End of training
Browse files- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2099 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.991111111111111,
|
| 3 |
+
"eval_accuracy": 0.954,
|
| 4 |
+
"eval_loss": 0.13787274062633514,
|
| 5 |
+
"eval_runtime": 1471.2362,
|
| 6 |
+
"eval_samples_per_second": 2.719,
|
| 7 |
+
"eval_steps_per_second": 0.085,
|
| 8 |
+
"total_flos": 8.940202118432686e+18,
|
| 9 |
+
"train_loss": 0.24847593615063568,
|
| 10 |
+
"train_runtime": 161511.588,
|
| 11 |
+
"train_samples_per_second": 2.229,
|
| 12 |
+
"train_steps_per_second": 0.017
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.991111111111111,
|
| 3 |
+
"eval_accuracy": 0.954,
|
| 4 |
+
"eval_loss": 0.13787274062633514,
|
| 5 |
+
"eval_runtime": 1471.2362,
|
| 6 |
+
"eval_samples_per_second": 2.719,
|
| 7 |
+
"eval_steps_per_second": 0.085
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.991111111111111,
|
| 3 |
+
"total_flos": 8.940202118432686e+18,
|
| 4 |
+
"train_loss": 0.24847593615063568,
|
| 5 |
+
"train_runtime": 161511.588,
|
| 6 |
+
"train_samples_per_second": 2.229,
|
| 7 |
+
"train_steps_per_second": 0.017
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2099 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.954,
|
| 3 |
+
"best_model_checkpoint": "swin-tiny-patch4-window7-224-swin-tiny-patch4-window7-224-finetuned-leukemia.v2.1/checkpoint-2810",
|
| 4 |
+
"epoch": 9.991111111111111,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2810,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.035555555555555556,
|
| 13 |
+
"grad_norm": 6.782552719116211,
|
| 14 |
+
"learning_rate": 1.7793594306049826e-06,
|
| 15 |
+
"loss": 0.6998,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.07111111111111111,
|
| 20 |
+
"grad_norm": 9.350823402404785,
|
| 21 |
+
"learning_rate": 3.558718861209965e-06,
|
| 22 |
+
"loss": 0.6905,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.10666666666666667,
|
| 27 |
+
"grad_norm": 10.970987319946289,
|
| 28 |
+
"learning_rate": 5.338078291814947e-06,
|
| 29 |
+
"loss": 0.6829,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.14222222222222222,
|
| 34 |
+
"grad_norm": 7.290198802947998,
|
| 35 |
+
"learning_rate": 7.11743772241993e-06,
|
| 36 |
+
"loss": 0.6713,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.17777777777777778,
|
| 41 |
+
"grad_norm": 10.42104721069336,
|
| 42 |
+
"learning_rate": 8.896797153024912e-06,
|
| 43 |
+
"loss": 0.6545,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.21333333333333335,
|
| 48 |
+
"grad_norm": 7.574017524719238,
|
| 49 |
+
"learning_rate": 1.0676156583629894e-05,
|
| 50 |
+
"loss": 0.6612,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.24888888888888888,
|
| 55 |
+
"grad_norm": 4.902874946594238,
|
| 56 |
+
"learning_rate": 1.2455516014234877e-05,
|
| 57 |
+
"loss": 0.6571,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.28444444444444444,
|
| 62 |
+
"grad_norm": 6.821291446685791,
|
| 63 |
+
"learning_rate": 1.423487544483986e-05,
|
| 64 |
+
"loss": 0.6425,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.32,
|
| 69 |
+
"grad_norm": 5.546478271484375,
|
| 70 |
+
"learning_rate": 1.601423487544484e-05,
|
| 71 |
+
"loss": 0.6333,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.35555555555555557,
|
| 76 |
+
"grad_norm": 8.954255104064941,
|
| 77 |
+
"learning_rate": 1.7793594306049825e-05,
|
| 78 |
+
"loss": 0.6294,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.39111111111111113,
|
| 83 |
+
"grad_norm": 8.712015151977539,
|
| 84 |
+
"learning_rate": 1.9572953736654805e-05,
|
| 85 |
+
"loss": 0.6262,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.4266666666666667,
|
| 90 |
+
"grad_norm": 8.461681365966797,
|
| 91 |
+
"learning_rate": 2.135231316725979e-05,
|
| 92 |
+
"loss": 0.594,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.4622222222222222,
|
| 97 |
+
"grad_norm": 15.287362098693848,
|
| 98 |
+
"learning_rate": 2.313167259786477e-05,
|
| 99 |
+
"loss": 0.5622,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.49777777777777776,
|
| 104 |
+
"grad_norm": 13.876784324645996,
|
| 105 |
+
"learning_rate": 2.4911032028469753e-05,
|
| 106 |
+
"loss": 0.5684,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.5333333333333333,
|
| 111 |
+
"grad_norm": 8.108331680297852,
|
| 112 |
+
"learning_rate": 2.669039145907473e-05,
|
| 113 |
+
"loss": 0.5268,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.5688888888888889,
|
| 118 |
+
"grad_norm": 10.684402465820312,
|
| 119 |
+
"learning_rate": 2.846975088967972e-05,
|
| 120 |
+
"loss": 0.5397,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.6044444444444445,
|
| 125 |
+
"grad_norm": 6.989006042480469,
|
| 126 |
+
"learning_rate": 3.02491103202847e-05,
|
| 127 |
+
"loss": 0.5337,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.64,
|
| 132 |
+
"grad_norm": 25.538442611694336,
|
| 133 |
+
"learning_rate": 3.202846975088968e-05,
|
| 134 |
+
"loss": 0.5013,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6755555555555556,
|
| 139 |
+
"grad_norm": 16.017742156982422,
|
| 140 |
+
"learning_rate": 3.380782918149467e-05,
|
| 141 |
+
"loss": 0.4775,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.7111111111111111,
|
| 146 |
+
"grad_norm": 8.600865364074707,
|
| 147 |
+
"learning_rate": 3.558718861209965e-05,
|
| 148 |
+
"loss": 0.5052,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.7466666666666667,
|
| 153 |
+
"grad_norm": 10.005355834960938,
|
| 154 |
+
"learning_rate": 3.736654804270463e-05,
|
| 155 |
+
"loss": 0.5027,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7822222222222223,
|
| 160 |
+
"grad_norm": 9.54813003540039,
|
| 161 |
+
"learning_rate": 3.914590747330961e-05,
|
| 162 |
+
"loss": 0.4906,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.8177777777777778,
|
| 167 |
+
"grad_norm": 26.83182144165039,
|
| 168 |
+
"learning_rate": 4.09252669039146e-05,
|
| 169 |
+
"loss": 0.4725,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.8533333333333334,
|
| 174 |
+
"grad_norm": 15.171409606933594,
|
| 175 |
+
"learning_rate": 4.270462633451958e-05,
|
| 176 |
+
"loss": 0.4698,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8888888888888888,
|
| 181 |
+
"grad_norm": 13.108918190002441,
|
| 182 |
+
"learning_rate": 4.448398576512456e-05,
|
| 183 |
+
"loss": 0.4294,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.9244444444444444,
|
| 188 |
+
"grad_norm": 9.245574951171875,
|
| 189 |
+
"learning_rate": 4.626334519572954e-05,
|
| 190 |
+
"loss": 0.4691,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.96,
|
| 195 |
+
"grad_norm": 24.02051544189453,
|
| 196 |
+
"learning_rate": 4.8042704626334526e-05,
|
| 197 |
+
"loss": 0.4528,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.9955555555555555,
|
| 202 |
+
"grad_norm": 15.161036491394043,
|
| 203 |
+
"learning_rate": 4.9822064056939506e-05,
|
| 204 |
+
"loss": 0.4215,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9991111111111111,
|
| 209 |
+
"eval_accuracy": 0.82925,
|
| 210 |
+
"eval_loss": 0.3879869878292084,
|
| 211 |
+
"eval_runtime": 1523.0822,
|
| 212 |
+
"eval_samples_per_second": 2.626,
|
| 213 |
+
"eval_steps_per_second": 0.082,
|
| 214 |
+
"step": 281
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 1.031111111111111,
|
| 218 |
+
"grad_norm": 14.298422813415527,
|
| 219 |
+
"learning_rate": 4.9822064056939506e-05,
|
| 220 |
+
"loss": 0.4261,
|
| 221 |
+
"step": 290
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.0666666666666667,
|
| 225 |
+
"grad_norm": 25.5822696685791,
|
| 226 |
+
"learning_rate": 4.962435745353895e-05,
|
| 227 |
+
"loss": 0.4725,
|
| 228 |
+
"step": 300
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.1022222222222222,
|
| 232 |
+
"grad_norm": 6.261577606201172,
|
| 233 |
+
"learning_rate": 4.9426650850138396e-05,
|
| 234 |
+
"loss": 0.3882,
|
| 235 |
+
"step": 310
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.1377777777777778,
|
| 239 |
+
"grad_norm": 13.508868217468262,
|
| 240 |
+
"learning_rate": 4.9228944246737844e-05,
|
| 241 |
+
"loss": 0.4532,
|
| 242 |
+
"step": 320
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.1733333333333333,
|
| 246 |
+
"grad_norm": 15.246574401855469,
|
| 247 |
+
"learning_rate": 4.903123764333729e-05,
|
| 248 |
+
"loss": 0.4165,
|
| 249 |
+
"step": 330
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.208888888888889,
|
| 253 |
+
"grad_norm": 8.761860847473145,
|
| 254 |
+
"learning_rate": 4.8833531039936733e-05,
|
| 255 |
+
"loss": 0.3892,
|
| 256 |
+
"step": 340
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.2444444444444445,
|
| 260 |
+
"grad_norm": 14.988066673278809,
|
| 261 |
+
"learning_rate": 4.863582443653618e-05,
|
| 262 |
+
"loss": 0.4098,
|
| 263 |
+
"step": 350
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.28,
|
| 267 |
+
"grad_norm": 25.65663719177246,
|
| 268 |
+
"learning_rate": 4.843811783313563e-05,
|
| 269 |
+
"loss": 0.4069,
|
| 270 |
+
"step": 360
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.3155555555555556,
|
| 274 |
+
"grad_norm": 8.681095123291016,
|
| 275 |
+
"learning_rate": 4.824041122973508e-05,
|
| 276 |
+
"loss": 0.3701,
|
| 277 |
+
"step": 370
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.3511111111111112,
|
| 281 |
+
"grad_norm": 11.527020454406738,
|
| 282 |
+
"learning_rate": 4.8042704626334526e-05,
|
| 283 |
+
"loss": 0.3814,
|
| 284 |
+
"step": 380
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.3866666666666667,
|
| 288 |
+
"grad_norm": 20.192686080932617,
|
| 289 |
+
"learning_rate": 4.784499802293397e-05,
|
| 290 |
+
"loss": 0.3811,
|
| 291 |
+
"step": 390
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.4222222222222223,
|
| 295 |
+
"grad_norm": 31.64283561706543,
|
| 296 |
+
"learning_rate": 4.7647291419533415e-05,
|
| 297 |
+
"loss": 0.3792,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.4577777777777778,
|
| 302 |
+
"grad_norm": 26.02726173400879,
|
| 303 |
+
"learning_rate": 4.7449584816132864e-05,
|
| 304 |
+
"loss": 0.3975,
|
| 305 |
+
"step": 410
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.4933333333333334,
|
| 309 |
+
"grad_norm": 19.982933044433594,
|
| 310 |
+
"learning_rate": 4.725187821273231e-05,
|
| 311 |
+
"loss": 0.3901,
|
| 312 |
+
"step": 420
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.528888888888889,
|
| 316 |
+
"grad_norm": 36.1695442199707,
|
| 317 |
+
"learning_rate": 4.705417160933175e-05,
|
| 318 |
+
"loss": 0.4053,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.5644444444444443,
|
| 323 |
+
"grad_norm": 16.495031356811523,
|
| 324 |
+
"learning_rate": 4.68564650059312e-05,
|
| 325 |
+
"loss": 0.4005,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.6,
|
| 330 |
+
"grad_norm": 31.459928512573242,
|
| 331 |
+
"learning_rate": 4.665875840253064e-05,
|
| 332 |
+
"loss": 0.3576,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.6355555555555554,
|
| 337 |
+
"grad_norm": 6.371111869812012,
|
| 338 |
+
"learning_rate": 4.64610517991301e-05,
|
| 339 |
+
"loss": 0.3354,
|
| 340 |
+
"step": 460
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.6711111111111112,
|
| 344 |
+
"grad_norm": 7.015986919403076,
|
| 345 |
+
"learning_rate": 4.626334519572954e-05,
|
| 346 |
+
"loss": 0.3222,
|
| 347 |
+
"step": 470
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.7066666666666666,
|
| 351 |
+
"grad_norm": 30.189485549926758,
|
| 352 |
+
"learning_rate": 4.606563859232899e-05,
|
| 353 |
+
"loss": 0.3579,
|
| 354 |
+
"step": 480
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.7422222222222223,
|
| 358 |
+
"grad_norm": 15.577569007873535,
|
| 359 |
+
"learning_rate": 4.586793198892843e-05,
|
| 360 |
+
"loss": 0.3805,
|
| 361 |
+
"step": 490
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.7777777777777777,
|
| 365 |
+
"grad_norm": 30.020050048828125,
|
| 366 |
+
"learning_rate": 4.5670225385527876e-05,
|
| 367 |
+
"loss": 0.4028,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.8133333333333335,
|
| 372 |
+
"grad_norm": 24.862821578979492,
|
| 373 |
+
"learning_rate": 4.5472518782127324e-05,
|
| 374 |
+
"loss": 0.3599,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.8488888888888888,
|
| 379 |
+
"grad_norm": 36.88605880737305,
|
| 380 |
+
"learning_rate": 4.527481217872677e-05,
|
| 381 |
+
"loss": 0.3329,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.8844444444444446,
|
| 386 |
+
"grad_norm": 32.23126983642578,
|
| 387 |
+
"learning_rate": 4.5077105575326214e-05,
|
| 388 |
+
"loss": 0.3715,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.92,
|
| 393 |
+
"grad_norm": 9.027021408081055,
|
| 394 |
+
"learning_rate": 4.487939897192566e-05,
|
| 395 |
+
"loss": 0.3678,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.9555555555555557,
|
| 400 |
+
"grad_norm": 16.631864547729492,
|
| 401 |
+
"learning_rate": 4.468169236852511e-05,
|
| 402 |
+
"loss": 0.3346,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.991111111111111,
|
| 407 |
+
"grad_norm": 19.838829040527344,
|
| 408 |
+
"learning_rate": 4.448398576512456e-05,
|
| 409 |
+
"loss": 0.3137,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.9982222222222221,
|
| 414 |
+
"eval_accuracy": 0.87875,
|
| 415 |
+
"eval_loss": 0.2897728383541107,
|
| 416 |
+
"eval_runtime": 1451.863,
|
| 417 |
+
"eval_samples_per_second": 2.755,
|
| 418 |
+
"eval_steps_per_second": 0.086,
|
| 419 |
+
"step": 562
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 2.026666666666667,
|
| 423 |
+
"grad_norm": 20.36936378479004,
|
| 424 |
+
"learning_rate": 4.4286279161724006e-05,
|
| 425 |
+
"loss": 0.3102,
|
| 426 |
+
"step": 570
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 2.062222222222222,
|
| 430 |
+
"grad_norm": 8.112001419067383,
|
| 431 |
+
"learning_rate": 4.408857255832345e-05,
|
| 432 |
+
"loss": 0.3337,
|
| 433 |
+
"step": 580
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 2.097777777777778,
|
| 437 |
+
"grad_norm": 27.735227584838867,
|
| 438 |
+
"learning_rate": 4.3890865954922896e-05,
|
| 439 |
+
"loss": 0.3596,
|
| 440 |
+
"step": 590
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.1333333333333333,
|
| 444 |
+
"grad_norm": 12.126736640930176,
|
| 445 |
+
"learning_rate": 4.3693159351522344e-05,
|
| 446 |
+
"loss": 0.4431,
|
| 447 |
+
"step": 600
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.168888888888889,
|
| 451 |
+
"grad_norm": 36.314876556396484,
|
| 452 |
+
"learning_rate": 4.349545274812179e-05,
|
| 453 |
+
"loss": 0.3472,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.2044444444444444,
|
| 458 |
+
"grad_norm": 8.522990226745605,
|
| 459 |
+
"learning_rate": 4.3297746144721233e-05,
|
| 460 |
+
"loss": 0.3179,
|
| 461 |
+
"step": 620
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.24,
|
| 465 |
+
"grad_norm": 13.799488067626953,
|
| 466 |
+
"learning_rate": 4.310003954132068e-05,
|
| 467 |
+
"loss": 0.3166,
|
| 468 |
+
"step": 630
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 2.2755555555555556,
|
| 472 |
+
"grad_norm": 10.06682300567627,
|
| 473 |
+
"learning_rate": 4.290233293792013e-05,
|
| 474 |
+
"loss": 0.3291,
|
| 475 |
+
"step": 640
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 2.311111111111111,
|
| 479 |
+
"grad_norm": 45.33847427368164,
|
| 480 |
+
"learning_rate": 4.270462633451958e-05,
|
| 481 |
+
"loss": 0.3048,
|
| 482 |
+
"step": 650
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.3466666666666667,
|
| 486 |
+
"grad_norm": 19.919029235839844,
|
| 487 |
+
"learning_rate": 4.250691973111902e-05,
|
| 488 |
+
"loss": 0.3771,
|
| 489 |
+
"step": 660
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.3822222222222225,
|
| 493 |
+
"grad_norm": 13.33297348022461,
|
| 494 |
+
"learning_rate": 4.230921312771847e-05,
|
| 495 |
+
"loss": 0.2913,
|
| 496 |
+
"step": 670
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.417777777777778,
|
| 500 |
+
"grad_norm": 10.342403411865234,
|
| 501 |
+
"learning_rate": 4.211150652431791e-05,
|
| 502 |
+
"loss": 0.329,
|
| 503 |
+
"step": 680
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.453333333333333,
|
| 507 |
+
"grad_norm": 14.087092399597168,
|
| 508 |
+
"learning_rate": 4.1913799920917364e-05,
|
| 509 |
+
"loss": 0.2843,
|
| 510 |
+
"step": 690
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.488888888888889,
|
| 514 |
+
"grad_norm": 20.755348205566406,
|
| 515 |
+
"learning_rate": 4.1716093317516805e-05,
|
| 516 |
+
"loss": 0.3078,
|
| 517 |
+
"step": 700
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.5244444444444447,
|
| 521 |
+
"grad_norm": 23.535104751586914,
|
| 522 |
+
"learning_rate": 4.151838671411625e-05,
|
| 523 |
+
"loss": 0.3001,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.56,
|
| 528 |
+
"grad_norm": 32.4664421081543,
|
| 529 |
+
"learning_rate": 4.13206801107157e-05,
|
| 530 |
+
"loss": 0.3346,
|
| 531 |
+
"step": 720
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.5955555555555554,
|
| 535 |
+
"grad_norm": 12.650872230529785,
|
| 536 |
+
"learning_rate": 4.112297350731515e-05,
|
| 537 |
+
"loss": 0.2884,
|
| 538 |
+
"step": 730
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.631111111111111,
|
| 542 |
+
"grad_norm": 11.952192306518555,
|
| 543 |
+
"learning_rate": 4.09252669039146e-05,
|
| 544 |
+
"loss": 0.2912,
|
| 545 |
+
"step": 740
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.6666666666666665,
|
| 549 |
+
"grad_norm": 9.068989753723145,
|
| 550 |
+
"learning_rate": 4.072756030051404e-05,
|
| 551 |
+
"loss": 0.2664,
|
| 552 |
+
"step": 750
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.7022222222222223,
|
| 556 |
+
"grad_norm": 47.9985466003418,
|
| 557 |
+
"learning_rate": 4.052985369711349e-05,
|
| 558 |
+
"loss": 0.2881,
|
| 559 |
+
"step": 760
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.7377777777777776,
|
| 563 |
+
"grad_norm": 23.085819244384766,
|
| 564 |
+
"learning_rate": 4.033214709371293e-05,
|
| 565 |
+
"loss": 0.3261,
|
| 566 |
+
"step": 770
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.7733333333333334,
|
| 570 |
+
"grad_norm": 9.50033950805664,
|
| 571 |
+
"learning_rate": 4.013444049031238e-05,
|
| 572 |
+
"loss": 0.3105,
|
| 573 |
+
"step": 780
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.8088888888888888,
|
| 577 |
+
"grad_norm": 13.149019241333008,
|
| 578 |
+
"learning_rate": 3.9936733886911825e-05,
|
| 579 |
+
"loss": 0.3034,
|
| 580 |
+
"step": 790
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.8444444444444446,
|
| 584 |
+
"grad_norm": 48.091148376464844,
|
| 585 |
+
"learning_rate": 3.973902728351127e-05,
|
| 586 |
+
"loss": 0.2799,
|
| 587 |
+
"step": 800
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.88,
|
| 591 |
+
"grad_norm": 11.447802543640137,
|
| 592 |
+
"learning_rate": 3.9541320680110714e-05,
|
| 593 |
+
"loss": 0.2979,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.9155555555555557,
|
| 598 |
+
"grad_norm": 22.164430618286133,
|
| 599 |
+
"learning_rate": 3.934361407671016e-05,
|
| 600 |
+
"loss": 0.2732,
|
| 601 |
+
"step": 820
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.951111111111111,
|
| 605 |
+
"grad_norm": 9.461747169494629,
|
| 606 |
+
"learning_rate": 3.914590747330961e-05,
|
| 607 |
+
"loss": 0.284,
|
| 608 |
+
"step": 830
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.986666666666667,
|
| 612 |
+
"grad_norm": 9.027481079101562,
|
| 613 |
+
"learning_rate": 3.894820086990906e-05,
|
| 614 |
+
"loss": 0.2631,
|
| 615 |
+
"step": 840
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.997333333333333,
|
| 619 |
+
"eval_accuracy": 0.907,
|
| 620 |
+
"eval_loss": 0.2382480353116989,
|
| 621 |
+
"eval_runtime": 1544.9741,
|
| 622 |
+
"eval_samples_per_second": 2.589,
|
| 623 |
+
"eval_steps_per_second": 0.081,
|
| 624 |
+
"step": 843
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 3.022222222222222,
|
| 628 |
+
"grad_norm": 21.90301513671875,
|
| 629 |
+
"learning_rate": 3.87504942665085e-05,
|
| 630 |
+
"loss": 0.2526,
|
| 631 |
+
"step": 850
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 3.057777777777778,
|
| 635 |
+
"grad_norm": 17.74726104736328,
|
| 636 |
+
"learning_rate": 3.855278766310795e-05,
|
| 637 |
+
"loss": 0.2289,
|
| 638 |
+
"step": 860
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 3.0933333333333333,
|
| 642 |
+
"grad_norm": 5.85569953918457,
|
| 643 |
+
"learning_rate": 3.8355081059707396e-05,
|
| 644 |
+
"loss": 0.3025,
|
| 645 |
+
"step": 870
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 3.128888888888889,
|
| 649 |
+
"grad_norm": 24.68912124633789,
|
| 650 |
+
"learning_rate": 3.8157374456306844e-05,
|
| 651 |
+
"loss": 0.2638,
|
| 652 |
+
"step": 880
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 3.1644444444444444,
|
| 656 |
+
"grad_norm": 10.496946334838867,
|
| 657 |
+
"learning_rate": 3.7959667852906285e-05,
|
| 658 |
+
"loss": 0.2397,
|
| 659 |
+
"step": 890
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 3.2,
|
| 663 |
+
"grad_norm": 13.301694869995117,
|
| 664 |
+
"learning_rate": 3.7761961249505734e-05,
|
| 665 |
+
"loss": 0.2785,
|
| 666 |
+
"step": 900
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 3.2355555555555555,
|
| 670 |
+
"grad_norm": 21.722679138183594,
|
| 671 |
+
"learning_rate": 3.756425464610518e-05,
|
| 672 |
+
"loss": 0.2962,
|
| 673 |
+
"step": 910
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 3.2711111111111113,
|
| 677 |
+
"grad_norm": 20.864559173583984,
|
| 678 |
+
"learning_rate": 3.736654804270463e-05,
|
| 679 |
+
"loss": 0.2644,
|
| 680 |
+
"step": 920
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 3.3066666666666666,
|
| 684 |
+
"grad_norm": 7.889869689941406,
|
| 685 |
+
"learning_rate": 3.716884143930408e-05,
|
| 686 |
+
"loss": 0.2887,
|
| 687 |
+
"step": 930
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.3422222222222224,
|
| 691 |
+
"grad_norm": 9.892852783203125,
|
| 692 |
+
"learning_rate": 3.697113483590352e-05,
|
| 693 |
+
"loss": 0.2638,
|
| 694 |
+
"step": 940
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.3777777777777778,
|
| 698 |
+
"grad_norm": 9.880372047424316,
|
| 699 |
+
"learning_rate": 3.677342823250297e-05,
|
| 700 |
+
"loss": 0.2559,
|
| 701 |
+
"step": 950
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.413333333333333,
|
| 705 |
+
"grad_norm": 17.303327560424805,
|
| 706 |
+
"learning_rate": 3.6575721629102416e-05,
|
| 707 |
+
"loss": 0.246,
|
| 708 |
+
"step": 960
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.448888888888889,
|
| 712 |
+
"grad_norm": 25.565025329589844,
|
| 713 |
+
"learning_rate": 3.6378015025701864e-05,
|
| 714 |
+
"loss": 0.2475,
|
| 715 |
+
"step": 970
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.4844444444444447,
|
| 719 |
+
"grad_norm": 18.166460037231445,
|
| 720 |
+
"learning_rate": 3.6180308422301305e-05,
|
| 721 |
+
"loss": 0.2515,
|
| 722 |
+
"step": 980
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.52,
|
| 726 |
+
"grad_norm": 12.403401374816895,
|
| 727 |
+
"learning_rate": 3.598260181890075e-05,
|
| 728 |
+
"loss": 0.2545,
|
| 729 |
+
"step": 990
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.5555555555555554,
|
| 733 |
+
"grad_norm": 41.32758712768555,
|
| 734 |
+
"learning_rate": 3.5784895215500194e-05,
|
| 735 |
+
"loss": 0.2446,
|
| 736 |
+
"step": 1000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.591111111111111,
|
| 740 |
+
"grad_norm": 20.999855041503906,
|
| 741 |
+
"learning_rate": 3.558718861209965e-05,
|
| 742 |
+
"loss": 0.2752,
|
| 743 |
+
"step": 1010
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.626666666666667,
|
| 747 |
+
"grad_norm": 10.489103317260742,
|
| 748 |
+
"learning_rate": 3.538948200869909e-05,
|
| 749 |
+
"loss": 0.2594,
|
| 750 |
+
"step": 1020
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.6622222222222223,
|
| 754 |
+
"grad_norm": 15.303348541259766,
|
| 755 |
+
"learning_rate": 3.519177540529854e-05,
|
| 756 |
+
"loss": 0.2784,
|
| 757 |
+
"step": 1030
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.6977777777777776,
|
| 761 |
+
"grad_norm": 28.007869720458984,
|
| 762 |
+
"learning_rate": 3.499406880189798e-05,
|
| 763 |
+
"loss": 0.249,
|
| 764 |
+
"step": 1040
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.7333333333333334,
|
| 768 |
+
"grad_norm": 11.705853462219238,
|
| 769 |
+
"learning_rate": 3.4796362198497435e-05,
|
| 770 |
+
"loss": 0.2441,
|
| 771 |
+
"step": 1050
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.7688888888888887,
|
| 775 |
+
"grad_norm": 10.978654861450195,
|
| 776 |
+
"learning_rate": 3.4598655595096876e-05,
|
| 777 |
+
"loss": 0.2476,
|
| 778 |
+
"step": 1060
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.8044444444444445,
|
| 782 |
+
"grad_norm": 21.1198787689209,
|
| 783 |
+
"learning_rate": 3.4400948991696325e-05,
|
| 784 |
+
"loss": 0.2498,
|
| 785 |
+
"step": 1070
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.84,
|
| 789 |
+
"grad_norm": 15.153464317321777,
|
| 790 |
+
"learning_rate": 3.420324238829577e-05,
|
| 791 |
+
"loss": 0.2599,
|
| 792 |
+
"step": 1080
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.8755555555555556,
|
| 796 |
+
"grad_norm": 13.55528450012207,
|
| 797 |
+
"learning_rate": 3.4005535784895214e-05,
|
| 798 |
+
"loss": 0.2381,
|
| 799 |
+
"step": 1090
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.911111111111111,
|
| 803 |
+
"grad_norm": 8.704148292541504,
|
| 804 |
+
"learning_rate": 3.380782918149467e-05,
|
| 805 |
+
"loss": 0.2108,
|
| 806 |
+
"step": 1100
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.9466666666666668,
|
| 810 |
+
"grad_norm": 9.983017921447754,
|
| 811 |
+
"learning_rate": 3.361012257809411e-05,
|
| 812 |
+
"loss": 0.2399,
|
| 813 |
+
"step": 1110
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.982222222222222,
|
| 817 |
+
"grad_norm": 7.5024003982543945,
|
| 818 |
+
"learning_rate": 3.341241597469356e-05,
|
| 819 |
+
"loss": 0.2338,
|
| 820 |
+
"step": 1120
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 4.0,
|
| 824 |
+
"eval_accuracy": 0.8575,
|
| 825 |
+
"eval_loss": 0.40904077887535095,
|
| 826 |
+
"eval_runtime": 1463.697,
|
| 827 |
+
"eval_samples_per_second": 2.733,
|
| 828 |
+
"eval_steps_per_second": 0.085,
|
| 829 |
+
"step": 1125
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 4.017777777777778,
|
| 833 |
+
"grad_norm": 16.46148109436035,
|
| 834 |
+
"learning_rate": 3.3214709371293e-05,
|
| 835 |
+
"loss": 0.2557,
|
| 836 |
+
"step": 1130
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.053333333333334,
|
| 840 |
+
"grad_norm": 7.5980939865112305,
|
| 841 |
+
"learning_rate": 3.301700276789245e-05,
|
| 842 |
+
"loss": 0.2543,
|
| 843 |
+
"step": 1140
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 4.088888888888889,
|
| 847 |
+
"grad_norm": 16.72043228149414,
|
| 848 |
+
"learning_rate": 3.2819296164491896e-05,
|
| 849 |
+
"loss": 0.2292,
|
| 850 |
+
"step": 1150
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 4.124444444444444,
|
| 854 |
+
"grad_norm": 26.61065101623535,
|
| 855 |
+
"learning_rate": 3.2621589561091344e-05,
|
| 856 |
+
"loss": 0.2954,
|
| 857 |
+
"step": 1160
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 4.16,
|
| 861 |
+
"grad_norm": 10.202407836914062,
|
| 862 |
+
"learning_rate": 3.2423882957690785e-05,
|
| 863 |
+
"loss": 0.2527,
|
| 864 |
+
"step": 1170
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 4.195555555555556,
|
| 868 |
+
"grad_norm": 25.287439346313477,
|
| 869 |
+
"learning_rate": 3.2226176354290234e-05,
|
| 870 |
+
"loss": 0.2308,
|
| 871 |
+
"step": 1180
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 4.231111111111111,
|
| 875 |
+
"grad_norm": 10.251760482788086,
|
| 876 |
+
"learning_rate": 3.202846975088968e-05,
|
| 877 |
+
"loss": 0.2262,
|
| 878 |
+
"step": 1190
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 4.266666666666667,
|
| 882 |
+
"grad_norm": 11.580595016479492,
|
| 883 |
+
"learning_rate": 3.183076314748913e-05,
|
| 884 |
+
"loss": 0.2135,
|
| 885 |
+
"step": 1200
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 4.302222222222222,
|
| 889 |
+
"grad_norm": 10.563607215881348,
|
| 890 |
+
"learning_rate": 3.163305654408857e-05,
|
| 891 |
+
"loss": 0.2226,
|
| 892 |
+
"step": 1210
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 4.337777777777778,
|
| 896 |
+
"grad_norm": 15.37928295135498,
|
| 897 |
+
"learning_rate": 3.143534994068802e-05,
|
| 898 |
+
"loss": 0.2237,
|
| 899 |
+
"step": 1220
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 4.373333333333333,
|
| 903 |
+
"grad_norm": 9.385284423828125,
|
| 904 |
+
"learning_rate": 3.123764333728747e-05,
|
| 905 |
+
"loss": 0.2696,
|
| 906 |
+
"step": 1230
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 4.408888888888889,
|
| 910 |
+
"grad_norm": 5.899435520172119,
|
| 911 |
+
"learning_rate": 3.1039936733886916e-05,
|
| 912 |
+
"loss": 0.2249,
|
| 913 |
+
"step": 1240
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 4.444444444444445,
|
| 917 |
+
"grad_norm": 9.673226356506348,
|
| 918 |
+
"learning_rate": 3.0842230130486364e-05,
|
| 919 |
+
"loss": 0.2159,
|
| 920 |
+
"step": 1250
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 4.48,
|
| 924 |
+
"grad_norm": 13.770596504211426,
|
| 925 |
+
"learning_rate": 3.0644523527085805e-05,
|
| 926 |
+
"loss": 0.2372,
|
| 927 |
+
"step": 1260
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 4.515555555555555,
|
| 931 |
+
"grad_norm": 7.797916412353516,
|
| 932 |
+
"learning_rate": 3.044681692368525e-05,
|
| 933 |
+
"loss": 0.206,
|
| 934 |
+
"step": 1270
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 4.551111111111111,
|
| 938 |
+
"grad_norm": 14.872694969177246,
|
| 939 |
+
"learning_rate": 3.02491103202847e-05,
|
| 940 |
+
"loss": 0.2215,
|
| 941 |
+
"step": 1280
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 4.586666666666667,
|
| 945 |
+
"grad_norm": 16.459089279174805,
|
| 946 |
+
"learning_rate": 3.0051403716884146e-05,
|
| 947 |
+
"loss": 0.1951,
|
| 948 |
+
"step": 1290
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 4.622222222222222,
|
| 952 |
+
"grad_norm": 19.72576141357422,
|
| 953 |
+
"learning_rate": 2.985369711348359e-05,
|
| 954 |
+
"loss": 0.2248,
|
| 955 |
+
"step": 1300
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 4.657777777777778,
|
| 959 |
+
"grad_norm": 11.154007911682129,
|
| 960 |
+
"learning_rate": 2.9655990510083035e-05,
|
| 961 |
+
"loss": 0.2112,
|
| 962 |
+
"step": 1310
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 4.693333333333333,
|
| 966 |
+
"grad_norm": 12.139333724975586,
|
| 967 |
+
"learning_rate": 2.9458283906682484e-05,
|
| 968 |
+
"loss": 0.1851,
|
| 969 |
+
"step": 1320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 4.728888888888889,
|
| 973 |
+
"grad_norm": 20.61275863647461,
|
| 974 |
+
"learning_rate": 2.9260577303281932e-05,
|
| 975 |
+
"loss": 0.2017,
|
| 976 |
+
"step": 1330
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 4.764444444444445,
|
| 980 |
+
"grad_norm": 8.542513847351074,
|
| 981 |
+
"learning_rate": 2.906287069988138e-05,
|
| 982 |
+
"loss": 0.1957,
|
| 983 |
+
"step": 1340
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 4.8,
|
| 987 |
+
"grad_norm": 19.356861114501953,
|
| 988 |
+
"learning_rate": 2.8865164096480825e-05,
|
| 989 |
+
"loss": 0.2341,
|
| 990 |
+
"step": 1350
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 4.835555555555556,
|
| 994 |
+
"grad_norm": 14.768260955810547,
|
| 995 |
+
"learning_rate": 2.866745749308027e-05,
|
| 996 |
+
"loss": 0.1954,
|
| 997 |
+
"step": 1360
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 4.871111111111111,
|
| 1001 |
+
"grad_norm": 9.00818157196045,
|
| 1002 |
+
"learning_rate": 2.846975088967972e-05,
|
| 1003 |
+
"loss": 0.2095,
|
| 1004 |
+
"step": 1370
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 4.906666666666666,
|
| 1008 |
+
"grad_norm": 27.53821563720703,
|
| 1009 |
+
"learning_rate": 2.8272044286279166e-05,
|
| 1010 |
+
"loss": 0.2237,
|
| 1011 |
+
"step": 1380
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 4.942222222222222,
|
| 1015 |
+
"grad_norm": 8.379549980163574,
|
| 1016 |
+
"learning_rate": 2.807433768287861e-05,
|
| 1017 |
+
"loss": 0.2137,
|
| 1018 |
+
"step": 1390
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 4.977777777777778,
|
| 1022 |
+
"grad_norm": 14.70921516418457,
|
| 1023 |
+
"learning_rate": 2.7876631079478055e-05,
|
| 1024 |
+
"loss": 0.1834,
|
| 1025 |
+
"step": 1400
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 4.999111111111111,
|
| 1029 |
+
"eval_accuracy": 0.8985,
|
| 1030 |
+
"eval_loss": 0.2476767748594284,
|
| 1031 |
+
"eval_runtime": 1541.9167,
|
| 1032 |
+
"eval_samples_per_second": 2.594,
|
| 1033 |
+
"eval_steps_per_second": 0.081,
|
| 1034 |
+
"step": 1406
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 5.013333333333334,
|
| 1038 |
+
"grad_norm": 4.798450469970703,
|
| 1039 |
+
"learning_rate": 2.76789244760775e-05,
|
| 1040 |
+
"loss": 0.2062,
|
| 1041 |
+
"step": 1410
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 5.0488888888888885,
|
| 1045 |
+
"grad_norm": 23.149574279785156,
|
| 1046 |
+
"learning_rate": 2.748121787267695e-05,
|
| 1047 |
+
"loss": 0.2002,
|
| 1048 |
+
"step": 1420
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 5.084444444444444,
|
| 1052 |
+
"grad_norm": 8.797248840332031,
|
| 1053 |
+
"learning_rate": 2.7283511269276396e-05,
|
| 1054 |
+
"loss": 0.2004,
|
| 1055 |
+
"step": 1430
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 5.12,
|
| 1059 |
+
"grad_norm": 11.881458282470703,
|
| 1060 |
+
"learning_rate": 2.708580466587584e-05,
|
| 1061 |
+
"loss": 0.1961,
|
| 1062 |
+
"step": 1440
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 5.155555555555556,
|
| 1066 |
+
"grad_norm": 12.706141471862793,
|
| 1067 |
+
"learning_rate": 2.6888098062475286e-05,
|
| 1068 |
+
"loss": 0.1797,
|
| 1069 |
+
"step": 1450
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 5.191111111111111,
|
| 1073 |
+
"grad_norm": 7.330083847045898,
|
| 1074 |
+
"learning_rate": 2.669039145907473e-05,
|
| 1075 |
+
"loss": 0.1711,
|
| 1076 |
+
"step": 1460
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 5.226666666666667,
|
| 1080 |
+
"grad_norm": 18.660188674926758,
|
| 1081 |
+
"learning_rate": 2.6492684855674182e-05,
|
| 1082 |
+
"loss": 0.1848,
|
| 1083 |
+
"step": 1470
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 5.262222222222222,
|
| 1087 |
+
"grad_norm": 7.45512056350708,
|
| 1088 |
+
"learning_rate": 2.6294978252273626e-05,
|
| 1089 |
+
"loss": 0.1827,
|
| 1090 |
+
"step": 1480
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 5.297777777777778,
|
| 1094 |
+
"grad_norm": 8.177336692810059,
|
| 1095 |
+
"learning_rate": 2.609727164887307e-05,
|
| 1096 |
+
"loss": 0.183,
|
| 1097 |
+
"step": 1490
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 5.333333333333333,
|
| 1101 |
+
"grad_norm": 14.408385276794434,
|
| 1102 |
+
"learning_rate": 2.589956504547252e-05,
|
| 1103 |
+
"loss": 0.1949,
|
| 1104 |
+
"step": 1500
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 5.368888888888889,
|
| 1108 |
+
"grad_norm": 7.131970405578613,
|
| 1109 |
+
"learning_rate": 2.5701858442071967e-05,
|
| 1110 |
+
"loss": 0.191,
|
| 1111 |
+
"step": 1510
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 5.404444444444445,
|
| 1115 |
+
"grad_norm": 14.019416809082031,
|
| 1116 |
+
"learning_rate": 2.5504151838671416e-05,
|
| 1117 |
+
"loss": 0.2121,
|
| 1118 |
+
"step": 1520
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 5.44,
|
| 1122 |
+
"grad_norm": 30.826915740966797,
|
| 1123 |
+
"learning_rate": 2.530644523527086e-05,
|
| 1124 |
+
"loss": 0.2317,
|
| 1125 |
+
"step": 1530
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 5.475555555555555,
|
| 1129 |
+
"grad_norm": 13.3015718460083,
|
| 1130 |
+
"learning_rate": 2.5108738631870305e-05,
|
| 1131 |
+
"loss": 0.2021,
|
| 1132 |
+
"step": 1540
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 5.511111111111111,
|
| 1136 |
+
"grad_norm": 24.24390983581543,
|
| 1137 |
+
"learning_rate": 2.4911032028469753e-05,
|
| 1138 |
+
"loss": 0.2085,
|
| 1139 |
+
"step": 1550
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 5.546666666666667,
|
| 1143 |
+
"grad_norm": 7.3809285163879395,
|
| 1144 |
+
"learning_rate": 2.4713325425069198e-05,
|
| 1145 |
+
"loss": 0.168,
|
| 1146 |
+
"step": 1560
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 5.582222222222223,
|
| 1150 |
+
"grad_norm": 17.8999080657959,
|
| 1151 |
+
"learning_rate": 2.4515618821668646e-05,
|
| 1152 |
+
"loss": 0.2103,
|
| 1153 |
+
"step": 1570
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 5.6177777777777775,
|
| 1157 |
+
"grad_norm": 19.31194305419922,
|
| 1158 |
+
"learning_rate": 2.431791221826809e-05,
|
| 1159 |
+
"loss": 0.1994,
|
| 1160 |
+
"step": 1580
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 5.653333333333333,
|
| 1164 |
+
"grad_norm": 6.58336067199707,
|
| 1165 |
+
"learning_rate": 2.412020561486754e-05,
|
| 1166 |
+
"loss": 0.1677,
|
| 1167 |
+
"step": 1590
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 5.688888888888889,
|
| 1171 |
+
"grad_norm": 12.235469818115234,
|
| 1172 |
+
"learning_rate": 2.3922499011466984e-05,
|
| 1173 |
+
"loss": 0.1707,
|
| 1174 |
+
"step": 1600
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 5.724444444444444,
|
| 1178 |
+
"grad_norm": 21.070587158203125,
|
| 1179 |
+
"learning_rate": 2.3724792408066432e-05,
|
| 1180 |
+
"loss": 0.1754,
|
| 1181 |
+
"step": 1610
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 5.76,
|
| 1185 |
+
"grad_norm": 20.813701629638672,
|
| 1186 |
+
"learning_rate": 2.3527085804665877e-05,
|
| 1187 |
+
"loss": 0.1576,
|
| 1188 |
+
"step": 1620
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 5.795555555555556,
|
| 1192 |
+
"grad_norm": 17.073057174682617,
|
| 1193 |
+
"learning_rate": 2.332937920126532e-05,
|
| 1194 |
+
"loss": 0.1631,
|
| 1195 |
+
"step": 1630
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 5.831111111111111,
|
| 1199 |
+
"grad_norm": 10.700809478759766,
|
| 1200 |
+
"learning_rate": 2.313167259786477e-05,
|
| 1201 |
+
"loss": 0.1854,
|
| 1202 |
+
"step": 1640
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 5.866666666666667,
|
| 1206 |
+
"grad_norm": 11.791812896728516,
|
| 1207 |
+
"learning_rate": 2.2933965994464214e-05,
|
| 1208 |
+
"loss": 0.187,
|
| 1209 |
+
"step": 1650
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 5.902222222222222,
|
| 1213 |
+
"grad_norm": 13.353257179260254,
|
| 1214 |
+
"learning_rate": 2.2736259391063662e-05,
|
| 1215 |
+
"loss": 0.1798,
|
| 1216 |
+
"step": 1660
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 5.937777777777778,
|
| 1220 |
+
"grad_norm": 13.19456672668457,
|
| 1221 |
+
"learning_rate": 2.2538552787663107e-05,
|
| 1222 |
+
"loss": 0.1817,
|
| 1223 |
+
"step": 1670
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 5.973333333333334,
|
| 1227 |
+
"grad_norm": 13.983536720275879,
|
| 1228 |
+
"learning_rate": 2.2340846184262555e-05,
|
| 1229 |
+
"loss": 0.2065,
|
| 1230 |
+
"step": 1680
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 5.998222222222222,
|
| 1234 |
+
"eval_accuracy": 0.95125,
|
| 1235 |
+
"eval_loss": 0.13313554227352142,
|
| 1236 |
+
"eval_runtime": 1541.5332,
|
| 1237 |
+
"eval_samples_per_second": 2.595,
|
| 1238 |
+
"eval_steps_per_second": 0.081,
|
| 1239 |
+
"step": 1687
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 6.0088888888888885,
|
| 1243 |
+
"grad_norm": 10.821728706359863,
|
| 1244 |
+
"learning_rate": 2.2143139580862003e-05,
|
| 1245 |
+
"loss": 0.1775,
|
| 1246 |
+
"step": 1690
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 6.044444444444444,
|
| 1250 |
+
"grad_norm": 16.398344039916992,
|
| 1251 |
+
"learning_rate": 2.1945432977461448e-05,
|
| 1252 |
+
"loss": 0.179,
|
| 1253 |
+
"step": 1700
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 6.08,
|
| 1257 |
+
"grad_norm": 18.149566650390625,
|
| 1258 |
+
"learning_rate": 2.1747726374060896e-05,
|
| 1259 |
+
"loss": 0.1822,
|
| 1260 |
+
"step": 1710
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 6.115555555555556,
|
| 1264 |
+
"grad_norm": 10.178900718688965,
|
| 1265 |
+
"learning_rate": 2.155001977066034e-05,
|
| 1266 |
+
"loss": 0.1779,
|
| 1267 |
+
"step": 1720
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 6.151111111111111,
|
| 1271 |
+
"grad_norm": 22.30776023864746,
|
| 1272 |
+
"learning_rate": 2.135231316725979e-05,
|
| 1273 |
+
"loss": 0.2134,
|
| 1274 |
+
"step": 1730
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 6.1866666666666665,
|
| 1278 |
+
"grad_norm": 10.142642974853516,
|
| 1279 |
+
"learning_rate": 2.1154606563859234e-05,
|
| 1280 |
+
"loss": 0.1676,
|
| 1281 |
+
"step": 1740
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 6.222222222222222,
|
| 1285 |
+
"grad_norm": 11.22693920135498,
|
| 1286 |
+
"learning_rate": 2.0956899960458682e-05,
|
| 1287 |
+
"loss": 0.153,
|
| 1288 |
+
"step": 1750
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 6.257777777777778,
|
| 1292 |
+
"grad_norm": 16.93440055847168,
|
| 1293 |
+
"learning_rate": 2.0759193357058127e-05,
|
| 1294 |
+
"loss": 0.1707,
|
| 1295 |
+
"step": 1760
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 6.293333333333333,
|
| 1299 |
+
"grad_norm": 15.217277526855469,
|
| 1300 |
+
"learning_rate": 2.0561486753657575e-05,
|
| 1301 |
+
"loss": 0.1661,
|
| 1302 |
+
"step": 1770
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 6.328888888888889,
|
| 1306 |
+
"grad_norm": 10.898624420166016,
|
| 1307 |
+
"learning_rate": 2.036378015025702e-05,
|
| 1308 |
+
"loss": 0.1621,
|
| 1309 |
+
"step": 1780
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 6.364444444444445,
|
| 1313 |
+
"grad_norm": 8.637317657470703,
|
| 1314 |
+
"learning_rate": 2.0166073546856464e-05,
|
| 1315 |
+
"loss": 0.1895,
|
| 1316 |
+
"step": 1790
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 6.4,
|
| 1320 |
+
"grad_norm": 9.220898628234863,
|
| 1321 |
+
"learning_rate": 1.9968366943455912e-05,
|
| 1322 |
+
"loss": 0.1598,
|
| 1323 |
+
"step": 1800
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 6.435555555555555,
|
| 1327 |
+
"grad_norm": 10.16035270690918,
|
| 1328 |
+
"learning_rate": 1.9770660340055357e-05,
|
| 1329 |
+
"loss": 0.1688,
|
| 1330 |
+
"step": 1810
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 6.471111111111111,
|
| 1334 |
+
"grad_norm": 12.095807075500488,
|
| 1335 |
+
"learning_rate": 1.9572953736654805e-05,
|
| 1336 |
+
"loss": 0.201,
|
| 1337 |
+
"step": 1820
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 6.506666666666667,
|
| 1341 |
+
"grad_norm": 23.264787673950195,
|
| 1342 |
+
"learning_rate": 1.937524713325425e-05,
|
| 1343 |
+
"loss": 0.1788,
|
| 1344 |
+
"step": 1830
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 6.542222222222223,
|
| 1348 |
+
"grad_norm": 14.32829475402832,
|
| 1349 |
+
"learning_rate": 1.9177540529853698e-05,
|
| 1350 |
+
"loss": 0.1723,
|
| 1351 |
+
"step": 1840
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 6.5777777777777775,
|
| 1355 |
+
"grad_norm": 6.539437770843506,
|
| 1356 |
+
"learning_rate": 1.8979833926453143e-05,
|
| 1357 |
+
"loss": 0.1761,
|
| 1358 |
+
"step": 1850
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 6.613333333333333,
|
| 1362 |
+
"grad_norm": 22.11655616760254,
|
| 1363 |
+
"learning_rate": 1.878212732305259e-05,
|
| 1364 |
+
"loss": 0.1652,
|
| 1365 |
+
"step": 1860
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 6.648888888888889,
|
| 1369 |
+
"grad_norm": 17.654651641845703,
|
| 1370 |
+
"learning_rate": 1.858442071965204e-05,
|
| 1371 |
+
"loss": 0.1959,
|
| 1372 |
+
"step": 1870
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 6.684444444444445,
|
| 1376 |
+
"grad_norm": 11.6602201461792,
|
| 1377 |
+
"learning_rate": 1.8386714116251484e-05,
|
| 1378 |
+
"loss": 0.1927,
|
| 1379 |
+
"step": 1880
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 6.72,
|
| 1383 |
+
"grad_norm": 13.29917049407959,
|
| 1384 |
+
"learning_rate": 1.8189007512850932e-05,
|
| 1385 |
+
"loss": 0.146,
|
| 1386 |
+
"step": 1890
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 6.7555555555555555,
|
| 1390 |
+
"grad_norm": 20.093284606933594,
|
| 1391 |
+
"learning_rate": 1.7991300909450377e-05,
|
| 1392 |
+
"loss": 0.1623,
|
| 1393 |
+
"step": 1900
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 6.791111111111111,
|
| 1397 |
+
"grad_norm": 8.822929382324219,
|
| 1398 |
+
"learning_rate": 1.7793594306049825e-05,
|
| 1399 |
+
"loss": 0.1647,
|
| 1400 |
+
"step": 1910
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 6.826666666666666,
|
| 1404 |
+
"grad_norm": 5.047310829162598,
|
| 1405 |
+
"learning_rate": 1.759588770264927e-05,
|
| 1406 |
+
"loss": 0.1479,
|
| 1407 |
+
"step": 1920
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 6.862222222222222,
|
| 1411 |
+
"grad_norm": 12.150626182556152,
|
| 1412 |
+
"learning_rate": 1.7398181099248718e-05,
|
| 1413 |
+
"loss": 0.1616,
|
| 1414 |
+
"step": 1930
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 6.897777777777778,
|
| 1418 |
+
"grad_norm": 9.621389389038086,
|
| 1419 |
+
"learning_rate": 1.7200474495848162e-05,
|
| 1420 |
+
"loss": 0.1556,
|
| 1421 |
+
"step": 1940
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 6.933333333333334,
|
| 1425 |
+
"grad_norm": 12.77206039428711,
|
| 1426 |
+
"learning_rate": 1.7002767892447607e-05,
|
| 1427 |
+
"loss": 0.1941,
|
| 1428 |
+
"step": 1950
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 6.968888888888889,
|
| 1432 |
+
"grad_norm": 12.906224250793457,
|
| 1433 |
+
"learning_rate": 1.6805061289047055e-05,
|
| 1434 |
+
"loss": 0.1555,
|
| 1435 |
+
"step": 1960
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 6.997333333333334,
|
| 1439 |
+
"eval_accuracy": 0.94725,
|
| 1440 |
+
"eval_loss": 0.13042479753494263,
|
| 1441 |
+
"eval_runtime": 1543.7566,
|
| 1442 |
+
"eval_samples_per_second": 2.591,
|
| 1443 |
+
"eval_steps_per_second": 0.081,
|
| 1444 |
+
"step": 1968
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 7.004444444444444,
|
| 1448 |
+
"grad_norm": 13.00174331665039,
|
| 1449 |
+
"learning_rate": 1.66073546856465e-05,
|
| 1450 |
+
"loss": 0.1638,
|
| 1451 |
+
"step": 1970
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 7.04,
|
| 1455 |
+
"grad_norm": 18.73016929626465,
|
| 1456 |
+
"learning_rate": 1.6409648082245948e-05,
|
| 1457 |
+
"loss": 0.1565,
|
| 1458 |
+
"step": 1980
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 7.075555555555556,
|
| 1462 |
+
"grad_norm": 7.3105010986328125,
|
| 1463 |
+
"learning_rate": 1.6211941478845393e-05,
|
| 1464 |
+
"loss": 0.1375,
|
| 1465 |
+
"step": 1990
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 7.111111111111111,
|
| 1469 |
+
"grad_norm": 9.498517036437988,
|
| 1470 |
+
"learning_rate": 1.601423487544484e-05,
|
| 1471 |
+
"loss": 0.1353,
|
| 1472 |
+
"step": 2000
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 7.1466666666666665,
|
| 1476 |
+
"grad_norm": 19.775840759277344,
|
| 1477 |
+
"learning_rate": 1.5816528272044286e-05,
|
| 1478 |
+
"loss": 0.144,
|
| 1479 |
+
"step": 2010
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 7.182222222222222,
|
| 1483 |
+
"grad_norm": 14.198583602905273,
|
| 1484 |
+
"learning_rate": 1.5618821668643734e-05,
|
| 1485 |
+
"loss": 0.1414,
|
| 1486 |
+
"step": 2020
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 7.217777777777778,
|
| 1490 |
+
"grad_norm": 9.943511962890625,
|
| 1491 |
+
"learning_rate": 1.5421115065243182e-05,
|
| 1492 |
+
"loss": 0.143,
|
| 1493 |
+
"step": 2030
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 7.253333333333333,
|
| 1497 |
+
"grad_norm": 9.717743873596191,
|
| 1498 |
+
"learning_rate": 1.5223408461842625e-05,
|
| 1499 |
+
"loss": 0.1644,
|
| 1500 |
+
"step": 2040
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 7.288888888888889,
|
| 1504 |
+
"grad_norm": 8.998052597045898,
|
| 1505 |
+
"learning_rate": 1.5025701858442073e-05,
|
| 1506 |
+
"loss": 0.1434,
|
| 1507 |
+
"step": 2050
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 7.3244444444444445,
|
| 1511 |
+
"grad_norm": 7.94929313659668,
|
| 1512 |
+
"learning_rate": 1.4827995255041518e-05,
|
| 1513 |
+
"loss": 0.1408,
|
| 1514 |
+
"step": 2060
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 7.36,
|
| 1518 |
+
"grad_norm": 7.949029922485352,
|
| 1519 |
+
"learning_rate": 1.4630288651640966e-05,
|
| 1520 |
+
"loss": 0.1426,
|
| 1521 |
+
"step": 2070
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 7.395555555555555,
|
| 1525 |
+
"grad_norm": 13.912178993225098,
|
| 1526 |
+
"learning_rate": 1.4432582048240412e-05,
|
| 1527 |
+
"loss": 0.1423,
|
| 1528 |
+
"step": 2080
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 7.431111111111111,
|
| 1532 |
+
"grad_norm": 9.720952987670898,
|
| 1533 |
+
"learning_rate": 1.423487544483986e-05,
|
| 1534 |
+
"loss": 0.1611,
|
| 1535 |
+
"step": 2090
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.466666666666667,
|
| 1539 |
+
"grad_norm": 15.604865074157715,
|
| 1540 |
+
"learning_rate": 1.4037168841439305e-05,
|
| 1541 |
+
"loss": 0.1355,
|
| 1542 |
+
"step": 2100
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 7.502222222222223,
|
| 1546 |
+
"grad_norm": 9.24775505065918,
|
| 1547 |
+
"learning_rate": 1.383946223803875e-05,
|
| 1548 |
+
"loss": 0.1471,
|
| 1549 |
+
"step": 2110
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 7.5377777777777775,
|
| 1553 |
+
"grad_norm": 10.455300331115723,
|
| 1554 |
+
"learning_rate": 1.3641755634638198e-05,
|
| 1555 |
+
"loss": 0.1271,
|
| 1556 |
+
"step": 2120
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.573333333333333,
|
| 1560 |
+
"grad_norm": 15.100184440612793,
|
| 1561 |
+
"learning_rate": 1.3444049031237643e-05,
|
| 1562 |
+
"loss": 0.1388,
|
| 1563 |
+
"step": 2130
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 7.608888888888889,
|
| 1567 |
+
"grad_norm": 10.345260620117188,
|
| 1568 |
+
"learning_rate": 1.3246342427837091e-05,
|
| 1569 |
+
"loss": 0.1354,
|
| 1570 |
+
"step": 2140
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 7.644444444444445,
|
| 1574 |
+
"grad_norm": 12.270657539367676,
|
| 1575 |
+
"learning_rate": 1.3048635824436536e-05,
|
| 1576 |
+
"loss": 0.154,
|
| 1577 |
+
"step": 2150
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 7.68,
|
| 1581 |
+
"grad_norm": 9.241147994995117,
|
| 1582 |
+
"learning_rate": 1.2850929221035984e-05,
|
| 1583 |
+
"loss": 0.1259,
|
| 1584 |
+
"step": 2160
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 7.7155555555555555,
|
| 1588 |
+
"grad_norm": 16.320280075073242,
|
| 1589 |
+
"learning_rate": 1.265322261763543e-05,
|
| 1590 |
+
"loss": 0.1351,
|
| 1591 |
+
"step": 2170
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 7.751111111111111,
|
| 1595 |
+
"grad_norm": 11.753779411315918,
|
| 1596 |
+
"learning_rate": 1.2455516014234877e-05,
|
| 1597 |
+
"loss": 0.1476,
|
| 1598 |
+
"step": 2180
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.786666666666667,
|
| 1602 |
+
"grad_norm": 10.116710662841797,
|
| 1603 |
+
"learning_rate": 1.2257809410834323e-05,
|
| 1604 |
+
"loss": 0.1461,
|
| 1605 |
+
"step": 2190
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.822222222222222,
|
| 1609 |
+
"grad_norm": 11.41565990447998,
|
| 1610 |
+
"learning_rate": 1.206010280743377e-05,
|
| 1611 |
+
"loss": 0.1505,
|
| 1612 |
+
"step": 2200
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.857777777777778,
|
| 1616 |
+
"grad_norm": 12.816720008850098,
|
| 1617 |
+
"learning_rate": 1.1862396204033216e-05,
|
| 1618 |
+
"loss": 0.1357,
|
| 1619 |
+
"step": 2210
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.8933333333333335,
|
| 1623 |
+
"grad_norm": 14.3566255569458,
|
| 1624 |
+
"learning_rate": 1.166468960063266e-05,
|
| 1625 |
+
"loss": 0.1433,
|
| 1626 |
+
"step": 2220
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.928888888888888,
|
| 1630 |
+
"grad_norm": 13.433180809020996,
|
| 1631 |
+
"learning_rate": 1.1466982997232107e-05,
|
| 1632 |
+
"loss": 0.1697,
|
| 1633 |
+
"step": 2230
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 7.964444444444444,
|
| 1637 |
+
"grad_norm": 10.5477933883667,
|
| 1638 |
+
"learning_rate": 1.1269276393831553e-05,
|
| 1639 |
+
"loss": 0.1321,
|
| 1640 |
+
"step": 2240
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 8.0,
|
| 1644 |
+
"grad_norm": 11.933510780334473,
|
| 1645 |
+
"learning_rate": 1.1071569790431002e-05,
|
| 1646 |
+
"loss": 0.1521,
|
| 1647 |
+
"step": 2250
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 8.0,
|
| 1651 |
+
"eval_accuracy": 0.92925,
|
| 1652 |
+
"eval_loss": 0.1837126612663269,
|
| 1653 |
+
"eval_runtime": 1563.6163,
|
| 1654 |
+
"eval_samples_per_second": 2.558,
|
| 1655 |
+
"eval_steps_per_second": 0.08,
|
| 1656 |
+
"step": 2250
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 8.035555555555556,
|
| 1660 |
+
"grad_norm": 22.18363380432129,
|
| 1661 |
+
"learning_rate": 1.0873863187030448e-05,
|
| 1662 |
+
"loss": 0.1396,
|
| 1663 |
+
"step": 2260
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 8.071111111111112,
|
| 1667 |
+
"grad_norm": 13.1504545211792,
|
| 1668 |
+
"learning_rate": 1.0676156583629894e-05,
|
| 1669 |
+
"loss": 0.1265,
|
| 1670 |
+
"step": 2270
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 8.106666666666667,
|
| 1674 |
+
"grad_norm": 21.022546768188477,
|
| 1675 |
+
"learning_rate": 1.0478449980229341e-05,
|
| 1676 |
+
"loss": 0.1476,
|
| 1677 |
+
"step": 2280
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 8.142222222222221,
|
| 1681 |
+
"grad_norm": 15.906450271606445,
|
| 1682 |
+
"learning_rate": 1.0280743376828787e-05,
|
| 1683 |
+
"loss": 0.1285,
|
| 1684 |
+
"step": 2290
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 8.177777777777777,
|
| 1688 |
+
"grad_norm": 10.43397331237793,
|
| 1689 |
+
"learning_rate": 1.0083036773428232e-05,
|
| 1690 |
+
"loss": 0.1336,
|
| 1691 |
+
"step": 2300
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 8.213333333333333,
|
| 1695 |
+
"grad_norm": 7.347660541534424,
|
| 1696 |
+
"learning_rate": 9.885330170027678e-06,
|
| 1697 |
+
"loss": 0.1418,
|
| 1698 |
+
"step": 2310
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 8.248888888888889,
|
| 1702 |
+
"grad_norm": 9.91401481628418,
|
| 1703 |
+
"learning_rate": 9.687623566627125e-06,
|
| 1704 |
+
"loss": 0.1212,
|
| 1705 |
+
"step": 2320
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 8.284444444444444,
|
| 1709 |
+
"grad_norm": 9.150163650512695,
|
| 1710 |
+
"learning_rate": 9.489916963226571e-06,
|
| 1711 |
+
"loss": 0.1478,
|
| 1712 |
+
"step": 2330
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 8.32,
|
| 1716 |
+
"grad_norm": 11.815370559692383,
|
| 1717 |
+
"learning_rate": 9.29221035982602e-06,
|
| 1718 |
+
"loss": 0.1258,
|
| 1719 |
+
"step": 2340
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 8.355555555555556,
|
| 1723 |
+
"grad_norm": 13.915023803710938,
|
| 1724 |
+
"learning_rate": 9.094503756425466e-06,
|
| 1725 |
+
"loss": 0.1257,
|
| 1726 |
+
"step": 2350
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 8.391111111111112,
|
| 1730 |
+
"grad_norm": 8.791934967041016,
|
| 1731 |
+
"learning_rate": 8.896797153024912e-06,
|
| 1732 |
+
"loss": 0.1342,
|
| 1733 |
+
"step": 2360
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 8.426666666666666,
|
| 1737 |
+
"grad_norm": 10.984416961669922,
|
| 1738 |
+
"learning_rate": 8.699090549624359e-06,
|
| 1739 |
+
"loss": 0.1292,
|
| 1740 |
+
"step": 2370
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 8.462222222222222,
|
| 1744 |
+
"grad_norm": 14.471619606018066,
|
| 1745 |
+
"learning_rate": 8.501383946223804e-06,
|
| 1746 |
+
"loss": 0.1221,
|
| 1747 |
+
"step": 2380
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 8.497777777777777,
|
| 1751 |
+
"grad_norm": 8.994637489318848,
|
| 1752 |
+
"learning_rate": 8.30367734282325e-06,
|
| 1753 |
+
"loss": 0.1246,
|
| 1754 |
+
"step": 2390
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 8.533333333333333,
|
| 1758 |
+
"grad_norm": 16.3120059967041,
|
| 1759 |
+
"learning_rate": 8.105970739422696e-06,
|
| 1760 |
+
"loss": 0.1345,
|
| 1761 |
+
"step": 2400
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 8.568888888888889,
|
| 1765 |
+
"grad_norm": 12.578171730041504,
|
| 1766 |
+
"learning_rate": 7.908264136022143e-06,
|
| 1767 |
+
"loss": 0.1386,
|
| 1768 |
+
"step": 2410
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 8.604444444444445,
|
| 1772 |
+
"grad_norm": 10.959859848022461,
|
| 1773 |
+
"learning_rate": 7.710557532621591e-06,
|
| 1774 |
+
"loss": 0.1367,
|
| 1775 |
+
"step": 2420
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 8.64,
|
| 1779 |
+
"grad_norm": 11.834918022155762,
|
| 1780 |
+
"learning_rate": 7.5128509292210365e-06,
|
| 1781 |
+
"loss": 0.136,
|
| 1782 |
+
"step": 2430
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 8.675555555555556,
|
| 1786 |
+
"grad_norm": 14.118315696716309,
|
| 1787 |
+
"learning_rate": 7.315144325820483e-06,
|
| 1788 |
+
"loss": 0.1232,
|
| 1789 |
+
"step": 2440
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 8.71111111111111,
|
| 1793 |
+
"grad_norm": 8.814461708068848,
|
| 1794 |
+
"learning_rate": 7.11743772241993e-06,
|
| 1795 |
+
"loss": 0.1052,
|
| 1796 |
+
"step": 2450
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 8.746666666666666,
|
| 1800 |
+
"grad_norm": 8.637076377868652,
|
| 1801 |
+
"learning_rate": 6.919731119019375e-06,
|
| 1802 |
+
"loss": 0.1062,
|
| 1803 |
+
"step": 2460
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 8.782222222222222,
|
| 1807 |
+
"grad_norm": 9.377944946289062,
|
| 1808 |
+
"learning_rate": 6.722024515618821e-06,
|
| 1809 |
+
"loss": 0.1334,
|
| 1810 |
+
"step": 2470
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 8.817777777777778,
|
| 1814 |
+
"grad_norm": 14.221964836120605,
|
| 1815 |
+
"learning_rate": 6.524317912218268e-06,
|
| 1816 |
+
"loss": 0.1387,
|
| 1817 |
+
"step": 2480
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 8.853333333333333,
|
| 1821 |
+
"grad_norm": 20.073240280151367,
|
| 1822 |
+
"learning_rate": 6.326611308817715e-06,
|
| 1823 |
+
"loss": 0.1294,
|
| 1824 |
+
"step": 2490
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 8.88888888888889,
|
| 1828 |
+
"grad_norm": 9.537182807922363,
|
| 1829 |
+
"learning_rate": 6.1289047054171615e-06,
|
| 1830 |
+
"loss": 0.1261,
|
| 1831 |
+
"step": 2500
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 8.924444444444445,
|
| 1835 |
+
"grad_norm": 12.444225311279297,
|
| 1836 |
+
"learning_rate": 5.931198102016608e-06,
|
| 1837 |
+
"loss": 0.129,
|
| 1838 |
+
"step": 2510
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 8.96,
|
| 1842 |
+
"grad_norm": 10.986297607421875,
|
| 1843 |
+
"learning_rate": 5.7334914986160535e-06,
|
| 1844 |
+
"loss": 0.1171,
|
| 1845 |
+
"step": 2520
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 8.995555555555555,
|
| 1849 |
+
"grad_norm": 5.872843265533447,
|
| 1850 |
+
"learning_rate": 5.535784895215501e-06,
|
| 1851 |
+
"loss": 0.1512,
|
| 1852 |
+
"step": 2530
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 8.99911111111111,
|
| 1856 |
+
"eval_accuracy": 0.9405,
|
| 1857 |
+
"eval_loss": 0.1708436906337738,
|
| 1858 |
+
"eval_runtime": 1529.8652,
|
| 1859 |
+
"eval_samples_per_second": 2.615,
|
| 1860 |
+
"eval_steps_per_second": 0.082,
|
| 1861 |
+
"step": 2531
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 9.03111111111111,
|
| 1865 |
+
"grad_norm": 7.123359680175781,
|
| 1866 |
+
"learning_rate": 5.338078291814947e-06,
|
| 1867 |
+
"loss": 0.1029,
|
| 1868 |
+
"step": 2540
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 9.066666666666666,
|
| 1872 |
+
"grad_norm": 14.264687538146973,
|
| 1873 |
+
"learning_rate": 5.140371688414394e-06,
|
| 1874 |
+
"loss": 0.1146,
|
| 1875 |
+
"step": 2550
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 9.102222222222222,
|
| 1879 |
+
"grad_norm": 6.885898113250732,
|
| 1880 |
+
"learning_rate": 4.942665085013839e-06,
|
| 1881 |
+
"loss": 0.1054,
|
| 1882 |
+
"step": 2560
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 9.137777777777778,
|
| 1886 |
+
"grad_norm": 5.8031086921691895,
|
| 1887 |
+
"learning_rate": 4.744958481613286e-06,
|
| 1888 |
+
"loss": 0.1048,
|
| 1889 |
+
"step": 2570
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 9.173333333333334,
|
| 1893 |
+
"grad_norm": 6.764829158782959,
|
| 1894 |
+
"learning_rate": 4.547251878212733e-06,
|
| 1895 |
+
"loss": 0.1031,
|
| 1896 |
+
"step": 2580
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 9.20888888888889,
|
| 1900 |
+
"grad_norm": 3.4530787467956543,
|
| 1901 |
+
"learning_rate": 4.349545274812179e-06,
|
| 1902 |
+
"loss": 0.1027,
|
| 1903 |
+
"step": 2590
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 9.244444444444444,
|
| 1907 |
+
"grad_norm": 19.80527114868164,
|
| 1908 |
+
"learning_rate": 4.151838671411625e-06,
|
| 1909 |
+
"loss": 0.1152,
|
| 1910 |
+
"step": 2600
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 9.28,
|
| 1914 |
+
"grad_norm": 12.928886413574219,
|
| 1915 |
+
"learning_rate": 3.954132068011071e-06,
|
| 1916 |
+
"loss": 0.1,
|
| 1917 |
+
"step": 2610
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 9.315555555555555,
|
| 1921 |
+
"grad_norm": 14.547968864440918,
|
| 1922 |
+
"learning_rate": 3.7564254646105183e-06,
|
| 1923 |
+
"loss": 0.1253,
|
| 1924 |
+
"step": 2620
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 9.351111111111111,
|
| 1928 |
+
"grad_norm": 8.808683395385742,
|
| 1929 |
+
"learning_rate": 3.558718861209965e-06,
|
| 1930 |
+
"loss": 0.116,
|
| 1931 |
+
"step": 2630
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 9.386666666666667,
|
| 1935 |
+
"grad_norm": 18.376087188720703,
|
| 1936 |
+
"learning_rate": 3.3610122578094107e-06,
|
| 1937 |
+
"loss": 0.1247,
|
| 1938 |
+
"step": 2640
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 9.422222222222222,
|
| 1942 |
+
"grad_norm": 9.673834800720215,
|
| 1943 |
+
"learning_rate": 3.1633056544088575e-06,
|
| 1944 |
+
"loss": 0.1322,
|
| 1945 |
+
"step": 2650
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 9.457777777777778,
|
| 1949 |
+
"grad_norm": 10.23854923248291,
|
| 1950 |
+
"learning_rate": 2.965599051008304e-06,
|
| 1951 |
+
"loss": 0.1233,
|
| 1952 |
+
"step": 2660
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 9.493333333333334,
|
| 1956 |
+
"grad_norm": 6.59231424331665,
|
| 1957 |
+
"learning_rate": 2.7678924476077504e-06,
|
| 1958 |
+
"loss": 0.1172,
|
| 1959 |
+
"step": 2670
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 9.528888888888888,
|
| 1963 |
+
"grad_norm": 4.731305122375488,
|
| 1964 |
+
"learning_rate": 2.570185844207197e-06,
|
| 1965 |
+
"loss": 0.0819,
|
| 1966 |
+
"step": 2680
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 9.564444444444444,
|
| 1970 |
+
"grad_norm": 8.215780258178711,
|
| 1971 |
+
"learning_rate": 2.372479240806643e-06,
|
| 1972 |
+
"loss": 0.128,
|
| 1973 |
+
"step": 2690
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 9.6,
|
| 1977 |
+
"grad_norm": 14.254526138305664,
|
| 1978 |
+
"learning_rate": 2.1747726374060897e-06,
|
| 1979 |
+
"loss": 0.1286,
|
| 1980 |
+
"step": 2700
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 9.635555555555555,
|
| 1984 |
+
"grad_norm": 13.742310523986816,
|
| 1985 |
+
"learning_rate": 1.9770660340055357e-06,
|
| 1986 |
+
"loss": 0.1205,
|
| 1987 |
+
"step": 2710
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 9.671111111111111,
|
| 1991 |
+
"grad_norm": 10.408873558044434,
|
| 1992 |
+
"learning_rate": 1.7793594306049826e-06,
|
| 1993 |
+
"loss": 0.1158,
|
| 1994 |
+
"step": 2720
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 9.706666666666667,
|
| 1998 |
+
"grad_norm": 5.809567451477051,
|
| 1999 |
+
"learning_rate": 1.5816528272044288e-06,
|
| 2000 |
+
"loss": 0.1195,
|
| 2001 |
+
"step": 2730
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 9.742222222222223,
|
| 2005 |
+
"grad_norm": 9.508487701416016,
|
| 2006 |
+
"learning_rate": 1.3839462238038752e-06,
|
| 2007 |
+
"loss": 0.117,
|
| 2008 |
+
"step": 2740
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 9.777777777777779,
|
| 2012 |
+
"grad_norm": 9.351639747619629,
|
| 2013 |
+
"learning_rate": 1.1862396204033214e-06,
|
| 2014 |
+
"loss": 0.1286,
|
| 2015 |
+
"step": 2750
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 9.813333333333333,
|
| 2019 |
+
"grad_norm": 11.513580322265625,
|
| 2020 |
+
"learning_rate": 9.885330170027678e-07,
|
| 2021 |
+
"loss": 0.0973,
|
| 2022 |
+
"step": 2760
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 9.848888888888888,
|
| 2026 |
+
"grad_norm": 6.790437698364258,
|
| 2027 |
+
"learning_rate": 7.908264136022144e-07,
|
| 2028 |
+
"loss": 0.1026,
|
| 2029 |
+
"step": 2770
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 9.884444444444444,
|
| 2033 |
+
"grad_norm": 12.467495918273926,
|
| 2034 |
+
"learning_rate": 5.931198102016607e-07,
|
| 2035 |
+
"loss": 0.108,
|
| 2036 |
+
"step": 2780
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 9.92,
|
| 2040 |
+
"grad_norm": 9.36400032043457,
|
| 2041 |
+
"learning_rate": 3.954132068011072e-07,
|
| 2042 |
+
"loss": 0.1155,
|
| 2043 |
+
"step": 2790
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 9.955555555555556,
|
| 2047 |
+
"grad_norm": 6.071395397186279,
|
| 2048 |
+
"learning_rate": 1.977066034005536e-07,
|
| 2049 |
+
"loss": 0.1256,
|
| 2050 |
+
"step": 2800
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 9.991111111111111,
|
| 2054 |
+
"grad_norm": 8.255842208862305,
|
| 2055 |
+
"learning_rate": 0.0,
|
| 2056 |
+
"loss": 0.119,
|
| 2057 |
+
"step": 2810
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 9.991111111111111,
|
| 2061 |
+
"eval_accuracy": 0.954,
|
| 2062 |
+
"eval_loss": 0.13787274062633514,
|
| 2063 |
+
"eval_runtime": 1553.5401,
|
| 2064 |
+
"eval_samples_per_second": 2.575,
|
| 2065 |
+
"eval_steps_per_second": 0.08,
|
| 2066 |
+
"step": 2810
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 9.991111111111111,
|
| 2070 |
+
"step": 2810,
|
| 2071 |
+
"total_flos": 8.940202118432686e+18,
|
| 2072 |
+
"train_loss": 0.24847593615063568,
|
| 2073 |
+
"train_runtime": 161511.588,
|
| 2074 |
+
"train_samples_per_second": 2.229,
|
| 2075 |
+
"train_steps_per_second": 0.017
|
| 2076 |
+
}
|
| 2077 |
+
],
|
| 2078 |
+
"logging_steps": 10,
|
| 2079 |
+
"max_steps": 2810,
|
| 2080 |
+
"num_input_tokens_seen": 0,
|
| 2081 |
+
"num_train_epochs": 10,
|
| 2082 |
+
"save_steps": 500,
|
| 2083 |
+
"stateful_callbacks": {
|
| 2084 |
+
"TrainerControl": {
|
| 2085 |
+
"args": {
|
| 2086 |
+
"should_epoch_stop": false,
|
| 2087 |
+
"should_evaluate": false,
|
| 2088 |
+
"should_log": false,
|
| 2089 |
+
"should_save": true,
|
| 2090 |
+
"should_training_stop": true
|
| 2091 |
+
},
|
| 2092 |
+
"attributes": {}
|
| 2093 |
+
}
|
| 2094 |
+
},
|
| 2095 |
+
"total_flos": 8.940202118432686e+18,
|
| 2096 |
+
"train_batch_size": 32,
|
| 2097 |
+
"trial_name": null,
|
| 2098 |
+
"trial_params": null
|
| 2099 |
+
}
|