Commit
•
831a441
1
Parent(s):
b8126ae
End of training
Browse files- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1574 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.998954339491112,
|
| 3 |
+
"eval_accuracy": 0.8684313725490196,
|
| 4 |
+
"eval_loss": 0.5095656514167786,
|
| 5 |
+
"eval_runtime": 203.3606,
|
| 6 |
+
"eval_samples_per_second": 50.157,
|
| 7 |
+
"eval_steps_per_second": 1.569,
|
| 8 |
+
"total_flos": 2.1353596059581743e+19,
|
| 9 |
+
"train_loss": 4.144423897296315,
|
| 10 |
+
"train_runtime": 8280.2172,
|
| 11 |
+
"train_samples_per_second": 33.26,
|
| 12 |
+
"train_steps_per_second": 0.26
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.998954339491112,
|
| 3 |
+
"eval_accuracy": 0.8684313725490196,
|
| 4 |
+
"eval_loss": 0.5095656514167786,
|
| 5 |
+
"eval_runtime": 203.3606,
|
| 6 |
+
"eval_samples_per_second": 50.157,
|
| 7 |
+
"eval_steps_per_second": 1.569
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.998954339491112,
|
| 3 |
+
"total_flos": 2.1353596059581743e+19,
|
| 4 |
+
"train_loss": 4.144423897296315,
|
| 5 |
+
"train_runtime": 8280.2172,
|
| 6 |
+
"train_samples_per_second": 33.26,
|
| 7 |
+
"train_steps_per_second": 0.26
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1574 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8684313725490196,
|
| 3 |
+
"best_model_checkpoint": "vit-base-patch16-224-finetuned-food102/checkpoint-2151",
|
| 4 |
+
"epoch": 2.998954339491112,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2151,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.013942140118508192,
|
| 13 |
+
"grad_norm": 11.970505714416504,
|
| 14 |
+
"learning_rate": 2.3148148148148148e-06,
|
| 15 |
+
"loss": 19.054,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.027884280237016383,
|
| 20 |
+
"grad_norm": 11.977810859680176,
|
| 21 |
+
"learning_rate": 4.6296296296296296e-06,
|
| 22 |
+
"loss": 19.0411,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04182642035552457,
|
| 27 |
+
"grad_norm": 12.143768310546875,
|
| 28 |
+
"learning_rate": 6.944444444444445e-06,
|
| 29 |
+
"loss": 19.0188,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.055768560474032766,
|
| 34 |
+
"grad_norm": 11.653632164001465,
|
| 35 |
+
"learning_rate": 9.259259259259259e-06,
|
| 36 |
+
"loss": 18.7585,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.06971070059254096,
|
| 41 |
+
"grad_norm": 13.25544261932373,
|
| 42 |
+
"learning_rate": 1.1574074074074075e-05,
|
| 43 |
+
"loss": 18.4989,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.08365284071104914,
|
| 48 |
+
"grad_norm": 12.296609878540039,
|
| 49 |
+
"learning_rate": 1.388888888888889e-05,
|
| 50 |
+
"loss": 18.199,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.09759498082955734,
|
| 55 |
+
"grad_norm": 11.88782024383545,
|
| 56 |
+
"learning_rate": 1.6203703703703704e-05,
|
| 57 |
+
"loss": 17.9952,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.11153712094806553,
|
| 62 |
+
"grad_norm": 11.943421363830566,
|
| 63 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 64 |
+
"loss": 17.5776,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.12547926106657373,
|
| 69 |
+
"grad_norm": 12.558123588562012,
|
| 70 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 71 |
+
"loss": 17.0623,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.13942140118508192,
|
| 76 |
+
"grad_norm": 13.311689376831055,
|
| 77 |
+
"learning_rate": 2.314814814814815e-05,
|
| 78 |
+
"loss": 16.5161,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1533635413035901,
|
| 83 |
+
"grad_norm": 14.52517318725586,
|
| 84 |
+
"learning_rate": 2.5462962962962965e-05,
|
| 85 |
+
"loss": 15.9671,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.16730568142209828,
|
| 90 |
+
"grad_norm": 12.310554504394531,
|
| 91 |
+
"learning_rate": 2.777777777777778e-05,
|
| 92 |
+
"loss": 15.2662,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.18124782154060648,
|
| 97 |
+
"grad_norm": 12.67236042022705,
|
| 98 |
+
"learning_rate": 3.0092592592592593e-05,
|
| 99 |
+
"loss": 14.4554,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.19518996165911467,
|
| 104 |
+
"grad_norm": 12.847444534301758,
|
| 105 |
+
"learning_rate": 3.240740740740741e-05,
|
| 106 |
+
"loss": 13.7588,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.20913210177762287,
|
| 111 |
+
"grad_norm": 12.77316665649414,
|
| 112 |
+
"learning_rate": 3.472222222222222e-05,
|
| 113 |
+
"loss": 13.0135,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.22307424189613106,
|
| 118 |
+
"grad_norm": 13.008152961730957,
|
| 119 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 120 |
+
"loss": 12.324,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.23701638201463926,
|
| 125 |
+
"grad_norm": 13.326498985290527,
|
| 126 |
+
"learning_rate": 3.935185185185186e-05,
|
| 127 |
+
"loss": 11.2685,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.25095852213314745,
|
| 132 |
+
"grad_norm": 13.166327476501465,
|
| 133 |
+
"learning_rate": 4.166666666666667e-05,
|
| 134 |
+
"loss": 10.491,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.26490066225165565,
|
| 139 |
+
"grad_norm": 12.491602897644043,
|
| 140 |
+
"learning_rate": 4.3981481481481486e-05,
|
| 141 |
+
"loss": 9.8871,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.27884280237016384,
|
| 146 |
+
"grad_norm": 12.803274154663086,
|
| 147 |
+
"learning_rate": 4.62962962962963e-05,
|
| 148 |
+
"loss": 9.0965,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.292784942488672,
|
| 153 |
+
"grad_norm": 13.318122863769531,
|
| 154 |
+
"learning_rate": 4.8611111111111115e-05,
|
| 155 |
+
"loss": 8.2912,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.3067270826071802,
|
| 160 |
+
"grad_norm": 12.178855895996094,
|
| 161 |
+
"learning_rate": 4.989664082687339e-05,
|
| 162 |
+
"loss": 8.0104,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3206692227256884,
|
| 167 |
+
"grad_norm": 11.760112762451172,
|
| 168 |
+
"learning_rate": 4.963824289405685e-05,
|
| 169 |
+
"loss": 7.2478,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.33461136284419657,
|
| 174 |
+
"grad_norm": 12.436824798583984,
|
| 175 |
+
"learning_rate": 4.9379844961240315e-05,
|
| 176 |
+
"loss": 6.8172,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.34855350296270476,
|
| 181 |
+
"grad_norm": 11.301292419433594,
|
| 182 |
+
"learning_rate": 4.912144702842377e-05,
|
| 183 |
+
"loss": 6.5386,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.36249564308121296,
|
| 188 |
+
"grad_norm": 12.099160194396973,
|
| 189 |
+
"learning_rate": 4.886304909560724e-05,
|
| 190 |
+
"loss": 6.1708,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.37643778319972115,
|
| 195 |
+
"grad_norm": 12.640030860900879,
|
| 196 |
+
"learning_rate": 4.86046511627907e-05,
|
| 197 |
+
"loss": 6.0798,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.39037992331822935,
|
| 202 |
+
"grad_norm": 12.914190292358398,
|
| 203 |
+
"learning_rate": 4.834625322997416e-05,
|
| 204 |
+
"loss": 5.5983,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.40432206343673754,
|
| 209 |
+
"grad_norm": 13.290894508361816,
|
| 210 |
+
"learning_rate": 4.8087855297157624e-05,
|
| 211 |
+
"loss": 5.6591,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.41826420355524574,
|
| 216 |
+
"grad_norm": 12.26599407196045,
|
| 217 |
+
"learning_rate": 4.782945736434109e-05,
|
| 218 |
+
"loss": 5.0549,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.43220634367375393,
|
| 223 |
+
"grad_norm": 13.02130126953125,
|
| 224 |
+
"learning_rate": 4.757105943152455e-05,
|
| 225 |
+
"loss": 4.8798,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.44614848379226213,
|
| 230 |
+
"grad_norm": 11.812455177307129,
|
| 231 |
+
"learning_rate": 4.7312661498708014e-05,
|
| 232 |
+
"loss": 4.8651,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.4600906239107703,
|
| 237 |
+
"grad_norm": 15.11469554901123,
|
| 238 |
+
"learning_rate": 4.705426356589148e-05,
|
| 239 |
+
"loss": 4.8677,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.4740327640292785,
|
| 244 |
+
"grad_norm": 13.028066635131836,
|
| 245 |
+
"learning_rate": 4.679586563307494e-05,
|
| 246 |
+
"loss": 4.7999,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.48797490414778666,
|
| 251 |
+
"grad_norm": 14.58755111694336,
|
| 252 |
+
"learning_rate": 4.65374677002584e-05,
|
| 253 |
+
"loss": 4.7765,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5019170442662949,
|
| 258 |
+
"grad_norm": 11.573223114013672,
|
| 259 |
+
"learning_rate": 4.627906976744186e-05,
|
| 260 |
+
"loss": 4.6033,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.515859184384803,
|
| 265 |
+
"grad_norm": 13.456429481506348,
|
| 266 |
+
"learning_rate": 4.602067183462532e-05,
|
| 267 |
+
"loss": 4.406,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.5298013245033113,
|
| 272 |
+
"grad_norm": 13.06003475189209,
|
| 273 |
+
"learning_rate": 4.5762273901808786e-05,
|
| 274 |
+
"loss": 4.4271,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.5437434646218194,
|
| 279 |
+
"grad_norm": 12.669313430786133,
|
| 280 |
+
"learning_rate": 4.550387596899225e-05,
|
| 281 |
+
"loss": 4.1339,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.5576856047403277,
|
| 286 |
+
"grad_norm": 12.845880508422852,
|
| 287 |
+
"learning_rate": 4.524547803617571e-05,
|
| 288 |
+
"loss": 4.3616,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.5716277448588358,
|
| 293 |
+
"grad_norm": 13.643671035766602,
|
| 294 |
+
"learning_rate": 4.4987080103359175e-05,
|
| 295 |
+
"loss": 4.03,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.585569884977344,
|
| 300 |
+
"grad_norm": 12.903581619262695,
|
| 301 |
+
"learning_rate": 4.472868217054264e-05,
|
| 302 |
+
"loss": 4.0456,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.5995120250958522,
|
| 307 |
+
"grad_norm": 11.996821403503418,
|
| 308 |
+
"learning_rate": 4.44702842377261e-05,
|
| 309 |
+
"loss": 4.1508,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.6134541652143604,
|
| 314 |
+
"grad_norm": 12.505786895751953,
|
| 315 |
+
"learning_rate": 4.421188630490956e-05,
|
| 316 |
+
"loss": 4.112,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.6273963053328686,
|
| 321 |
+
"grad_norm": 12.409116744995117,
|
| 322 |
+
"learning_rate": 4.395348837209303e-05,
|
| 323 |
+
"loss": 3.8887,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.6413384454513767,
|
| 328 |
+
"grad_norm": 12.7697114944458,
|
| 329 |
+
"learning_rate": 4.3695090439276484e-05,
|
| 330 |
+
"loss": 3.7454,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.655280585569885,
|
| 335 |
+
"grad_norm": 13.008294105529785,
|
| 336 |
+
"learning_rate": 4.3436692506459954e-05,
|
| 337 |
+
"loss": 3.879,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.6692227256883931,
|
| 342 |
+
"grad_norm": 12.83389949798584,
|
| 343 |
+
"learning_rate": 4.317829457364341e-05,
|
| 344 |
+
"loss": 4.032,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.6831648658069014,
|
| 349 |
+
"grad_norm": 13.29801082611084,
|
| 350 |
+
"learning_rate": 4.2919896640826874e-05,
|
| 351 |
+
"loss": 3.9523,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.6971070059254095,
|
| 356 |
+
"grad_norm": 13.225083351135254,
|
| 357 |
+
"learning_rate": 4.266149870801034e-05,
|
| 358 |
+
"loss": 3.7454,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.7110491460439178,
|
| 363 |
+
"grad_norm": 13.245360374450684,
|
| 364 |
+
"learning_rate": 4.24031007751938e-05,
|
| 365 |
+
"loss": 3.5761,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.7249912861624259,
|
| 370 |
+
"grad_norm": 12.166769981384277,
|
| 371 |
+
"learning_rate": 4.214470284237726e-05,
|
| 372 |
+
"loss": 3.8125,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.7389334262809342,
|
| 377 |
+
"grad_norm": 13.542282104492188,
|
| 378 |
+
"learning_rate": 4.1886304909560726e-05,
|
| 379 |
+
"loss": 3.6979,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.7528755663994423,
|
| 384 |
+
"grad_norm": 17.77124786376953,
|
| 385 |
+
"learning_rate": 4.162790697674418e-05,
|
| 386 |
+
"loss": 3.4578,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.7668177065179506,
|
| 391 |
+
"grad_norm": 13.022871971130371,
|
| 392 |
+
"learning_rate": 4.136950904392765e-05,
|
| 393 |
+
"loss": 3.4832,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.7807598466364587,
|
| 398 |
+
"grad_norm": 15.890631675720215,
|
| 399 |
+
"learning_rate": 4.111111111111111e-05,
|
| 400 |
+
"loss": 3.5997,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.7947019867549668,
|
| 405 |
+
"grad_norm": 14.40475845336914,
|
| 406 |
+
"learning_rate": 4.085271317829458e-05,
|
| 407 |
+
"loss": 3.6687,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.8086441268734751,
|
| 412 |
+
"grad_norm": 10.683858871459961,
|
| 413 |
+
"learning_rate": 4.0594315245478035e-05,
|
| 414 |
+
"loss": 3.3821,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.8225862669919832,
|
| 419 |
+
"grad_norm": 14.170791625976562,
|
| 420 |
+
"learning_rate": 4.0335917312661505e-05,
|
| 421 |
+
"loss": 3.4403,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.8365284071104915,
|
| 426 |
+
"grad_norm": 12.744939804077148,
|
| 427 |
+
"learning_rate": 4.007751937984496e-05,
|
| 428 |
+
"loss": 3.1932,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.8504705472289996,
|
| 433 |
+
"grad_norm": 12.260272979736328,
|
| 434 |
+
"learning_rate": 3.9819121447028425e-05,
|
| 435 |
+
"loss": 3.5403,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.8644126873475079,
|
| 440 |
+
"grad_norm": 14.6715669631958,
|
| 441 |
+
"learning_rate": 3.956072351421189e-05,
|
| 442 |
+
"loss": 3.4033,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.878354827466016,
|
| 447 |
+
"grad_norm": 12.32928466796875,
|
| 448 |
+
"learning_rate": 3.930232558139535e-05,
|
| 449 |
+
"loss": 3.2728,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.8922969675845243,
|
| 454 |
+
"grad_norm": 13.14637279510498,
|
| 455 |
+
"learning_rate": 3.9043927648578814e-05,
|
| 456 |
+
"loss": 3.3631,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.9062391077030324,
|
| 461 |
+
"grad_norm": 11.95271110534668,
|
| 462 |
+
"learning_rate": 3.878552971576228e-05,
|
| 463 |
+
"loss": 3.2798,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.9201812478215406,
|
| 468 |
+
"grad_norm": 12.834661483764648,
|
| 469 |
+
"learning_rate": 3.8527131782945734e-05,
|
| 470 |
+
"loss": 3.5264,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.9341233879400488,
|
| 475 |
+
"grad_norm": 13.227917671203613,
|
| 476 |
+
"learning_rate": 3.8268733850129204e-05,
|
| 477 |
+
"loss": 3.5221,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.948065528058557,
|
| 482 |
+
"grad_norm": 13.852123260498047,
|
| 483 |
+
"learning_rate": 3.801033591731266e-05,
|
| 484 |
+
"loss": 3.1419,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.9620076681770652,
|
| 489 |
+
"grad_norm": 16.48082160949707,
|
| 490 |
+
"learning_rate": 3.775193798449613e-05,
|
| 491 |
+
"loss": 3.4653,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.9759498082955733,
|
| 496 |
+
"grad_norm": 11.60905933380127,
|
| 497 |
+
"learning_rate": 3.7493540051679586e-05,
|
| 498 |
+
"loss": 3.1008,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.9898919484140816,
|
| 503 |
+
"grad_norm": 25.217321395874023,
|
| 504 |
+
"learning_rate": 3.7235142118863056e-05,
|
| 505 |
+
"loss": 3.3941,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.9996514464970373,
|
| 510 |
+
"eval_accuracy": 0.8350980392156863,
|
| 511 |
+
"eval_loss": 0.6625040769577026,
|
| 512 |
+
"eval_runtime": 286.4591,
|
| 513 |
+
"eval_samples_per_second": 35.607,
|
| 514 |
+
"eval_steps_per_second": 1.114,
|
| 515 |
+
"step": 717
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.0038340885325898,
|
| 519 |
+
"grad_norm": 10.983098983764648,
|
| 520 |
+
"learning_rate": 3.697674418604651e-05,
|
| 521 |
+
"loss": 3.2282,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.0177762286510978,
|
| 526 |
+
"grad_norm": 12.984674453735352,
|
| 527 |
+
"learning_rate": 3.6718346253229976e-05,
|
| 528 |
+
"loss": 2.9876,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.031718368769606,
|
| 533 |
+
"grad_norm": 11.363933563232422,
|
| 534 |
+
"learning_rate": 3.645994832041344e-05,
|
| 535 |
+
"loss": 3.1042,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.0456605088881143,
|
| 540 |
+
"grad_norm": 12.959346771240234,
|
| 541 |
+
"learning_rate": 3.62015503875969e-05,
|
| 542 |
+
"loss": 2.8314,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.0596026490066226,
|
| 547 |
+
"grad_norm": 13.28182601928711,
|
| 548 |
+
"learning_rate": 3.5943152454780365e-05,
|
| 549 |
+
"loss": 2.9215,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.0735447891251306,
|
| 554 |
+
"grad_norm": 12.953139305114746,
|
| 555 |
+
"learning_rate": 3.568475452196383e-05,
|
| 556 |
+
"loss": 2.9363,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.0874869292436389,
|
| 561 |
+
"grad_norm": 12.209274291992188,
|
| 562 |
+
"learning_rate": 3.5426356589147285e-05,
|
| 563 |
+
"loss": 2.9331,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.1014290693621471,
|
| 568 |
+
"grad_norm": 12.68194580078125,
|
| 569 |
+
"learning_rate": 3.5167958656330755e-05,
|
| 570 |
+
"loss": 2.9515,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.1153712094806554,
|
| 575 |
+
"grad_norm": 11.559549331665039,
|
| 576 |
+
"learning_rate": 3.490956072351421e-05,
|
| 577 |
+
"loss": 2.7768,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.1293133495991634,
|
| 582 |
+
"grad_norm": 13.342930793762207,
|
| 583 |
+
"learning_rate": 3.465116279069768e-05,
|
| 584 |
+
"loss": 2.8238,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.1432554897176717,
|
| 589 |
+
"grad_norm": 14.305107116699219,
|
| 590 |
+
"learning_rate": 3.439276485788114e-05,
|
| 591 |
+
"loss": 2.9598,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.15719762983618,
|
| 596 |
+
"grad_norm": 12.436918258666992,
|
| 597 |
+
"learning_rate": 3.41343669250646e-05,
|
| 598 |
+
"loss": 2.8277,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.1711397699546882,
|
| 603 |
+
"grad_norm": 12.794795036315918,
|
| 604 |
+
"learning_rate": 3.3875968992248064e-05,
|
| 605 |
+
"loss": 2.7674,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.1850819100731962,
|
| 610 |
+
"grad_norm": 12.380959510803223,
|
| 611 |
+
"learning_rate": 3.361757105943153e-05,
|
| 612 |
+
"loss": 2.6084,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.1990240501917044,
|
| 617 |
+
"grad_norm": 16.531293869018555,
|
| 618 |
+
"learning_rate": 3.335917312661499e-05,
|
| 619 |
+
"loss": 2.6154,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.2129661903102127,
|
| 624 |
+
"grad_norm": 12.086883544921875,
|
| 625 |
+
"learning_rate": 3.310077519379845e-05,
|
| 626 |
+
"loss": 2.9116,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.2269083304287207,
|
| 631 |
+
"grad_norm": 11.90904712677002,
|
| 632 |
+
"learning_rate": 3.2842377260981916e-05,
|
| 633 |
+
"loss": 2.7052,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.240850470547229,
|
| 638 |
+
"grad_norm": 13.168009757995605,
|
| 639 |
+
"learning_rate": 3.258397932816537e-05,
|
| 640 |
+
"loss": 2.5401,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.2547926106657372,
|
| 645 |
+
"grad_norm": 13.133540153503418,
|
| 646 |
+
"learning_rate": 3.2325581395348836e-05,
|
| 647 |
+
"loss": 3.089,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.2687347507842452,
|
| 652 |
+
"grad_norm": 10.270415306091309,
|
| 653 |
+
"learning_rate": 3.20671834625323e-05,
|
| 654 |
+
"loss": 2.5938,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.2826768909027535,
|
| 659 |
+
"grad_norm": 13.742632865905762,
|
| 660 |
+
"learning_rate": 3.180878552971576e-05,
|
| 661 |
+
"loss": 2.6841,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.2966190310212617,
|
| 666 |
+
"grad_norm": 10.14291000366211,
|
| 667 |
+
"learning_rate": 3.1550387596899225e-05,
|
| 668 |
+
"loss": 2.5288,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.31056117113977,
|
| 673 |
+
"grad_norm": 11.036438941955566,
|
| 674 |
+
"learning_rate": 3.129198966408269e-05,
|
| 675 |
+
"loss": 2.8172,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.3245033112582782,
|
| 680 |
+
"grad_norm": 11.163543701171875,
|
| 681 |
+
"learning_rate": 3.103359173126615e-05,
|
| 682 |
+
"loss": 2.8402,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.3384454513767863,
|
| 687 |
+
"grad_norm": 16.406246185302734,
|
| 688 |
+
"learning_rate": 3.0775193798449615e-05,
|
| 689 |
+
"loss": 2.8161,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.3523875914952945,
|
| 694 |
+
"grad_norm": 11.284817695617676,
|
| 695 |
+
"learning_rate": 3.0516795865633074e-05,
|
| 696 |
+
"loss": 2.448,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.3663297316138028,
|
| 701 |
+
"grad_norm": 12.130209922790527,
|
| 702 |
+
"learning_rate": 3.025839793281654e-05,
|
| 703 |
+
"loss": 2.8401,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.3802718717323108,
|
| 708 |
+
"grad_norm": 11.599443435668945,
|
| 709 |
+
"learning_rate": 3e-05,
|
| 710 |
+
"loss": 2.6025,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.394214011850819,
|
| 715 |
+
"grad_norm": 11.705489158630371,
|
| 716 |
+
"learning_rate": 2.9741602067183467e-05,
|
| 717 |
+
"loss": 2.6507,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.4081561519693273,
|
| 722 |
+
"grad_norm": 11.031926155090332,
|
| 723 |
+
"learning_rate": 2.9483204134366927e-05,
|
| 724 |
+
"loss": 2.8557,
|
| 725 |
+
"step": 1010
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.4220982920878356,
|
| 729 |
+
"grad_norm": 13.046774864196777,
|
| 730 |
+
"learning_rate": 2.9224806201550387e-05,
|
| 731 |
+
"loss": 2.972,
|
| 732 |
+
"step": 1020
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.4360404322063438,
|
| 736 |
+
"grad_norm": 12.700828552246094,
|
| 737 |
+
"learning_rate": 2.8966408268733853e-05,
|
| 738 |
+
"loss": 2.7807,
|
| 739 |
+
"step": 1030
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.4499825723248518,
|
| 743 |
+
"grad_norm": 15.926816940307617,
|
| 744 |
+
"learning_rate": 2.8708010335917313e-05,
|
| 745 |
+
"loss": 2.5443,
|
| 746 |
+
"step": 1040
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.46392471244336,
|
| 750 |
+
"grad_norm": 14.657876968383789,
|
| 751 |
+
"learning_rate": 2.844961240310078e-05,
|
| 752 |
+
"loss": 2.8499,
|
| 753 |
+
"step": 1050
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.4778668525618683,
|
| 757 |
+
"grad_norm": 12.2283935546875,
|
| 758 |
+
"learning_rate": 2.819121447028424e-05,
|
| 759 |
+
"loss": 2.6669,
|
| 760 |
+
"step": 1060
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.4918089926803764,
|
| 764 |
+
"grad_norm": 14.132109642028809,
|
| 765 |
+
"learning_rate": 2.7932816537467703e-05,
|
| 766 |
+
"loss": 2.75,
|
| 767 |
+
"step": 1070
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.5057511327988846,
|
| 771 |
+
"grad_norm": 15.497398376464844,
|
| 772 |
+
"learning_rate": 2.7674418604651166e-05,
|
| 773 |
+
"loss": 2.6658,
|
| 774 |
+
"step": 1080
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.5196932729173929,
|
| 778 |
+
"grad_norm": 39.895477294921875,
|
| 779 |
+
"learning_rate": 2.7416020671834625e-05,
|
| 780 |
+
"loss": 2.8124,
|
| 781 |
+
"step": 1090
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.533635413035901,
|
| 785 |
+
"grad_norm": 12.836750030517578,
|
| 786 |
+
"learning_rate": 2.7157622739018092e-05,
|
| 787 |
+
"loss": 2.7151,
|
| 788 |
+
"step": 1100
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.5475775531544091,
|
| 792 |
+
"grad_norm": 12.65983772277832,
|
| 793 |
+
"learning_rate": 2.6899224806201552e-05,
|
| 794 |
+
"loss": 2.5187,
|
| 795 |
+
"step": 1110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.5615196932729174,
|
| 799 |
+
"grad_norm": 12.410466194152832,
|
| 800 |
+
"learning_rate": 2.6640826873385015e-05,
|
| 801 |
+
"loss": 2.7021,
|
| 802 |
+
"step": 1120
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.5754618333914254,
|
| 806 |
+
"grad_norm": 16.45282554626465,
|
| 807 |
+
"learning_rate": 2.6382428940568478e-05,
|
| 808 |
+
"loss": 2.8277,
|
| 809 |
+
"step": 1130
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.589403973509934,
|
| 813 |
+
"grad_norm": 12.401958465576172,
|
| 814 |
+
"learning_rate": 2.6124031007751938e-05,
|
| 815 |
+
"loss": 2.6465,
|
| 816 |
+
"step": 1140
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.603346113628442,
|
| 820 |
+
"grad_norm": 13.451824188232422,
|
| 821 |
+
"learning_rate": 2.58656330749354e-05,
|
| 822 |
+
"loss": 2.7568,
|
| 823 |
+
"step": 1150
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.6172882537469502,
|
| 827 |
+
"grad_norm": 12.007116317749023,
|
| 828 |
+
"learning_rate": 2.5607235142118864e-05,
|
| 829 |
+
"loss": 2.5732,
|
| 830 |
+
"step": 1160
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.6312303938654584,
|
| 834 |
+
"grad_norm": 13.369595527648926,
|
| 835 |
+
"learning_rate": 2.5348837209302327e-05,
|
| 836 |
+
"loss": 2.5993,
|
| 837 |
+
"step": 1170
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.6451725339839665,
|
| 841 |
+
"grad_norm": 13.079323768615723,
|
| 842 |
+
"learning_rate": 2.5090439276485787e-05,
|
| 843 |
+
"loss": 2.557,
|
| 844 |
+
"step": 1180
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.6591146741024747,
|
| 848 |
+
"grad_norm": 12.571920394897461,
|
| 849 |
+
"learning_rate": 2.483204134366925e-05,
|
| 850 |
+
"loss": 2.5582,
|
| 851 |
+
"step": 1190
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.673056814220983,
|
| 855 |
+
"grad_norm": 12.451131820678711,
|
| 856 |
+
"learning_rate": 2.4573643410852713e-05,
|
| 857 |
+
"loss": 2.485,
|
| 858 |
+
"step": 1200
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.686998954339491,
|
| 862 |
+
"grad_norm": 11.426173210144043,
|
| 863 |
+
"learning_rate": 2.4315245478036177e-05,
|
| 864 |
+
"loss": 2.8228,
|
| 865 |
+
"step": 1210
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.7009410944579995,
|
| 869 |
+
"grad_norm": 13.152283668518066,
|
| 870 |
+
"learning_rate": 2.405684754521964e-05,
|
| 871 |
+
"loss": 2.9743,
|
| 872 |
+
"step": 1220
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.7148832345765075,
|
| 876 |
+
"grad_norm": 12.218315124511719,
|
| 877 |
+
"learning_rate": 2.37984496124031e-05,
|
| 878 |
+
"loss": 2.5617,
|
| 879 |
+
"step": 1230
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.7288253746950157,
|
| 883 |
+
"grad_norm": 16.104169845581055,
|
| 884 |
+
"learning_rate": 2.3540051679586563e-05,
|
| 885 |
+
"loss": 2.5886,
|
| 886 |
+
"step": 1240
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.742767514813524,
|
| 890 |
+
"grad_norm": 9.216697692871094,
|
| 891 |
+
"learning_rate": 2.3281653746770026e-05,
|
| 892 |
+
"loss": 2.2698,
|
| 893 |
+
"step": 1250
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.756709654932032,
|
| 897 |
+
"grad_norm": 12.568550109863281,
|
| 898 |
+
"learning_rate": 2.302325581395349e-05,
|
| 899 |
+
"loss": 2.7008,
|
| 900 |
+
"step": 1260
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.7706517950505403,
|
| 904 |
+
"grad_norm": 13.100058555603027,
|
| 905 |
+
"learning_rate": 2.2764857881136952e-05,
|
| 906 |
+
"loss": 2.5122,
|
| 907 |
+
"step": 1270
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.7845939351690485,
|
| 911 |
+
"grad_norm": 14.60105037689209,
|
| 912 |
+
"learning_rate": 2.2506459948320415e-05,
|
| 913 |
+
"loss": 2.7869,
|
| 914 |
+
"step": 1280
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.7985360752875565,
|
| 918 |
+
"grad_norm": 11.809743881225586,
|
| 919 |
+
"learning_rate": 2.2248062015503875e-05,
|
| 920 |
+
"loss": 3.0241,
|
| 921 |
+
"step": 1290
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.8124782154060648,
|
| 925 |
+
"grad_norm": 11.19316291809082,
|
| 926 |
+
"learning_rate": 2.1989664082687338e-05,
|
| 927 |
+
"loss": 2.3842,
|
| 928 |
+
"step": 1300
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.826420355524573,
|
| 932 |
+
"grad_norm": 11.623698234558105,
|
| 933 |
+
"learning_rate": 2.17312661498708e-05,
|
| 934 |
+
"loss": 2.5165,
|
| 935 |
+
"step": 1310
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.840362495643081,
|
| 939 |
+
"grad_norm": 12.553630828857422,
|
| 940 |
+
"learning_rate": 2.1472868217054264e-05,
|
| 941 |
+
"loss": 2.6636,
|
| 942 |
+
"step": 1320
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.8543046357615895,
|
| 946 |
+
"grad_norm": 16.92484474182129,
|
| 947 |
+
"learning_rate": 2.1214470284237728e-05,
|
| 948 |
+
"loss": 2.8168,
|
| 949 |
+
"step": 1330
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.8682467758800976,
|
| 953 |
+
"grad_norm": 11.145873069763184,
|
| 954 |
+
"learning_rate": 2.095607235142119e-05,
|
| 955 |
+
"loss": 2.498,
|
| 956 |
+
"step": 1340
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.8821889159986058,
|
| 960 |
+
"grad_norm": 13.681021690368652,
|
| 961 |
+
"learning_rate": 2.069767441860465e-05,
|
| 962 |
+
"loss": 2.5061,
|
| 963 |
+
"step": 1350
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.896131056117114,
|
| 967 |
+
"grad_norm": 13.941627502441406,
|
| 968 |
+
"learning_rate": 2.0439276485788114e-05,
|
| 969 |
+
"loss": 2.5961,
|
| 970 |
+
"step": 1360
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.910073196235622,
|
| 974 |
+
"grad_norm": 10.970531463623047,
|
| 975 |
+
"learning_rate": 2.0180878552971577e-05,
|
| 976 |
+
"loss": 2.5375,
|
| 977 |
+
"step": 1370
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.9240153363541304,
|
| 981 |
+
"grad_norm": 11.011059761047363,
|
| 982 |
+
"learning_rate": 1.992248062015504e-05,
|
| 983 |
+
"loss": 2.6054,
|
| 984 |
+
"step": 1380
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.9379574764726386,
|
| 988 |
+
"grad_norm": 11.878082275390625,
|
| 989 |
+
"learning_rate": 1.9664082687338503e-05,
|
| 990 |
+
"loss": 2.487,
|
| 991 |
+
"step": 1390
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.9518996165911466,
|
| 995 |
+
"grad_norm": 12.48755931854248,
|
| 996 |
+
"learning_rate": 1.9405684754521966e-05,
|
| 997 |
+
"loss": 2.8289,
|
| 998 |
+
"step": 1400
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.9658417567096549,
|
| 1002 |
+
"grad_norm": 14.737470626831055,
|
| 1003 |
+
"learning_rate": 1.9147286821705426e-05,
|
| 1004 |
+
"loss": 2.5594,
|
| 1005 |
+
"step": 1410
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.9797838968281631,
|
| 1009 |
+
"grad_norm": 11.880918502807617,
|
| 1010 |
+
"learning_rate": 1.888888888888889e-05,
|
| 1011 |
+
"loss": 2.4736,
|
| 1012 |
+
"step": 1420
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.9937260369466712,
|
| 1016 |
+
"grad_norm": 14.554669380187988,
|
| 1017 |
+
"learning_rate": 1.8630490956072352e-05,
|
| 1018 |
+
"loss": 2.6442,
|
| 1019 |
+
"step": 1430
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.9993028929940746,
|
| 1023 |
+
"eval_accuracy": 0.8597058823529412,
|
| 1024 |
+
"eval_loss": 0.5420134663581848,
|
| 1025 |
+
"eval_runtime": 226.1399,
|
| 1026 |
+
"eval_samples_per_second": 45.105,
|
| 1027 |
+
"eval_steps_per_second": 1.411,
|
| 1028 |
+
"step": 1434
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.0076681770651796,
|
| 1032 |
+
"grad_norm": 11.852781295776367,
|
| 1033 |
+
"learning_rate": 1.8372093023255815e-05,
|
| 1034 |
+
"loss": 2.2307,
|
| 1035 |
+
"step": 1440
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 2.0216103171836877,
|
| 1039 |
+
"grad_norm": 12.653249740600586,
|
| 1040 |
+
"learning_rate": 1.811369509043928e-05,
|
| 1041 |
+
"loss": 2.3448,
|
| 1042 |
+
"step": 1450
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 2.0355524573021957,
|
| 1046 |
+
"grad_norm": 15.625718116760254,
|
| 1047 |
+
"learning_rate": 1.7855297157622742e-05,
|
| 1048 |
+
"loss": 2.4701,
|
| 1049 |
+
"step": 1460
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 2.049494597420704,
|
| 1053 |
+
"grad_norm": 10.152451515197754,
|
| 1054 |
+
"learning_rate": 1.75968992248062e-05,
|
| 1055 |
+
"loss": 2.2279,
|
| 1056 |
+
"step": 1470
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.063436737539212,
|
| 1060 |
+
"grad_norm": 11.112920761108398,
|
| 1061 |
+
"learning_rate": 1.7338501291989665e-05,
|
| 1062 |
+
"loss": 2.5778,
|
| 1063 |
+
"step": 1480
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.0773788776577207,
|
| 1067 |
+
"grad_norm": 13.829360008239746,
|
| 1068 |
+
"learning_rate": 1.7080103359173128e-05,
|
| 1069 |
+
"loss": 2.2665,
|
| 1070 |
+
"step": 1490
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.0913210177762287,
|
| 1074 |
+
"grad_norm": 12.493084907531738,
|
| 1075 |
+
"learning_rate": 1.682170542635659e-05,
|
| 1076 |
+
"loss": 2.5002,
|
| 1077 |
+
"step": 1500
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.1052631578947367,
|
| 1081 |
+
"grad_norm": 11.931866645812988,
|
| 1082 |
+
"learning_rate": 1.6563307493540054e-05,
|
| 1083 |
+
"loss": 2.3208,
|
| 1084 |
+
"step": 1510
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.119205298013245,
|
| 1088 |
+
"grad_norm": 10.55782413482666,
|
| 1089 |
+
"learning_rate": 1.6304909560723514e-05,
|
| 1090 |
+
"loss": 2.2226,
|
| 1091 |
+
"step": 1520
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.1331474381317532,
|
| 1095 |
+
"grad_norm": 14.521842002868652,
|
| 1096 |
+
"learning_rate": 1.6046511627906977e-05,
|
| 1097 |
+
"loss": 2.1835,
|
| 1098 |
+
"step": 1530
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.1470895782502613,
|
| 1102 |
+
"grad_norm": 12.054328918457031,
|
| 1103 |
+
"learning_rate": 1.578811369509044e-05,
|
| 1104 |
+
"loss": 2.3775,
|
| 1105 |
+
"step": 1540
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.1610317183687697,
|
| 1109 |
+
"grad_norm": 15.088787078857422,
|
| 1110 |
+
"learning_rate": 1.5529715762273903e-05,
|
| 1111 |
+
"loss": 2.4388,
|
| 1112 |
+
"step": 1550
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.1749738584872778,
|
| 1116 |
+
"grad_norm": 13.146328926086426,
|
| 1117 |
+
"learning_rate": 1.5271317829457366e-05,
|
| 1118 |
+
"loss": 2.5277,
|
| 1119 |
+
"step": 1560
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.188915998605786,
|
| 1123 |
+
"grad_norm": 12.280598640441895,
|
| 1124 |
+
"learning_rate": 1.5012919896640828e-05,
|
| 1125 |
+
"loss": 2.3422,
|
| 1126 |
+
"step": 1570
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.2028581387242943,
|
| 1130 |
+
"grad_norm": 12.004185676574707,
|
| 1131 |
+
"learning_rate": 1.475452196382429e-05,
|
| 1132 |
+
"loss": 2.4908,
|
| 1133 |
+
"step": 1580
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.2168002788428023,
|
| 1137 |
+
"grad_norm": 10.80855941772461,
|
| 1138 |
+
"learning_rate": 1.4496124031007753e-05,
|
| 1139 |
+
"loss": 2.357,
|
| 1140 |
+
"step": 1590
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.2307424189613108,
|
| 1144 |
+
"grad_norm": 11.332206726074219,
|
| 1145 |
+
"learning_rate": 1.4237726098191214e-05,
|
| 1146 |
+
"loss": 2.2628,
|
| 1147 |
+
"step": 1600
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 2.244684559079819,
|
| 1151 |
+
"grad_norm": 13.836607933044434,
|
| 1152 |
+
"learning_rate": 1.3979328165374677e-05,
|
| 1153 |
+
"loss": 2.1193,
|
| 1154 |
+
"step": 1610
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.258626699198327,
|
| 1158 |
+
"grad_norm": 11.600666999816895,
|
| 1159 |
+
"learning_rate": 1.372093023255814e-05,
|
| 1160 |
+
"loss": 2.4166,
|
| 1161 |
+
"step": 1620
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.2725688393168353,
|
| 1165 |
+
"grad_norm": 14.399880409240723,
|
| 1166 |
+
"learning_rate": 1.3462532299741603e-05,
|
| 1167 |
+
"loss": 2.2878,
|
| 1168 |
+
"step": 1630
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.2865109794353433,
|
| 1172 |
+
"grad_norm": 16.899677276611328,
|
| 1173 |
+
"learning_rate": 1.3204134366925063e-05,
|
| 1174 |
+
"loss": 2.26,
|
| 1175 |
+
"step": 1640
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.3004531195538513,
|
| 1179 |
+
"grad_norm": 12.676505088806152,
|
| 1180 |
+
"learning_rate": 1.2945736434108526e-05,
|
| 1181 |
+
"loss": 2.5939,
|
| 1182 |
+
"step": 1650
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.31439525967236,
|
| 1186 |
+
"grad_norm": 11.750486373901367,
|
| 1187 |
+
"learning_rate": 1.268733850129199e-05,
|
| 1188 |
+
"loss": 2.2246,
|
| 1189 |
+
"step": 1660
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.328337399790868,
|
| 1193 |
+
"grad_norm": 14.333374977111816,
|
| 1194 |
+
"learning_rate": 1.2428940568475453e-05,
|
| 1195 |
+
"loss": 2.4163,
|
| 1196 |
+
"step": 1670
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.3422795399093763,
|
| 1200 |
+
"grad_norm": 12.187148094177246,
|
| 1201 |
+
"learning_rate": 1.2170542635658916e-05,
|
| 1202 |
+
"loss": 2.3381,
|
| 1203 |
+
"step": 1680
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.3562216800278843,
|
| 1207 |
+
"grad_norm": 10.980198860168457,
|
| 1208 |
+
"learning_rate": 1.1912144702842377e-05,
|
| 1209 |
+
"loss": 2.3502,
|
| 1210 |
+
"step": 1690
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 2.3701638201463924,
|
| 1214 |
+
"grad_norm": 13.697635650634766,
|
| 1215 |
+
"learning_rate": 1.165374677002584e-05,
|
| 1216 |
+
"loss": 2.2907,
|
| 1217 |
+
"step": 1700
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 2.384105960264901,
|
| 1221 |
+
"grad_norm": 12.195317268371582,
|
| 1222 |
+
"learning_rate": 1.1395348837209304e-05,
|
| 1223 |
+
"loss": 2.2207,
|
| 1224 |
+
"step": 1710
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 2.398048100383409,
|
| 1228 |
+
"grad_norm": 39.64952850341797,
|
| 1229 |
+
"learning_rate": 1.1136950904392765e-05,
|
| 1230 |
+
"loss": 2.2268,
|
| 1231 |
+
"step": 1720
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 2.411990240501917,
|
| 1235 |
+
"grad_norm": 11.313281059265137,
|
| 1236 |
+
"learning_rate": 1.0878552971576228e-05,
|
| 1237 |
+
"loss": 2.2816,
|
| 1238 |
+
"step": 1730
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.4259323806204254,
|
| 1242 |
+
"grad_norm": 13.669633865356445,
|
| 1243 |
+
"learning_rate": 1.0620155038759691e-05,
|
| 1244 |
+
"loss": 2.4572,
|
| 1245 |
+
"step": 1740
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 2.4398745207389334,
|
| 1249 |
+
"grad_norm": 13.479631423950195,
|
| 1250 |
+
"learning_rate": 1.0361757105943153e-05,
|
| 1251 |
+
"loss": 2.094,
|
| 1252 |
+
"step": 1750
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 2.4538166608574414,
|
| 1256 |
+
"grad_norm": 10.151293754577637,
|
| 1257 |
+
"learning_rate": 1.0103359173126616e-05,
|
| 1258 |
+
"loss": 2.0609,
|
| 1259 |
+
"step": 1760
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 2.46775880097595,
|
| 1263 |
+
"grad_norm": 14.358535766601562,
|
| 1264 |
+
"learning_rate": 9.844961240310079e-06,
|
| 1265 |
+
"loss": 2.4335,
|
| 1266 |
+
"step": 1770
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 2.481700941094458,
|
| 1270 |
+
"grad_norm": 14.903388023376465,
|
| 1271 |
+
"learning_rate": 9.58656330749354e-06,
|
| 1272 |
+
"loss": 2.2056,
|
| 1273 |
+
"step": 1780
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 2.4956430812129664,
|
| 1277 |
+
"grad_norm": 12.196182250976562,
|
| 1278 |
+
"learning_rate": 9.328165374677004e-06,
|
| 1279 |
+
"loss": 2.358,
|
| 1280 |
+
"step": 1790
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.5095852213314744,
|
| 1284 |
+
"grad_norm": 10.977519989013672,
|
| 1285 |
+
"learning_rate": 9.069767441860467e-06,
|
| 1286 |
+
"loss": 2.1615,
|
| 1287 |
+
"step": 1800
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 2.5235273614499825,
|
| 1291 |
+
"grad_norm": 10.606942176818848,
|
| 1292 |
+
"learning_rate": 8.811369509043928e-06,
|
| 1293 |
+
"loss": 2.1802,
|
| 1294 |
+
"step": 1810
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 2.5374695015684905,
|
| 1298 |
+
"grad_norm": 11.948677062988281,
|
| 1299 |
+
"learning_rate": 8.552971576227391e-06,
|
| 1300 |
+
"loss": 2.3221,
|
| 1301 |
+
"step": 1820
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 2.551411641686999,
|
| 1305 |
+
"grad_norm": 8.935708999633789,
|
| 1306 |
+
"learning_rate": 8.294573643410853e-06,
|
| 1307 |
+
"loss": 2.0471,
|
| 1308 |
+
"step": 1830
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 2.565353781805507,
|
| 1312 |
+
"grad_norm": 11.44541072845459,
|
| 1313 |
+
"learning_rate": 8.036175710594316e-06,
|
| 1314 |
+
"loss": 2.1508,
|
| 1315 |
+
"step": 1840
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 2.5792959219240155,
|
| 1319 |
+
"grad_norm": 12.39538288116455,
|
| 1320 |
+
"learning_rate": 7.777777777777777e-06,
|
| 1321 |
+
"loss": 2.253,
|
| 1322 |
+
"step": 1850
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.5932380620425235,
|
| 1326 |
+
"grad_norm": 16.37424087524414,
|
| 1327 |
+
"learning_rate": 7.5193798449612415e-06,
|
| 1328 |
+
"loss": 2.2232,
|
| 1329 |
+
"step": 1860
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 2.6071802021610315,
|
| 1333 |
+
"grad_norm": 11.820206642150879,
|
| 1334 |
+
"learning_rate": 7.260981912144703e-06,
|
| 1335 |
+
"loss": 2.4425,
|
| 1336 |
+
"step": 1870
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 2.62112234227954,
|
| 1340 |
+
"grad_norm": 12.931533813476562,
|
| 1341 |
+
"learning_rate": 7.002583979328166e-06,
|
| 1342 |
+
"loss": 2.2895,
|
| 1343 |
+
"step": 1880
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 2.635064482398048,
|
| 1347 |
+
"grad_norm": 13.43673038482666,
|
| 1348 |
+
"learning_rate": 6.744186046511629e-06,
|
| 1349 |
+
"loss": 2.2439,
|
| 1350 |
+
"step": 1890
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 2.6490066225165565,
|
| 1354 |
+
"grad_norm": 10.057219505310059,
|
| 1355 |
+
"learning_rate": 6.485788113695091e-06,
|
| 1356 |
+
"loss": 2.2317,
|
| 1357 |
+
"step": 1900
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 2.6629487626350645,
|
| 1361 |
+
"grad_norm": 15.320812225341797,
|
| 1362 |
+
"learning_rate": 6.227390180878554e-06,
|
| 1363 |
+
"loss": 2.3445,
|
| 1364 |
+
"step": 1910
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 2.6768909027535726,
|
| 1368 |
+
"grad_norm": 13.997590065002441,
|
| 1369 |
+
"learning_rate": 5.968992248062015e-06,
|
| 1370 |
+
"loss": 2.1295,
|
| 1371 |
+
"step": 1920
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 2.690833042872081,
|
| 1375 |
+
"grad_norm": 10.897136688232422,
|
| 1376 |
+
"learning_rate": 5.710594315245478e-06,
|
| 1377 |
+
"loss": 2.0328,
|
| 1378 |
+
"step": 1930
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 2.704775182990589,
|
| 1382 |
+
"grad_norm": 11.251324653625488,
|
| 1383 |
+
"learning_rate": 5.452196382428941e-06,
|
| 1384 |
+
"loss": 2.2964,
|
| 1385 |
+
"step": 1940
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 2.718717323109097,
|
| 1389 |
+
"grad_norm": 12.688488960266113,
|
| 1390 |
+
"learning_rate": 5.193798449612403e-06,
|
| 1391 |
+
"loss": 2.585,
|
| 1392 |
+
"step": 1950
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 2.7326594632276056,
|
| 1396 |
+
"grad_norm": 15.246435165405273,
|
| 1397 |
+
"learning_rate": 4.935400516795865e-06,
|
| 1398 |
+
"loss": 2.4457,
|
| 1399 |
+
"step": 1960
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 2.7466016033461136,
|
| 1403 |
+
"grad_norm": 11.022492408752441,
|
| 1404 |
+
"learning_rate": 4.6770025839793285e-06,
|
| 1405 |
+
"loss": 1.9957,
|
| 1406 |
+
"step": 1970
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 2.7605437434646216,
|
| 1410 |
+
"grad_norm": 15.17203426361084,
|
| 1411 |
+
"learning_rate": 4.418604651162791e-06,
|
| 1412 |
+
"loss": 2.3373,
|
| 1413 |
+
"step": 1980
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 2.77448588358313,
|
| 1417 |
+
"grad_norm": 13.542671203613281,
|
| 1418 |
+
"learning_rate": 4.160206718346253e-06,
|
| 1419 |
+
"loss": 2.392,
|
| 1420 |
+
"step": 1990
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 2.788428023701638,
|
| 1424 |
+
"grad_norm": 15.951026916503906,
|
| 1425 |
+
"learning_rate": 3.901808785529716e-06,
|
| 1426 |
+
"loss": 1.9829,
|
| 1427 |
+
"step": 2000
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 2.8023701638201466,
|
| 1431 |
+
"grad_norm": 13.995731353759766,
|
| 1432 |
+
"learning_rate": 3.6434108527131786e-06,
|
| 1433 |
+
"loss": 2.4399,
|
| 1434 |
+
"step": 2010
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 2.8163123039386546,
|
| 1438 |
+
"grad_norm": 11.442358016967773,
|
| 1439 |
+
"learning_rate": 3.385012919896641e-06,
|
| 1440 |
+
"loss": 2.2327,
|
| 1441 |
+
"step": 2020
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 2.8302544440571626,
|
| 1445 |
+
"grad_norm": 13.158218383789062,
|
| 1446 |
+
"learning_rate": 3.1266149870801036e-06,
|
| 1447 |
+
"loss": 2.1585,
|
| 1448 |
+
"step": 2030
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 2.844196584175671,
|
| 1452 |
+
"grad_norm": 14.946197509765625,
|
| 1453 |
+
"learning_rate": 2.868217054263566e-06,
|
| 1454 |
+
"loss": 2.2202,
|
| 1455 |
+
"step": 2040
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 2.858138724294179,
|
| 1459 |
+
"grad_norm": 13.085625648498535,
|
| 1460 |
+
"learning_rate": 2.6098191214470286e-06,
|
| 1461 |
+
"loss": 2.3849,
|
| 1462 |
+
"step": 2050
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 2.8720808644126876,
|
| 1466 |
+
"grad_norm": 11.152565002441406,
|
| 1467 |
+
"learning_rate": 2.351421188630491e-06,
|
| 1468 |
+
"loss": 2.3714,
|
| 1469 |
+
"step": 2060
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 2.8860230045311956,
|
| 1473 |
+
"grad_norm": 13.09467601776123,
|
| 1474 |
+
"learning_rate": 2.0930232558139536e-06,
|
| 1475 |
+
"loss": 2.1876,
|
| 1476 |
+
"step": 2070
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 2.8999651446497037,
|
| 1480 |
+
"grad_norm": 17.58858871459961,
|
| 1481 |
+
"learning_rate": 1.8346253229974162e-06,
|
| 1482 |
+
"loss": 2.2959,
|
| 1483 |
+
"step": 2080
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 2.9139072847682117,
|
| 1487 |
+
"grad_norm": 16.1025447845459,
|
| 1488 |
+
"learning_rate": 1.5762273901808785e-06,
|
| 1489 |
+
"loss": 2.3838,
|
| 1490 |
+
"step": 2090
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 2.92784942488672,
|
| 1494 |
+
"grad_norm": 13.546030044555664,
|
| 1495 |
+
"learning_rate": 1.3178294573643412e-06,
|
| 1496 |
+
"loss": 2.14,
|
| 1497 |
+
"step": 2100
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 2.941791565005228,
|
| 1501 |
+
"grad_norm": 12.100831985473633,
|
| 1502 |
+
"learning_rate": 1.0594315245478035e-06,
|
| 1503 |
+
"loss": 2.0447,
|
| 1504 |
+
"step": 2110
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 2.9557337051237367,
|
| 1508 |
+
"grad_norm": 12.555295944213867,
|
| 1509 |
+
"learning_rate": 8.010335917312662e-07,
|
| 1510 |
+
"loss": 2.4814,
|
| 1511 |
+
"step": 2120
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 2.9696758452422447,
|
| 1515 |
+
"grad_norm": 11.319598197937012,
|
| 1516 |
+
"learning_rate": 5.426356589147287e-07,
|
| 1517 |
+
"loss": 2.3802,
|
| 1518 |
+
"step": 2130
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 2.9836179853607527,
|
| 1522 |
+
"grad_norm": 10.149794578552246,
|
| 1523 |
+
"learning_rate": 2.842377260981912e-07,
|
| 1524 |
+
"loss": 2.3564,
|
| 1525 |
+
"step": 2140
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 2.997560125479261,
|
| 1529 |
+
"grad_norm": 13.360361099243164,
|
| 1530 |
+
"learning_rate": 2.5839793281653747e-08,
|
| 1531 |
+
"loss": 2.1182,
|
| 1532 |
+
"step": 2150
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 2.998954339491112,
|
| 1536 |
+
"eval_accuracy": 0.8684313725490196,
|
| 1537 |
+
"eval_loss": 0.5095656514167786,
|
| 1538 |
+
"eval_runtime": 245.9211,
|
| 1539 |
+
"eval_samples_per_second": 41.477,
|
| 1540 |
+
"eval_steps_per_second": 1.297,
|
| 1541 |
+
"step": 2151
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 2.998954339491112,
|
| 1545 |
+
"step": 2151,
|
| 1546 |
+
"total_flos": 2.1353596059581743e+19,
|
| 1547 |
+
"train_loss": 4.144423897296315,
|
| 1548 |
+
"train_runtime": 8280.2172,
|
| 1549 |
+
"train_samples_per_second": 33.26,
|
| 1550 |
+
"train_steps_per_second": 0.26
|
| 1551 |
+
}
|
| 1552 |
+
],
|
| 1553 |
+
"logging_steps": 10,
|
| 1554 |
+
"max_steps": 2151,
|
| 1555 |
+
"num_input_tokens_seen": 0,
|
| 1556 |
+
"num_train_epochs": 3,
|
| 1557 |
+
"save_steps": 500,
|
| 1558 |
+
"stateful_callbacks": {
|
| 1559 |
+
"TrainerControl": {
|
| 1560 |
+
"args": {
|
| 1561 |
+
"should_epoch_stop": false,
|
| 1562 |
+
"should_evaluate": false,
|
| 1563 |
+
"should_log": false,
|
| 1564 |
+
"should_save": true,
|
| 1565 |
+
"should_training_stop": true
|
| 1566 |
+
},
|
| 1567 |
+
"attributes": {}
|
| 1568 |
+
}
|
| 1569 |
+
},
|
| 1570 |
+
"total_flos": 2.1353596059581743e+19,
|
| 1571 |
+
"train_batch_size": 32,
|
| 1572 |
+
"trial_name": null,
|
| 1573 |
+
"trial_params": null
|
| 1574 |
+
}
|