073,
|
| 2729 |
+
"grad_norm": 3.4217112064361572,
|
| 2730 |
+
"learning_rate": 4.250000000000001e-06,
|
| 2731 |
+
"loss": 2.0808,
|
| 2732 |
+
"step": 389
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.9737827715355806,
|
| 2736 |
+
"grad_norm": 3.281397581100464,
|
| 2737 |
+
"learning_rate": 4.125e-06,
|
| 2738 |
+
"loss": 2.0552,
|
| 2739 |
+
"step": 390
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.9762796504369539,
|
| 2743 |
+
"grad_norm": 3.3025379180908203,
|
| 2744 |
+
"learning_rate": 4.000000000000001e-06,
|
| 2745 |
+
"loss": 2.0196,
|
| 2746 |
+
"step": 391
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.978776529338327,
|
| 2750 |
+
"grad_norm": 3.261284112930298,
|
| 2751 |
+
"learning_rate": 3.875e-06,
|
| 2752 |
+
"loss": 1.903,
|
| 2753 |
+
"step": 392
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.9812734082397003,
|
| 2757 |
+
"grad_norm": 3.3036253452301025,
|
| 2758 |
+
"learning_rate": 3.75e-06,
|
| 2759 |
+
"loss": 2.0455,
|
| 2760 |
+
"step": 393
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.9837702871410736,
|
| 2764 |
+
"grad_norm": 3.4235925674438477,
|
| 2765 |
+
"learning_rate": 3.625e-06,
|
| 2766 |
+
"loss": 2.0393,
|
| 2767 |
+
"step": 394
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.9862671660424469,
|
| 2771 |
+
"grad_norm": 3.347456455230713,
|
| 2772 |
+
"learning_rate": 3.5000000000000004e-06,
|
| 2773 |
+
"loss": 2.0227,
|
| 2774 |
+
"step": 395
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.9887640449438202,
|
| 2778 |
+
"grad_norm": 3.3116769790649414,
|
| 2779 |
+
"learning_rate": 3.3750000000000003e-06,
|
| 2780 |
+
"loss": 2.0891,
|
| 2781 |
+
"step": 396
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.9912609238451935,
|
| 2785 |
+
"grad_norm": 3.2966830730438232,
|
| 2786 |
+
"learning_rate": 3.2500000000000002e-06,
|
| 2787 |
+
"loss": 2.0784,
|
| 2788 |
+
"step": 397
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.9937578027465668,
|
| 2792 |
+
"grad_norm": 3.4097273349761963,
|
| 2793 |
+
"learning_rate": 3.125e-06,
|
| 2794 |
+
"loss": 2.0751,
|
| 2795 |
+
"step": 398
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.9962546816479401,
|
| 2799 |
+
"grad_norm": 3.3749406337738037,
|
| 2800 |
+
"learning_rate": 3e-06,
|
| 2801 |
+
"loss": 2.0916,
|
| 2802 |
+
"step": 399
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.9987515605493134,
|
| 2806 |
+
"grad_norm": 3.322603225708008,
|
| 2807 |
+
"learning_rate": 2.8750000000000004e-06,
|
| 2808 |
+
"loss": 2.0949,
|
| 2809 |
+
"step": 400
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 0.9987515605493134,
|
| 2813 |
+
"step": 400,
|
| 2814 |
+
"total_flos": 2.847507751108608e+17,
|
| 2815 |
+
"train_loss": 2.1486885741353037,
|
| 2816 |
+
"train_runtime": 92894.5901,
|
| 2817 |
+
"train_samples_per_second": 0.241,
|
| 2818 |
+
"train_steps_per_second": 0.004
|
| 2819 |
+
}
|
| 2820 |
+
],
|
| 2821 |
+
"logging_steps": 1.0,
|
| 2822 |
+
"max_steps": 400,
|
| 2823 |
+
"num_input_tokens_seen": 0,
|
| 2824 |
+
"num_train_epochs": 1,
|
| 2825 |
+
"save_steps": 400,
|
| 2826 |
+
"stateful_callbacks": {
|
| 2827 |
+
"TrainerControl": {
|
| 2828 |
+
"args": {
|
| 2829 |
+
"should_epoch_stop": false,
|
| 2830 |
+
"should_evaluate": false,
|
| 2831 |
+
"should_log": false,
|
| 2832 |
+
"should_save": true,
|
| 2833 |
+
"should_training_stop": true
|
| 2834 |
+
},
|
| 2835 |
+
"attributes": {}
|
| 2836 |
+
}
|
| 2837 |
+
},
|
| 2838 |
+
"total_flos": 2.847507751108608e+17,
|
| 2839 |
+
"train_batch_size": 14,
|
| 2840 |
+
"trial_name": null,
|
| 2841 |
+
"trial_params": null
|
| 2842 |
+
}
|