Model save
Browse files- all_results.json +6 -1
- train_results.json +8 -0
- trainer_state.json +3497 -0
all_results.json
CHANGED
|
@@ -4,5 +4,10 @@
|
|
| 4 |
"eval_loss": 0.08026164770126343,
|
| 5 |
"eval_runtime": 73.8075,
|
| 6 |
"eval_samples_per_second": 203.231,
|
| 7 |
-
"eval_steps_per_second": 12.709
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
}
|
|
|
|
| 4 |
"eval_loss": 0.08026164770126343,
|
| 5 |
"eval_runtime": 73.8075,
|
| 6 |
"eval_samples_per_second": 203.231,
|
| 7 |
+
"eval_steps_per_second": 12.709,
|
| 8 |
+
"total_flos": 5.42482821328896e+18,
|
| 9 |
+
"train_loss": 0.13969962793986365,
|
| 10 |
+
"train_runtime": 4718.8358,
|
| 11 |
+
"train_samples_per_second": 14.834,
|
| 12 |
+
"train_steps_per_second": 0.927
|
| 13 |
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.0,
|
| 3 |
+
"total_flos": 5.42482821328896e+18,
|
| 4 |
+
"train_loss": 0.13969962793986365,
|
| 5 |
+
"train_runtime": 4718.8358,
|
| 6 |
+
"train_samples_per_second": 14.834,
|
| 7 |
+
"train_steps_per_second": 0.927
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3497 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.08026164770126343,
|
| 3 |
+
"best_model_checkpoint": "./vit-base-cifar10/checkpoint-4300",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 4376,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.004570383912248629,
|
| 13 |
+
"grad_norm": 7.92866325378418,
|
| 14 |
+
"learning_rate": 0.00019954296160877515,
|
| 15 |
+
"loss": 0.2503,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.009140767824497258,
|
| 20 |
+
"grad_norm": 0.6853145956993103,
|
| 21 |
+
"learning_rate": 0.00019908592321755028,
|
| 22 |
+
"loss": 0.2887,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.013711151736745886,
|
| 27 |
+
"grad_norm": 8.078825950622559,
|
| 28 |
+
"learning_rate": 0.00019862888482632542,
|
| 29 |
+
"loss": 0.3522,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.018281535648994516,
|
| 34 |
+
"grad_norm": 0.5457130670547485,
|
| 35 |
+
"learning_rate": 0.00019817184643510056,
|
| 36 |
+
"loss": 0.2236,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.022851919561243144,
|
| 41 |
+
"grad_norm": 7.070144176483154,
|
| 42 |
+
"learning_rate": 0.0001977148080438757,
|
| 43 |
+
"loss": 0.3022,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.027422303473491772,
|
| 48 |
+
"grad_norm": 6.712225437164307,
|
| 49 |
+
"learning_rate": 0.00019725776965265083,
|
| 50 |
+
"loss": 0.2947,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.031992687385740404,
|
| 55 |
+
"grad_norm": 0.18951156735420227,
|
| 56 |
+
"learning_rate": 0.00019680073126142596,
|
| 57 |
+
"loss": 0.2885,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.03656307129798903,
|
| 62 |
+
"grad_norm": 5.366819858551025,
|
| 63 |
+
"learning_rate": 0.0001963436928702011,
|
| 64 |
+
"loss": 0.2822,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.04113345521023766,
|
| 69 |
+
"grad_norm": 1.0978411436080933,
|
| 70 |
+
"learning_rate": 0.00019588665447897624,
|
| 71 |
+
"loss": 0.2124,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.04570383912248629,
|
| 76 |
+
"grad_norm": 0.3122899532318115,
|
| 77 |
+
"learning_rate": 0.0001954296160877514,
|
| 78 |
+
"loss": 0.1043,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.04570383912248629,
|
| 83 |
+
"eval_accuracy": 0.919,
|
| 84 |
+
"eval_loss": 0.285547137260437,
|
| 85 |
+
"eval_runtime": 71.7238,
|
| 86 |
+
"eval_samples_per_second": 209.136,
|
| 87 |
+
"eval_steps_per_second": 13.078,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.050274223034734916,
|
| 92 |
+
"grad_norm": 0.872364342212677,
|
| 93 |
+
"learning_rate": 0.00019497257769652654,
|
| 94 |
+
"loss": 0.1525,
|
| 95 |
+
"step": 110
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.054844606946983544,
|
| 99 |
+
"grad_norm": 0.4638870358467102,
|
| 100 |
+
"learning_rate": 0.00019451553930530167,
|
| 101 |
+
"loss": 0.269,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.05941499085923217,
|
| 106 |
+
"grad_norm": 4.21981143951416,
|
| 107 |
+
"learning_rate": 0.0001940585009140768,
|
| 108 |
+
"loss": 0.3743,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.06398537477148081,
|
| 113 |
+
"grad_norm": 0.4475237727165222,
|
| 114 |
+
"learning_rate": 0.00019360146252285195,
|
| 115 |
+
"loss": 0.2608,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.06855575868372943,
|
| 120 |
+
"grad_norm": 6.892650127410889,
|
| 121 |
+
"learning_rate": 0.00019314442413162706,
|
| 122 |
+
"loss": 0.271,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.07312614259597806,
|
| 127 |
+
"grad_norm": 5.913930416107178,
|
| 128 |
+
"learning_rate": 0.0001926873857404022,
|
| 129 |
+
"loss": 0.1474,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.07769652650822668,
|
| 134 |
+
"grad_norm": 2.620058298110962,
|
| 135 |
+
"learning_rate": 0.00019223034734917733,
|
| 136 |
+
"loss": 0.1543,
|
| 137 |
+
"step": 170
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.08226691042047532,
|
| 141 |
+
"grad_norm": 8.168417930603027,
|
| 142 |
+
"learning_rate": 0.00019177330895795246,
|
| 143 |
+
"loss": 0.4235,
|
| 144 |
+
"step": 180
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.08683729433272395,
|
| 148 |
+
"grad_norm": 6.655405044555664,
|
| 149 |
+
"learning_rate": 0.0001913162705667276,
|
| 150 |
+
"loss": 0.2081,
|
| 151 |
+
"step": 190
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.09140767824497258,
|
| 155 |
+
"grad_norm": 0.3367229104042053,
|
| 156 |
+
"learning_rate": 0.00019085923217550274,
|
| 157 |
+
"loss": 0.2671,
|
| 158 |
+
"step": 200
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.09140767824497258,
|
| 162 |
+
"eval_accuracy": 0.9015333333333333,
|
| 163 |
+
"eval_loss": 0.3649848401546478,
|
| 164 |
+
"eval_runtime": 70.014,
|
| 165 |
+
"eval_samples_per_second": 214.243,
|
| 166 |
+
"eval_steps_per_second": 13.397,
|
| 167 |
+
"step": 200
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.09597806215722121,
|
| 171 |
+
"grad_norm": 7.15602970123291,
|
| 172 |
+
"learning_rate": 0.00019040219378427787,
|
| 173 |
+
"loss": 0.5093,
|
| 174 |
+
"step": 210
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.10054844606946983,
|
| 178 |
+
"grad_norm": 3.229771137237549,
|
| 179 |
+
"learning_rate": 0.000189945155393053,
|
| 180 |
+
"loss": 0.4777,
|
| 181 |
+
"step": 220
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.10511882998171847,
|
| 185 |
+
"grad_norm": 7.438943862915039,
|
| 186 |
+
"learning_rate": 0.00018948811700182815,
|
| 187 |
+
"loss": 0.407,
|
| 188 |
+
"step": 230
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.10968921389396709,
|
| 192 |
+
"grad_norm": 8.4913969039917,
|
| 193 |
+
"learning_rate": 0.00018903107861060328,
|
| 194 |
+
"loss": 0.338,
|
| 195 |
+
"step": 240
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.11425959780621572,
|
| 199 |
+
"grad_norm": 1.3278025388717651,
|
| 200 |
+
"learning_rate": 0.00018857404021937845,
|
| 201 |
+
"loss": 0.3518,
|
| 202 |
+
"step": 250
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.11882998171846434,
|
| 206 |
+
"grad_norm": 9.502877235412598,
|
| 207 |
+
"learning_rate": 0.00018811700182815358,
|
| 208 |
+
"loss": 0.4069,
|
| 209 |
+
"step": 260
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.12340036563071298,
|
| 213 |
+
"grad_norm": 3.407360553741455,
|
| 214 |
+
"learning_rate": 0.00018765996343692872,
|
| 215 |
+
"loss": 0.2093,
|
| 216 |
+
"step": 270
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.12797074954296161,
|
| 220 |
+
"grad_norm": 9.035057067871094,
|
| 221 |
+
"learning_rate": 0.00018720292504570386,
|
| 222 |
+
"loss": 0.3598,
|
| 223 |
+
"step": 280
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 0.13254113345521024,
|
| 227 |
+
"grad_norm": 0.40073174238204956,
|
| 228 |
+
"learning_rate": 0.000186745886654479,
|
| 229 |
+
"loss": 0.2389,
|
| 230 |
+
"step": 290
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.13711151736745886,
|
| 234 |
+
"grad_norm": 0.32648196816444397,
|
| 235 |
+
"learning_rate": 0.00018628884826325413,
|
| 236 |
+
"loss": 0.2935,
|
| 237 |
+
"step": 300
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.13711151736745886,
|
| 241 |
+
"eval_accuracy": 0.9066666666666666,
|
| 242 |
+
"eval_loss": 0.31670910120010376,
|
| 243 |
+
"eval_runtime": 69.8802,
|
| 244 |
+
"eval_samples_per_second": 214.653,
|
| 245 |
+
"eval_steps_per_second": 13.423,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.1416819012797075,
|
| 250 |
+
"grad_norm": 5.271216869354248,
|
| 251 |
+
"learning_rate": 0.00018583180987202927,
|
| 252 |
+
"loss": 0.3934,
|
| 253 |
+
"step": 310
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.14625228519195613,
|
| 257 |
+
"grad_norm": 6.032111167907715,
|
| 258 |
+
"learning_rate": 0.0001853747714808044,
|
| 259 |
+
"loss": 0.2739,
|
| 260 |
+
"step": 320
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.15082266910420475,
|
| 264 |
+
"grad_norm": 1.9436883926391602,
|
| 265 |
+
"learning_rate": 0.00018491773308957954,
|
| 266 |
+
"loss": 0.2362,
|
| 267 |
+
"step": 330
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.15539305301645337,
|
| 271 |
+
"grad_norm": 7.990470886230469,
|
| 272 |
+
"learning_rate": 0.00018446069469835467,
|
| 273 |
+
"loss": 0.422,
|
| 274 |
+
"step": 340
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.15996343692870202,
|
| 278 |
+
"grad_norm": 3.6017348766326904,
|
| 279 |
+
"learning_rate": 0.0001840036563071298,
|
| 280 |
+
"loss": 0.3151,
|
| 281 |
+
"step": 350
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.16453382084095064,
|
| 285 |
+
"grad_norm": 4.97476053237915,
|
| 286 |
+
"learning_rate": 0.00018354661791590495,
|
| 287 |
+
"loss": 0.486,
|
| 288 |
+
"step": 360
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.16910420475319926,
|
| 292 |
+
"grad_norm": 5.418762683868408,
|
| 293 |
+
"learning_rate": 0.00018308957952468008,
|
| 294 |
+
"loss": 0.5847,
|
| 295 |
+
"step": 370
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.1736745886654479,
|
| 299 |
+
"grad_norm": 2.143413782119751,
|
| 300 |
+
"learning_rate": 0.00018263254113345522,
|
| 301 |
+
"loss": 0.3188,
|
| 302 |
+
"step": 380
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.17824497257769653,
|
| 306 |
+
"grad_norm": 5.15682315826416,
|
| 307 |
+
"learning_rate": 0.00018217550274223036,
|
| 308 |
+
"loss": 0.27,
|
| 309 |
+
"step": 390
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.18281535648994515,
|
| 313 |
+
"grad_norm": 4.638512134552002,
|
| 314 |
+
"learning_rate": 0.0001817184643510055,
|
| 315 |
+
"loss": 0.27,
|
| 316 |
+
"step": 400
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 0.18281535648994515,
|
| 320 |
+
"eval_accuracy": 0.8922,
|
| 321 |
+
"eval_loss": 0.35180962085723877,
|
| 322 |
+
"eval_runtime": 69.6193,
|
| 323 |
+
"eval_samples_per_second": 215.458,
|
| 324 |
+
"eval_steps_per_second": 13.473,
|
| 325 |
+
"step": 400
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.18738574040219377,
|
| 329 |
+
"grad_norm": 7.933419704437256,
|
| 330 |
+
"learning_rate": 0.00018126142595978063,
|
| 331 |
+
"loss": 0.2467,
|
| 332 |
+
"step": 410
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.19195612431444242,
|
| 336 |
+
"grad_norm": 2.749178886413574,
|
| 337 |
+
"learning_rate": 0.00018080438756855577,
|
| 338 |
+
"loss": 0.283,
|
| 339 |
+
"step": 420
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.19652650822669104,
|
| 343 |
+
"grad_norm": 4.662679672241211,
|
| 344 |
+
"learning_rate": 0.0001803473491773309,
|
| 345 |
+
"loss": 0.3666,
|
| 346 |
+
"step": 430
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.20109689213893966,
|
| 350 |
+
"grad_norm": 8.538412094116211,
|
| 351 |
+
"learning_rate": 0.00017989031078610604,
|
| 352 |
+
"loss": 0.3315,
|
| 353 |
+
"step": 440
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.2056672760511883,
|
| 357 |
+
"grad_norm": 7.226283550262451,
|
| 358 |
+
"learning_rate": 0.00017943327239488118,
|
| 359 |
+
"loss": 0.2998,
|
| 360 |
+
"step": 450
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.21023765996343693,
|
| 364 |
+
"grad_norm": 7.400086402893066,
|
| 365 |
+
"learning_rate": 0.0001789762340036563,
|
| 366 |
+
"loss": 0.3298,
|
| 367 |
+
"step": 460
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.21480804387568556,
|
| 371 |
+
"grad_norm": 2.4320435523986816,
|
| 372 |
+
"learning_rate": 0.00017851919561243145,
|
| 373 |
+
"loss": 0.3493,
|
| 374 |
+
"step": 470
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.21937842778793418,
|
| 378 |
+
"grad_norm": 2.891914129257202,
|
| 379 |
+
"learning_rate": 0.00017806215722120658,
|
| 380 |
+
"loss": 0.2929,
|
| 381 |
+
"step": 480
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.22394881170018283,
|
| 385 |
+
"grad_norm": 4.258346080780029,
|
| 386 |
+
"learning_rate": 0.00017760511882998172,
|
| 387 |
+
"loss": 0.3847,
|
| 388 |
+
"step": 490
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.22851919561243145,
|
| 392 |
+
"grad_norm": 3.8576903343200684,
|
| 393 |
+
"learning_rate": 0.00017714808043875686,
|
| 394 |
+
"loss": 0.3634,
|
| 395 |
+
"step": 500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.22851919561243145,
|
| 399 |
+
"eval_accuracy": 0.8953333333333333,
|
| 400 |
+
"eval_loss": 0.36601880192756653,
|
| 401 |
+
"eval_runtime": 70.2316,
|
| 402 |
+
"eval_samples_per_second": 213.579,
|
| 403 |
+
"eval_steps_per_second": 13.356,
|
| 404 |
+
"step": 500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.23308957952468007,
|
| 408 |
+
"grad_norm": 6.1643853187561035,
|
| 409 |
+
"learning_rate": 0.000176691042047532,
|
| 410 |
+
"loss": 0.3764,
|
| 411 |
+
"step": 510
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.2376599634369287,
|
| 415 |
+
"grad_norm": 0.7884778380393982,
|
| 416 |
+
"learning_rate": 0.00017623400365630713,
|
| 417 |
+
"loss": 0.2016,
|
| 418 |
+
"step": 520
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.24223034734917734,
|
| 422 |
+
"grad_norm": 3.4569339752197266,
|
| 423 |
+
"learning_rate": 0.00017577696526508227,
|
| 424 |
+
"loss": 0.3292,
|
| 425 |
+
"step": 530
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.24680073126142596,
|
| 429 |
+
"grad_norm": 0.9117717146873474,
|
| 430 |
+
"learning_rate": 0.0001753199268738574,
|
| 431 |
+
"loss": 0.1951,
|
| 432 |
+
"step": 540
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.2513711151736746,
|
| 436 |
+
"grad_norm": 11.449209213256836,
|
| 437 |
+
"learning_rate": 0.00017486288848263254,
|
| 438 |
+
"loss": 0.3511,
|
| 439 |
+
"step": 550
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 0.25594149908592323,
|
| 443 |
+
"grad_norm": 3.824899435043335,
|
| 444 |
+
"learning_rate": 0.00017440585009140768,
|
| 445 |
+
"loss": 0.2262,
|
| 446 |
+
"step": 560
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.26051188299817185,
|
| 450 |
+
"grad_norm": 7.360519886016846,
|
| 451 |
+
"learning_rate": 0.0001739488117001828,
|
| 452 |
+
"loss": 0.2491,
|
| 453 |
+
"step": 570
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.26508226691042047,
|
| 457 |
+
"grad_norm": 6.038581848144531,
|
| 458 |
+
"learning_rate": 0.00017349177330895795,
|
| 459 |
+
"loss": 0.2548,
|
| 460 |
+
"step": 580
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.2696526508226691,
|
| 464 |
+
"grad_norm": 3.3812975883483887,
|
| 465 |
+
"learning_rate": 0.0001730347349177331,
|
| 466 |
+
"loss": 0.4146,
|
| 467 |
+
"step": 590
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.2742230347349177,
|
| 471 |
+
"grad_norm": 0.8644607067108154,
|
| 472 |
+
"learning_rate": 0.00017257769652650825,
|
| 473 |
+
"loss": 0.2559,
|
| 474 |
+
"step": 600
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.2742230347349177,
|
| 478 |
+
"eval_accuracy": 0.8901333333333333,
|
| 479 |
+
"eval_loss": 0.3964242935180664,
|
| 480 |
+
"eval_runtime": 70.6896,
|
| 481 |
+
"eval_samples_per_second": 212.195,
|
| 482 |
+
"eval_steps_per_second": 13.269,
|
| 483 |
+
"step": 600
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.27879341864716634,
|
| 487 |
+
"grad_norm": 0.19820252060890198,
|
| 488 |
+
"learning_rate": 0.00017212065813528338,
|
| 489 |
+
"loss": 0.2866,
|
| 490 |
+
"step": 610
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 0.283363802559415,
|
| 494 |
+
"grad_norm": 4.149306774139404,
|
| 495 |
+
"learning_rate": 0.00017166361974405852,
|
| 496 |
+
"loss": 0.2711,
|
| 497 |
+
"step": 620
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.28793418647166363,
|
| 501 |
+
"grad_norm": 0.7687764167785645,
|
| 502 |
+
"learning_rate": 0.00017120658135283366,
|
| 503 |
+
"loss": 0.2035,
|
| 504 |
+
"step": 630
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.29250457038391225,
|
| 508 |
+
"grad_norm": 3.1452839374542236,
|
| 509 |
+
"learning_rate": 0.0001707495429616088,
|
| 510 |
+
"loss": 0.3511,
|
| 511 |
+
"step": 640
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 0.2970749542961609,
|
| 515 |
+
"grad_norm": 4.324541091918945,
|
| 516 |
+
"learning_rate": 0.00017029250457038393,
|
| 517 |
+
"loss": 0.2869,
|
| 518 |
+
"step": 650
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.3016453382084095,
|
| 522 |
+
"grad_norm": 0.06479712575674057,
|
| 523 |
+
"learning_rate": 0.00016983546617915907,
|
| 524 |
+
"loss": 0.1225,
|
| 525 |
+
"step": 660
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.3062157221206581,
|
| 529 |
+
"grad_norm": 1.430450677871704,
|
| 530 |
+
"learning_rate": 0.0001693784277879342,
|
| 531 |
+
"loss": 0.2164,
|
| 532 |
+
"step": 670
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 0.31078610603290674,
|
| 536 |
+
"grad_norm": 3.9292774200439453,
|
| 537 |
+
"learning_rate": 0.00016892138939670934,
|
| 538 |
+
"loss": 0.3039,
|
| 539 |
+
"step": 680
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 0.3153564899451554,
|
| 543 |
+
"grad_norm": 5.680319309234619,
|
| 544 |
+
"learning_rate": 0.00016846435100548448,
|
| 545 |
+
"loss": 0.2702,
|
| 546 |
+
"step": 690
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 0.31992687385740404,
|
| 550 |
+
"grad_norm": 0.42744606733322144,
|
| 551 |
+
"learning_rate": 0.0001680073126142596,
|
| 552 |
+
"loss": 0.197,
|
| 553 |
+
"step": 700
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 0.31992687385740404,
|
| 557 |
+
"eval_accuracy": 0.9252666666666667,
|
| 558 |
+
"eval_loss": 0.24806976318359375,
|
| 559 |
+
"eval_runtime": 70.2676,
|
| 560 |
+
"eval_samples_per_second": 213.47,
|
| 561 |
+
"eval_steps_per_second": 13.349,
|
| 562 |
+
"step": 700
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.32449725776965266,
|
| 566 |
+
"grad_norm": 5.1783447265625,
|
| 567 |
+
"learning_rate": 0.00016755027422303475,
|
| 568 |
+
"loss": 0.4019,
|
| 569 |
+
"step": 710
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.3290676416819013,
|
| 573 |
+
"grad_norm": 4.345694541931152,
|
| 574 |
+
"learning_rate": 0.00016709323583180986,
|
| 575 |
+
"loss": 0.191,
|
| 576 |
+
"step": 720
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.3336380255941499,
|
| 580 |
+
"grad_norm": 5.100950241088867,
|
| 581 |
+
"learning_rate": 0.000166636197440585,
|
| 582 |
+
"loss": 0.2976,
|
| 583 |
+
"step": 730
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.3382084095063985,
|
| 587 |
+
"grad_norm": 6.052485942840576,
|
| 588 |
+
"learning_rate": 0.00016617915904936016,
|
| 589 |
+
"loss": 0.3266,
|
| 590 |
+
"step": 740
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.34277879341864714,
|
| 594 |
+
"grad_norm": 2.7631659507751465,
|
| 595 |
+
"learning_rate": 0.0001657221206581353,
|
| 596 |
+
"loss": 0.3042,
|
| 597 |
+
"step": 750
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.3473491773308958,
|
| 601 |
+
"grad_norm": 3.2634143829345703,
|
| 602 |
+
"learning_rate": 0.00016526508226691043,
|
| 603 |
+
"loss": 0.1501,
|
| 604 |
+
"step": 760
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.35191956124314444,
|
| 608 |
+
"grad_norm": 5.064187049865723,
|
| 609 |
+
"learning_rate": 0.00016480804387568557,
|
| 610 |
+
"loss": 0.3669,
|
| 611 |
+
"step": 770
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.35648994515539306,
|
| 615 |
+
"grad_norm": 6.092987060546875,
|
| 616 |
+
"learning_rate": 0.0001643510054844607,
|
| 617 |
+
"loss": 0.2448,
|
| 618 |
+
"step": 780
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.3610603290676417,
|
| 622 |
+
"grad_norm": 4.66456413269043,
|
| 623 |
+
"learning_rate": 0.00016389396709323584,
|
| 624 |
+
"loss": 0.3027,
|
| 625 |
+
"step": 790
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.3656307129798903,
|
| 629 |
+
"grad_norm": 6.245533466339111,
|
| 630 |
+
"learning_rate": 0.00016343692870201098,
|
| 631 |
+
"loss": 0.2594,
|
| 632 |
+
"step": 800
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.3656307129798903,
|
| 636 |
+
"eval_accuracy": 0.923,
|
| 637 |
+
"eval_loss": 0.24855366349220276,
|
| 638 |
+
"eval_runtime": 70.4374,
|
| 639 |
+
"eval_samples_per_second": 212.955,
|
| 640 |
+
"eval_steps_per_second": 13.317,
|
| 641 |
+
"step": 800
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.3702010968921389,
|
| 645 |
+
"grad_norm": 0.3337598741054535,
|
| 646 |
+
"learning_rate": 0.0001629798903107861,
|
| 647 |
+
"loss": 0.2141,
|
| 648 |
+
"step": 810
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.37477148080438755,
|
| 652 |
+
"grad_norm": 3.9450998306274414,
|
| 653 |
+
"learning_rate": 0.00016252285191956125,
|
| 654 |
+
"loss": 0.2284,
|
| 655 |
+
"step": 820
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.3793418647166362,
|
| 659 |
+
"grad_norm": 4.09628438949585,
|
| 660 |
+
"learning_rate": 0.00016206581352833639,
|
| 661 |
+
"loss": 0.2501,
|
| 662 |
+
"step": 830
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.38391224862888484,
|
| 666 |
+
"grad_norm": 3.9277567863464355,
|
| 667 |
+
"learning_rate": 0.00016160877513711152,
|
| 668 |
+
"loss": 0.1436,
|
| 669 |
+
"step": 840
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.38848263254113347,
|
| 673 |
+
"grad_norm": 1.291695237159729,
|
| 674 |
+
"learning_rate": 0.00016115173674588666,
|
| 675 |
+
"loss": 0.3047,
|
| 676 |
+
"step": 850
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.3930530164533821,
|
| 680 |
+
"grad_norm": 3.195793867111206,
|
| 681 |
+
"learning_rate": 0.0001606946983546618,
|
| 682 |
+
"loss": 0.2684,
|
| 683 |
+
"step": 860
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.3976234003656307,
|
| 687 |
+
"grad_norm": 9.900090217590332,
|
| 688 |
+
"learning_rate": 0.00016023765996343693,
|
| 689 |
+
"loss": 0.2006,
|
| 690 |
+
"step": 870
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.40219378427787933,
|
| 694 |
+
"grad_norm": 7.173875331878662,
|
| 695 |
+
"learning_rate": 0.00015978062157221207,
|
| 696 |
+
"loss": 0.3154,
|
| 697 |
+
"step": 880
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.40676416819012795,
|
| 701 |
+
"grad_norm": 6.029903411865234,
|
| 702 |
+
"learning_rate": 0.0001593235831809872,
|
| 703 |
+
"loss": 0.3175,
|
| 704 |
+
"step": 890
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.4113345521023766,
|
| 708 |
+
"grad_norm": 6.628474235534668,
|
| 709 |
+
"learning_rate": 0.00015886654478976234,
|
| 710 |
+
"loss": 0.4545,
|
| 711 |
+
"step": 900
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.4113345521023766,
|
| 715 |
+
"eval_accuracy": 0.9,
|
| 716 |
+
"eval_loss": 0.327102929353714,
|
| 717 |
+
"eval_runtime": 70.3538,
|
| 718 |
+
"eval_samples_per_second": 213.208,
|
| 719 |
+
"eval_steps_per_second": 13.333,
|
| 720 |
+
"step": 900
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.41590493601462525,
|
| 724 |
+
"grad_norm": 4.372049808502197,
|
| 725 |
+
"learning_rate": 0.00015840950639853748,
|
| 726 |
+
"loss": 0.2577,
|
| 727 |
+
"step": 910
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.42047531992687387,
|
| 731 |
+
"grad_norm": 5.5920562744140625,
|
| 732 |
+
"learning_rate": 0.0001579524680073126,
|
| 733 |
+
"loss": 0.2078,
|
| 734 |
+
"step": 920
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.4250457038391225,
|
| 738 |
+
"grad_norm": 0.5012129545211792,
|
| 739 |
+
"learning_rate": 0.00015749542961608778,
|
| 740 |
+
"loss": 0.4423,
|
| 741 |
+
"step": 930
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.4296160877513711,
|
| 745 |
+
"grad_norm": 4.091048717498779,
|
| 746 |
+
"learning_rate": 0.0001570383912248629,
|
| 747 |
+
"loss": 0.2831,
|
| 748 |
+
"step": 940
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 0.43418647166361973,
|
| 752 |
+
"grad_norm": 3.677157163619995,
|
| 753 |
+
"learning_rate": 0.00015658135283363805,
|
| 754 |
+
"loss": 0.3525,
|
| 755 |
+
"step": 950
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 0.43875685557586835,
|
| 759 |
+
"grad_norm": 2.8373067378997803,
|
| 760 |
+
"learning_rate": 0.00015612431444241319,
|
| 761 |
+
"loss": 0.1623,
|
| 762 |
+
"step": 960
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.443327239488117,
|
| 766 |
+
"grad_norm": 7.7640156745910645,
|
| 767 |
+
"learning_rate": 0.00015566727605118832,
|
| 768 |
+
"loss": 0.2469,
|
| 769 |
+
"step": 970
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.44789762340036565,
|
| 773 |
+
"grad_norm": 0.34190094470977783,
|
| 774 |
+
"learning_rate": 0.00015521023765996346,
|
| 775 |
+
"loss": 0.2173,
|
| 776 |
+
"step": 980
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.4524680073126143,
|
| 780 |
+
"grad_norm": 0.42838719487190247,
|
| 781 |
+
"learning_rate": 0.0001547531992687386,
|
| 782 |
+
"loss": 0.1671,
|
| 783 |
+
"step": 990
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.4570383912248629,
|
| 787 |
+
"grad_norm": 1.7106258869171143,
|
| 788 |
+
"learning_rate": 0.00015429616087751373,
|
| 789 |
+
"loss": 0.1243,
|
| 790 |
+
"step": 1000
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 0.4570383912248629,
|
| 794 |
+
"eval_accuracy": 0.9268666666666666,
|
| 795 |
+
"eval_loss": 0.24482281506061554,
|
| 796 |
+
"eval_runtime": 70.0663,
|
| 797 |
+
"eval_samples_per_second": 214.083,
|
| 798 |
+
"eval_steps_per_second": 13.387,
|
| 799 |
+
"step": 1000
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 0.4616087751371115,
|
| 803 |
+
"grad_norm": 4.664043426513672,
|
| 804 |
+
"learning_rate": 0.00015383912248628884,
|
| 805 |
+
"loss": 0.2915,
|
| 806 |
+
"step": 1010
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.46617915904936014,
|
| 810 |
+
"grad_norm": 6.955049991607666,
|
| 811 |
+
"learning_rate": 0.00015338208409506398,
|
| 812 |
+
"loss": 0.2288,
|
| 813 |
+
"step": 1020
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.47074954296160876,
|
| 817 |
+
"grad_norm": 6.842510223388672,
|
| 818 |
+
"learning_rate": 0.00015292504570383911,
|
| 819 |
+
"loss": 0.4217,
|
| 820 |
+
"step": 1030
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 0.4753199268738574,
|
| 824 |
+
"grad_norm": 2.6210012435913086,
|
| 825 |
+
"learning_rate": 0.00015246800731261425,
|
| 826 |
+
"loss": 0.3455,
|
| 827 |
+
"step": 1040
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 0.47989031078610606,
|
| 831 |
+
"grad_norm": 0.2048071324825287,
|
| 832 |
+
"learning_rate": 0.0001520109689213894,
|
| 833 |
+
"loss": 0.1599,
|
| 834 |
+
"step": 1050
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 0.4844606946983547,
|
| 838 |
+
"grad_norm": 2.204970121383667,
|
| 839 |
+
"learning_rate": 0.00015155393053016452,
|
| 840 |
+
"loss": 0.2539,
|
| 841 |
+
"step": 1060
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.4890310786106033,
|
| 845 |
+
"grad_norm": 4.208988189697266,
|
| 846 |
+
"learning_rate": 0.00015109689213893966,
|
| 847 |
+
"loss": 0.1719,
|
| 848 |
+
"step": 1070
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.4936014625228519,
|
| 852 |
+
"grad_norm": 1.6201022863388062,
|
| 853 |
+
"learning_rate": 0.00015063985374771482,
|
| 854 |
+
"loss": 0.103,
|
| 855 |
+
"step": 1080
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.49817184643510054,
|
| 859 |
+
"grad_norm": 2.211272954940796,
|
| 860 |
+
"learning_rate": 0.00015018281535648996,
|
| 861 |
+
"loss": 0.1956,
|
| 862 |
+
"step": 1090
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 0.5027422303473492,
|
| 866 |
+
"grad_norm": 7.531051158905029,
|
| 867 |
+
"learning_rate": 0.0001497257769652651,
|
| 868 |
+
"loss": 0.3593,
|
| 869 |
+
"step": 1100
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 0.5027422303473492,
|
| 873 |
+
"eval_accuracy": 0.9354,
|
| 874 |
+
"eval_loss": 0.2118164449930191,
|
| 875 |
+
"eval_runtime": 70.235,
|
| 876 |
+
"eval_samples_per_second": 213.569,
|
| 877 |
+
"eval_steps_per_second": 13.355,
|
| 878 |
+
"step": 1100
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.5073126142595978,
|
| 882 |
+
"grad_norm": 3.182407855987549,
|
| 883 |
+
"learning_rate": 0.00014926873857404023,
|
| 884 |
+
"loss": 0.3023,
|
| 885 |
+
"step": 1110
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.5118829981718465,
|
| 889 |
+
"grad_norm": 3.975743293762207,
|
| 890 |
+
"learning_rate": 0.00014881170018281537,
|
| 891 |
+
"loss": 0.2358,
|
| 892 |
+
"step": 1120
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.5164533820840951,
|
| 896 |
+
"grad_norm": 4.823325157165527,
|
| 897 |
+
"learning_rate": 0.0001483546617915905,
|
| 898 |
+
"loss": 0.2031,
|
| 899 |
+
"step": 1130
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.5210237659963437,
|
| 903 |
+
"grad_norm": 0.8731510043144226,
|
| 904 |
+
"learning_rate": 0.00014789762340036564,
|
| 905 |
+
"loss": 0.2254,
|
| 906 |
+
"step": 1140
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.5255941499085923,
|
| 910 |
+
"grad_norm": 1.2150533199310303,
|
| 911 |
+
"learning_rate": 0.00014744058500914078,
|
| 912 |
+
"loss": 0.312,
|
| 913 |
+
"step": 1150
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.5301645338208409,
|
| 917 |
+
"grad_norm": 1.9706271886825562,
|
| 918 |
+
"learning_rate": 0.00014698354661791591,
|
| 919 |
+
"loss": 0.1619,
|
| 920 |
+
"step": 1160
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.5347349177330896,
|
| 924 |
+
"grad_norm": 0.36833953857421875,
|
| 925 |
+
"learning_rate": 0.00014652650822669105,
|
| 926 |
+
"loss": 0.1976,
|
| 927 |
+
"step": 1170
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.5393053016453382,
|
| 931 |
+
"grad_norm": 3.9230244159698486,
|
| 932 |
+
"learning_rate": 0.0001460694698354662,
|
| 933 |
+
"loss": 0.2928,
|
| 934 |
+
"step": 1180
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.5438756855575868,
|
| 938 |
+
"grad_norm": 4.378619194030762,
|
| 939 |
+
"learning_rate": 0.00014561243144424132,
|
| 940 |
+
"loss": 0.2008,
|
| 941 |
+
"step": 1190
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.5484460694698354,
|
| 945 |
+
"grad_norm": 0.8093172907829285,
|
| 946 |
+
"learning_rate": 0.00014515539305301646,
|
| 947 |
+
"loss": 0.1375,
|
| 948 |
+
"step": 1200
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.5484460694698354,
|
| 952 |
+
"eval_accuracy": 0.9348666666666666,
|
| 953 |
+
"eval_loss": 0.22052045166492462,
|
| 954 |
+
"eval_runtime": 70.7854,
|
| 955 |
+
"eval_samples_per_second": 211.908,
|
| 956 |
+
"eval_steps_per_second": 13.251,
|
| 957 |
+
"step": 1200
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.553016453382084,
|
| 961 |
+
"grad_norm": 8.147080421447754,
|
| 962 |
+
"learning_rate": 0.0001446983546617916,
|
| 963 |
+
"loss": 0.1968,
|
| 964 |
+
"step": 1210
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.5575868372943327,
|
| 968 |
+
"grad_norm": 4.382139682769775,
|
| 969 |
+
"learning_rate": 0.00014424131627056673,
|
| 970 |
+
"loss": 0.2315,
|
| 971 |
+
"step": 1220
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.5621572212065814,
|
| 975 |
+
"grad_norm": 2.122556209564209,
|
| 976 |
+
"learning_rate": 0.00014378427787934187,
|
| 977 |
+
"loss": 0.1646,
|
| 978 |
+
"step": 1230
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.56672760511883,
|
| 982 |
+
"grad_norm": 4.841265678405762,
|
| 983 |
+
"learning_rate": 0.000143327239488117,
|
| 984 |
+
"loss": 0.2223,
|
| 985 |
+
"step": 1240
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.5712979890310786,
|
| 989 |
+
"grad_norm": 0.27766382694244385,
|
| 990 |
+
"learning_rate": 0.00014287020109689214,
|
| 991 |
+
"loss": 0.2387,
|
| 992 |
+
"step": 1250
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.5758683729433273,
|
| 996 |
+
"grad_norm": 1.7641736268997192,
|
| 997 |
+
"learning_rate": 0.00014241316270566728,
|
| 998 |
+
"loss": 0.1726,
|
| 999 |
+
"step": 1260
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.5804387568555759,
|
| 1003 |
+
"grad_norm": 6.29493522644043,
|
| 1004 |
+
"learning_rate": 0.00014195612431444244,
|
| 1005 |
+
"loss": 0.4514,
|
| 1006 |
+
"step": 1270
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.5850091407678245,
|
| 1010 |
+
"grad_norm": 0.2854464054107666,
|
| 1011 |
+
"learning_rate": 0.00014149908592321758,
|
| 1012 |
+
"loss": 0.1004,
|
| 1013 |
+
"step": 1280
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.5895795246800731,
|
| 1017 |
+
"grad_norm": 5.928961753845215,
|
| 1018 |
+
"learning_rate": 0.00014104204753199271,
|
| 1019 |
+
"loss": 0.2575,
|
| 1020 |
+
"step": 1290
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.5941499085923218,
|
| 1024 |
+
"grad_norm": 0.3155369758605957,
|
| 1025 |
+
"learning_rate": 0.00014058500914076782,
|
| 1026 |
+
"loss": 0.1521,
|
| 1027 |
+
"step": 1300
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 0.5941499085923218,
|
| 1031 |
+
"eval_accuracy": 0.9376,
|
| 1032 |
+
"eval_loss": 0.200880229473114,
|
| 1033 |
+
"eval_runtime": 70.0018,
|
| 1034 |
+
"eval_samples_per_second": 214.28,
|
| 1035 |
+
"eval_steps_per_second": 13.4,
|
| 1036 |
+
"step": 1300
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 0.5987202925045704,
|
| 1040 |
+
"grad_norm": 5.347257614135742,
|
| 1041 |
+
"learning_rate": 0.00014012797074954296,
|
| 1042 |
+
"loss": 0.178,
|
| 1043 |
+
"step": 1310
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 0.603290676416819,
|
| 1047 |
+
"grad_norm": 8.616045951843262,
|
| 1048 |
+
"learning_rate": 0.0001396709323583181,
|
| 1049 |
+
"loss": 0.2389,
|
| 1050 |
+
"step": 1320
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 0.6078610603290676,
|
| 1054 |
+
"grad_norm": 3.013582229614258,
|
| 1055 |
+
"learning_rate": 0.00013921389396709323,
|
| 1056 |
+
"loss": 0.1769,
|
| 1057 |
+
"step": 1330
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 0.6124314442413162,
|
| 1061 |
+
"grad_norm": 5.738295078277588,
|
| 1062 |
+
"learning_rate": 0.00013875685557586837,
|
| 1063 |
+
"loss": 0.1847,
|
| 1064 |
+
"step": 1340
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.6170018281535649,
|
| 1068 |
+
"grad_norm": 0.47330769896507263,
|
| 1069 |
+
"learning_rate": 0.0001382998171846435,
|
| 1070 |
+
"loss": 0.1451,
|
| 1071 |
+
"step": 1350
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.6215722120658135,
|
| 1075 |
+
"grad_norm": 4.576214790344238,
|
| 1076 |
+
"learning_rate": 0.00013784277879341864,
|
| 1077 |
+
"loss": 0.2656,
|
| 1078 |
+
"step": 1360
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 0.6261425959780622,
|
| 1082 |
+
"grad_norm": 2.6387856006622314,
|
| 1083 |
+
"learning_rate": 0.00013738574040219378,
|
| 1084 |
+
"loss": 0.1181,
|
| 1085 |
+
"step": 1370
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 0.6307129798903108,
|
| 1089 |
+
"grad_norm": 0.5952144265174866,
|
| 1090 |
+
"learning_rate": 0.00013692870201096892,
|
| 1091 |
+
"loss": 0.1286,
|
| 1092 |
+
"step": 1380
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 0.6352833638025595,
|
| 1096 |
+
"grad_norm": 5.721958160400391,
|
| 1097 |
+
"learning_rate": 0.00013647166361974405,
|
| 1098 |
+
"loss": 0.2129,
|
| 1099 |
+
"step": 1390
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 0.6398537477148081,
|
| 1103 |
+
"grad_norm": 1.397820234298706,
|
| 1104 |
+
"learning_rate": 0.0001360146252285192,
|
| 1105 |
+
"loss": 0.1237,
|
| 1106 |
+
"step": 1400
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.6398537477148081,
|
| 1110 |
+
"eval_accuracy": 0.9444666666666667,
|
| 1111 |
+
"eval_loss": 0.18025201559066772,
|
| 1112 |
+
"eval_runtime": 69.8721,
|
| 1113 |
+
"eval_samples_per_second": 214.678,
|
| 1114 |
+
"eval_steps_per_second": 13.425,
|
| 1115 |
+
"step": 1400
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.6444241316270567,
|
| 1119 |
+
"grad_norm": 2.166053533554077,
|
| 1120 |
+
"learning_rate": 0.00013555758683729432,
|
| 1121 |
+
"loss": 0.2256,
|
| 1122 |
+
"step": 1410
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.6489945155393053,
|
| 1126 |
+
"grad_norm": 1.5744013786315918,
|
| 1127 |
+
"learning_rate": 0.0001351005484460695,
|
| 1128 |
+
"loss": 0.085,
|
| 1129 |
+
"step": 1420
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.6535648994515539,
|
| 1133 |
+
"grad_norm": 5.0064167976379395,
|
| 1134 |
+
"learning_rate": 0.00013464351005484462,
|
| 1135 |
+
"loss": 0.1947,
|
| 1136 |
+
"step": 1430
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.6581352833638026,
|
| 1140 |
+
"grad_norm": 6.537513732910156,
|
| 1141 |
+
"learning_rate": 0.00013418647166361976,
|
| 1142 |
+
"loss": 0.095,
|
| 1143 |
+
"step": 1440
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.6627056672760512,
|
| 1147 |
+
"grad_norm": 10.165826797485352,
|
| 1148 |
+
"learning_rate": 0.0001337294332723949,
|
| 1149 |
+
"loss": 0.2335,
|
| 1150 |
+
"step": 1450
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.6672760511882998,
|
| 1154 |
+
"grad_norm": 2.5049922466278076,
|
| 1155 |
+
"learning_rate": 0.00013327239488117003,
|
| 1156 |
+
"loss": 0.1958,
|
| 1157 |
+
"step": 1460
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.6718464351005484,
|
| 1161 |
+
"grad_norm": 6.945699214935303,
|
| 1162 |
+
"learning_rate": 0.00013281535648994517,
|
| 1163 |
+
"loss": 0.141,
|
| 1164 |
+
"step": 1470
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.676416819012797,
|
| 1168 |
+
"grad_norm": 0.31921499967575073,
|
| 1169 |
+
"learning_rate": 0.0001323583180987203,
|
| 1170 |
+
"loss": 0.1275,
|
| 1171 |
+
"step": 1480
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.6809872029250457,
|
| 1175 |
+
"grad_norm": 6.356455326080322,
|
| 1176 |
+
"learning_rate": 0.00013190127970749544,
|
| 1177 |
+
"loss": 0.2715,
|
| 1178 |
+
"step": 1490
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.6855575868372943,
|
| 1182 |
+
"grad_norm": 0.11626709252595901,
|
| 1183 |
+
"learning_rate": 0.00013144424131627058,
|
| 1184 |
+
"loss": 0.2214,
|
| 1185 |
+
"step": 1500
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.6855575868372943,
|
| 1189 |
+
"eval_accuracy": 0.9394666666666667,
|
| 1190 |
+
"eval_loss": 0.20262379944324493,
|
| 1191 |
+
"eval_runtime": 70.7416,
|
| 1192 |
+
"eval_samples_per_second": 212.039,
|
| 1193 |
+
"eval_steps_per_second": 13.26,
|
| 1194 |
+
"step": 1500
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 0.6901279707495429,
|
| 1198 |
+
"grad_norm": 10.25921630859375,
|
| 1199 |
+
"learning_rate": 0.00013098720292504572,
|
| 1200 |
+
"loss": 0.3417,
|
| 1201 |
+
"step": 1510
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 0.6946983546617916,
|
| 1205 |
+
"grad_norm": 0.09503093361854553,
|
| 1206 |
+
"learning_rate": 0.00013053016453382085,
|
| 1207 |
+
"loss": 0.1734,
|
| 1208 |
+
"step": 1520
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.6992687385740403,
|
| 1212 |
+
"grad_norm": 4.171358108520508,
|
| 1213 |
+
"learning_rate": 0.000130073126142596,
|
| 1214 |
+
"loss": 0.1565,
|
| 1215 |
+
"step": 1530
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.7038391224862889,
|
| 1219 |
+
"grad_norm": 1.6032589673995972,
|
| 1220 |
+
"learning_rate": 0.00012961608775137112,
|
| 1221 |
+
"loss": 0.279,
|
| 1222 |
+
"step": 1540
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.7084095063985375,
|
| 1226 |
+
"grad_norm": 4.150899410247803,
|
| 1227 |
+
"learning_rate": 0.00012915904936014626,
|
| 1228 |
+
"loss": 0.261,
|
| 1229 |
+
"step": 1550
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.7129798903107861,
|
| 1233 |
+
"grad_norm": 3.3375895023345947,
|
| 1234 |
+
"learning_rate": 0.0001287020109689214,
|
| 1235 |
+
"loss": 0.2531,
|
| 1236 |
+
"step": 1560
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 0.7175502742230347,
|
| 1240 |
+
"grad_norm": 6.167548179626465,
|
| 1241 |
+
"learning_rate": 0.00012824497257769653,
|
| 1242 |
+
"loss": 0.2602,
|
| 1243 |
+
"step": 1570
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 0.7221206581352834,
|
| 1247 |
+
"grad_norm": 2.9471616744995117,
|
| 1248 |
+
"learning_rate": 0.00012778793418647167,
|
| 1249 |
+
"loss": 0.2944,
|
| 1250 |
+
"step": 1580
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.726691042047532,
|
| 1254 |
+
"grad_norm": 1.4548298120498657,
|
| 1255 |
+
"learning_rate": 0.0001273308957952468,
|
| 1256 |
+
"loss": 0.0746,
|
| 1257 |
+
"step": 1590
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.7312614259597806,
|
| 1261 |
+
"grad_norm": 6.375178337097168,
|
| 1262 |
+
"learning_rate": 0.00012687385740402194,
|
| 1263 |
+
"loss": 0.1324,
|
| 1264 |
+
"step": 1600
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 0.7312614259597806,
|
| 1268 |
+
"eval_accuracy": 0.9493333333333334,
|
| 1269 |
+
"eval_loss": 0.16354645788669586,
|
| 1270 |
+
"eval_runtime": 70.3467,
|
| 1271 |
+
"eval_samples_per_second": 213.23,
|
| 1272 |
+
"eval_steps_per_second": 13.334,
|
| 1273 |
+
"step": 1600
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.7358318098720292,
|
| 1277 |
+
"grad_norm": 6.716129302978516,
|
| 1278 |
+
"learning_rate": 0.00012641681901279708,
|
| 1279 |
+
"loss": 0.2021,
|
| 1280 |
+
"step": 1610
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.7404021937842779,
|
| 1284 |
+
"grad_norm": 0.047931790351867676,
|
| 1285 |
+
"learning_rate": 0.00012595978062157222,
|
| 1286 |
+
"loss": 0.1456,
|
| 1287 |
+
"step": 1620
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.7449725776965265,
|
| 1291 |
+
"grad_norm": 0.08262607455253601,
|
| 1292 |
+
"learning_rate": 0.00012550274223034735,
|
| 1293 |
+
"loss": 0.1167,
|
| 1294 |
+
"step": 1630
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.7495429616087751,
|
| 1298 |
+
"grad_norm": 0.21918249130249023,
|
| 1299 |
+
"learning_rate": 0.0001250457038391225,
|
| 1300 |
+
"loss": 0.1558,
|
| 1301 |
+
"step": 1640
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.7541133455210237,
|
| 1305 |
+
"grad_norm": 3.03836989402771,
|
| 1306 |
+
"learning_rate": 0.00012458866544789763,
|
| 1307 |
+
"loss": 0.1506,
|
| 1308 |
+
"step": 1650
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.7586837294332724,
|
| 1312 |
+
"grad_norm": 9.57582950592041,
|
| 1313 |
+
"learning_rate": 0.00012413162705667276,
|
| 1314 |
+
"loss": 0.2193,
|
| 1315 |
+
"step": 1660
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.7632541133455211,
|
| 1319 |
+
"grad_norm": 0.2596251964569092,
|
| 1320 |
+
"learning_rate": 0.0001236745886654479,
|
| 1321 |
+
"loss": 0.1041,
|
| 1322 |
+
"step": 1670
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.7678244972577697,
|
| 1326 |
+
"grad_norm": 11.114243507385254,
|
| 1327 |
+
"learning_rate": 0.00012321755027422303,
|
| 1328 |
+
"loss": 0.1528,
|
| 1329 |
+
"step": 1680
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.7723948811700183,
|
| 1333 |
+
"grad_norm": 5.171649932861328,
|
| 1334 |
+
"learning_rate": 0.00012276051188299817,
|
| 1335 |
+
"loss": 0.1856,
|
| 1336 |
+
"step": 1690
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.7769652650822669,
|
| 1340 |
+
"grad_norm": 4.471177577972412,
|
| 1341 |
+
"learning_rate": 0.0001223034734917733,
|
| 1342 |
+
"loss": 0.1864,
|
| 1343 |
+
"step": 1700
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.7769652650822669,
|
| 1347 |
+
"eval_accuracy": 0.9493333333333334,
|
| 1348 |
+
"eval_loss": 0.16721826791763306,
|
| 1349 |
+
"eval_runtime": 70.5466,
|
| 1350 |
+
"eval_samples_per_second": 212.625,
|
| 1351 |
+
"eval_steps_per_second": 13.296,
|
| 1352 |
+
"step": 1700
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.7815356489945156,
|
| 1356 |
+
"grad_norm": 4.236210346221924,
|
| 1357 |
+
"learning_rate": 0.00012184643510054844,
|
| 1358 |
+
"loss": 0.1898,
|
| 1359 |
+
"step": 1710
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.7861060329067642,
|
| 1363 |
+
"grad_norm": 3.0980677604675293,
|
| 1364 |
+
"learning_rate": 0.0001213893967093236,
|
| 1365 |
+
"loss": 0.1925,
|
| 1366 |
+
"step": 1720
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 0.7906764168190128,
|
| 1370 |
+
"grad_norm": 2.8537585735321045,
|
| 1371 |
+
"learning_rate": 0.00012093235831809873,
|
| 1372 |
+
"loss": 0.1141,
|
| 1373 |
+
"step": 1730
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 0.7952468007312614,
|
| 1377 |
+
"grad_norm": 3.942676305770874,
|
| 1378 |
+
"learning_rate": 0.00012047531992687387,
|
| 1379 |
+
"loss": 0.2004,
|
| 1380 |
+
"step": 1740
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 0.79981718464351,
|
| 1384 |
+
"grad_norm": 2.8048367500305176,
|
| 1385 |
+
"learning_rate": 0.000120018281535649,
|
| 1386 |
+
"loss": 0.2308,
|
| 1387 |
+
"step": 1750
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 0.8043875685557587,
|
| 1391 |
+
"grad_norm": 2.2184743881225586,
|
| 1392 |
+
"learning_rate": 0.00011956124314442414,
|
| 1393 |
+
"loss": 0.1032,
|
| 1394 |
+
"step": 1760
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.8089579524680073,
|
| 1398 |
+
"grad_norm": 12.887242317199707,
|
| 1399 |
+
"learning_rate": 0.00011910420475319928,
|
| 1400 |
+
"loss": 0.2534,
|
| 1401 |
+
"step": 1770
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.8135283363802559,
|
| 1405 |
+
"grad_norm": 0.6864253878593445,
|
| 1406 |
+
"learning_rate": 0.00011864716636197441,
|
| 1407 |
+
"loss": 0.1879,
|
| 1408 |
+
"step": 1780
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 0.8180987202925045,
|
| 1412 |
+
"grad_norm": 3.042908191680908,
|
| 1413 |
+
"learning_rate": 0.00011819012797074955,
|
| 1414 |
+
"loss": 0.2072,
|
| 1415 |
+
"step": 1790
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 0.8226691042047533,
|
| 1419 |
+
"grad_norm": 0.42097190022468567,
|
| 1420 |
+
"learning_rate": 0.00011773308957952469,
|
| 1421 |
+
"loss": 0.128,
|
| 1422 |
+
"step": 1800
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 0.8226691042047533,
|
| 1426 |
+
"eval_accuracy": 0.9409333333333333,
|
| 1427 |
+
"eval_loss": 0.2014516144990921,
|
| 1428 |
+
"eval_runtime": 71.2106,
|
| 1429 |
+
"eval_samples_per_second": 210.643,
|
| 1430 |
+
"eval_steps_per_second": 13.172,
|
| 1431 |
+
"step": 1800
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.8272394881170019,
|
| 1435 |
+
"grad_norm": 5.429555892944336,
|
| 1436 |
+
"learning_rate": 0.00011727605118829984,
|
| 1437 |
+
"loss": 0.099,
|
| 1438 |
+
"step": 1810
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.8318098720292505,
|
| 1442 |
+
"grad_norm": 6.33099889755249,
|
| 1443 |
+
"learning_rate": 0.00011681901279707497,
|
| 1444 |
+
"loss": 0.1546,
|
| 1445 |
+
"step": 1820
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.8363802559414991,
|
| 1449 |
+
"grad_norm": 5.382350921630859,
|
| 1450 |
+
"learning_rate": 0.00011636197440585011,
|
| 1451 |
+
"loss": 0.2645,
|
| 1452 |
+
"step": 1830
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.8409506398537477,
|
| 1456 |
+
"grad_norm": 2.201632022857666,
|
| 1457 |
+
"learning_rate": 0.00011590493601462524,
|
| 1458 |
+
"loss": 0.2078,
|
| 1459 |
+
"step": 1840
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.8455210237659964,
|
| 1463 |
+
"grad_norm": 7.094603538513184,
|
| 1464 |
+
"learning_rate": 0.00011544789762340038,
|
| 1465 |
+
"loss": 0.3097,
|
| 1466 |
+
"step": 1850
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.850091407678245,
|
| 1470 |
+
"grad_norm": 0.3324418067932129,
|
| 1471 |
+
"learning_rate": 0.00011499085923217552,
|
| 1472 |
+
"loss": 0.1623,
|
| 1473 |
+
"step": 1860
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.8546617915904936,
|
| 1477 |
+
"grad_norm": 0.12678645551204681,
|
| 1478 |
+
"learning_rate": 0.00011453382084095065,
|
| 1479 |
+
"loss": 0.1275,
|
| 1480 |
+
"step": 1870
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.8592321755027422,
|
| 1484 |
+
"grad_norm": 2.2118465900421143,
|
| 1485 |
+
"learning_rate": 0.00011407678244972578,
|
| 1486 |
+
"loss": 0.1834,
|
| 1487 |
+
"step": 1880
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.8638025594149908,
|
| 1491 |
+
"grad_norm": 11.703352928161621,
|
| 1492 |
+
"learning_rate": 0.00011361974405850091,
|
| 1493 |
+
"loss": 0.165,
|
| 1494 |
+
"step": 1890
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.8683729433272395,
|
| 1498 |
+
"grad_norm": 5.437178611755371,
|
| 1499 |
+
"learning_rate": 0.00011316270566727605,
|
| 1500 |
+
"loss": 0.121,
|
| 1501 |
+
"step": 1900
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.8683729433272395,
|
| 1505 |
+
"eval_accuracy": 0.9451333333333334,
|
| 1506 |
+
"eval_loss": 0.17528271675109863,
|
| 1507 |
+
"eval_runtime": 70.5119,
|
| 1508 |
+
"eval_samples_per_second": 212.73,
|
| 1509 |
+
"eval_steps_per_second": 13.303,
|
| 1510 |
+
"step": 1900
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 0.8729433272394881,
|
| 1514 |
+
"grad_norm": 2.402738332748413,
|
| 1515 |
+
"learning_rate": 0.00011270566727605119,
|
| 1516 |
+
"loss": 0.1896,
|
| 1517 |
+
"step": 1910
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 0.8775137111517367,
|
| 1521 |
+
"grad_norm": 0.6686537861824036,
|
| 1522 |
+
"learning_rate": 0.00011224862888482632,
|
| 1523 |
+
"loss": 0.1535,
|
| 1524 |
+
"step": 1920
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 0.8820840950639853,
|
| 1528 |
+
"grad_norm": 3.365333080291748,
|
| 1529 |
+
"learning_rate": 0.00011179159049360146,
|
| 1530 |
+
"loss": 0.1135,
|
| 1531 |
+
"step": 1930
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 0.886654478976234,
|
| 1535 |
+
"grad_norm": 3.6913065910339355,
|
| 1536 |
+
"learning_rate": 0.0001113345521023766,
|
| 1537 |
+
"loss": 0.1391,
|
| 1538 |
+
"step": 1940
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.8912248628884827,
|
| 1542 |
+
"grad_norm": 7.079347133636475,
|
| 1543 |
+
"learning_rate": 0.00011087751371115173,
|
| 1544 |
+
"loss": 0.1528,
|
| 1545 |
+
"step": 1950
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.8957952468007313,
|
| 1549 |
+
"grad_norm": 2.3773577213287354,
|
| 1550 |
+
"learning_rate": 0.00011042047531992688,
|
| 1551 |
+
"loss": 0.1345,
|
| 1552 |
+
"step": 1960
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 0.9003656307129799,
|
| 1556 |
+
"grad_norm": 3.536985158920288,
|
| 1557 |
+
"learning_rate": 0.00010996343692870202,
|
| 1558 |
+
"loss": 0.1762,
|
| 1559 |
+
"step": 1970
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 0.9049360146252285,
|
| 1563 |
+
"grad_norm": 0.4026005268096924,
|
| 1564 |
+
"learning_rate": 0.00010950639853747715,
|
| 1565 |
+
"loss": 0.2182,
|
| 1566 |
+
"step": 1980
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 0.9095063985374772,
|
| 1570 |
+
"grad_norm": 2.521723747253418,
|
| 1571 |
+
"learning_rate": 0.00010904936014625229,
|
| 1572 |
+
"loss": 0.1531,
|
| 1573 |
+
"step": 1990
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 0.9140767824497258,
|
| 1577 |
+
"grad_norm": 0.42068588733673096,
|
| 1578 |
+
"learning_rate": 0.00010859232175502743,
|
| 1579 |
+
"loss": 0.1918,
|
| 1580 |
+
"step": 2000
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.9140767824497258,
|
| 1584 |
+
"eval_accuracy": 0.9588,
|
| 1585 |
+
"eval_loss": 0.13700534403324127,
|
| 1586 |
+
"eval_runtime": 70.8331,
|
| 1587 |
+
"eval_samples_per_second": 211.766,
|
| 1588 |
+
"eval_steps_per_second": 13.242,
|
| 1589 |
+
"step": 2000
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 0.9186471663619744,
|
| 1593 |
+
"grad_norm": 0.27603089809417725,
|
| 1594 |
+
"learning_rate": 0.00010813528336380256,
|
| 1595 |
+
"loss": 0.172,
|
| 1596 |
+
"step": 2010
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 0.923217550274223,
|
| 1600 |
+
"grad_norm": 1.674926996231079,
|
| 1601 |
+
"learning_rate": 0.0001076782449725777,
|
| 1602 |
+
"loss": 0.1129,
|
| 1603 |
+
"step": 2020
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 0.9277879341864717,
|
| 1607 |
+
"grad_norm": 0.18493302166461945,
|
| 1608 |
+
"learning_rate": 0.00010722120658135284,
|
| 1609 |
+
"loss": 0.0895,
|
| 1610 |
+
"step": 2030
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 0.9323583180987203,
|
| 1614 |
+
"grad_norm": 4.929255485534668,
|
| 1615 |
+
"learning_rate": 0.00010676416819012797,
|
| 1616 |
+
"loss": 0.0942,
|
| 1617 |
+
"step": 2040
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 0.9369287020109689,
|
| 1621 |
+
"grad_norm": 2.887568950653076,
|
| 1622 |
+
"learning_rate": 0.00010630712979890312,
|
| 1623 |
+
"loss": 0.1769,
|
| 1624 |
+
"step": 2050
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 0.9414990859232175,
|
| 1628 |
+
"grad_norm": 2.5681095123291016,
|
| 1629 |
+
"learning_rate": 0.00010585009140767826,
|
| 1630 |
+
"loss": 0.1939,
|
| 1631 |
+
"step": 2060
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 0.9460694698354661,
|
| 1635 |
+
"grad_norm": 9.683144569396973,
|
| 1636 |
+
"learning_rate": 0.0001053930530164534,
|
| 1637 |
+
"loss": 0.212,
|
| 1638 |
+
"step": 2070
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 0.9506398537477148,
|
| 1642 |
+
"grad_norm": 0.2908894419670105,
|
| 1643 |
+
"learning_rate": 0.00010493601462522853,
|
| 1644 |
+
"loss": 0.1635,
|
| 1645 |
+
"step": 2080
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 0.9552102376599635,
|
| 1649 |
+
"grad_norm": 5.623505115509033,
|
| 1650 |
+
"learning_rate": 0.00010447897623400367,
|
| 1651 |
+
"loss": 0.0803,
|
| 1652 |
+
"step": 2090
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 0.9597806215722121,
|
| 1656 |
+
"grad_norm": 1.6659337282180786,
|
| 1657 |
+
"learning_rate": 0.0001040219378427788,
|
| 1658 |
+
"loss": 0.1658,
|
| 1659 |
+
"step": 2100
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 0.9597806215722121,
|
| 1663 |
+
"eval_accuracy": 0.9534666666666667,
|
| 1664 |
+
"eval_loss": 0.15428349375724792,
|
| 1665 |
+
"eval_runtime": 70.1951,
|
| 1666 |
+
"eval_samples_per_second": 213.69,
|
| 1667 |
+
"eval_steps_per_second": 13.363,
|
| 1668 |
+
"step": 2100
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.9643510054844607,
|
| 1672 |
+
"grad_norm": 3.0254101753234863,
|
| 1673 |
+
"learning_rate": 0.00010356489945155394,
|
| 1674 |
+
"loss": 0.2228,
|
| 1675 |
+
"step": 2110
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.9689213893967094,
|
| 1679 |
+
"grad_norm": 4.310238838195801,
|
| 1680 |
+
"learning_rate": 0.00010310786106032908,
|
| 1681 |
+
"loss": 0.1241,
|
| 1682 |
+
"step": 2120
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.973491773308958,
|
| 1686 |
+
"grad_norm": 3.578296422958374,
|
| 1687 |
+
"learning_rate": 0.00010265082266910421,
|
| 1688 |
+
"loss": 0.212,
|
| 1689 |
+
"step": 2130
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.9780621572212066,
|
| 1693 |
+
"grad_norm": 1.9950228929519653,
|
| 1694 |
+
"learning_rate": 0.00010219378427787935,
|
| 1695 |
+
"loss": 0.2508,
|
| 1696 |
+
"step": 2140
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.9826325411334552,
|
| 1700 |
+
"grad_norm": 0.7868995666503906,
|
| 1701 |
+
"learning_rate": 0.0001017367458866545,
|
| 1702 |
+
"loss": 0.1755,
|
| 1703 |
+
"step": 2150
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.9872029250457038,
|
| 1707 |
+
"grad_norm": 2.1217503547668457,
|
| 1708 |
+
"learning_rate": 0.00010127970749542961,
|
| 1709 |
+
"loss": 0.1431,
|
| 1710 |
+
"step": 2160
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.9917733089579525,
|
| 1714 |
+
"grad_norm": 0.1202094554901123,
|
| 1715 |
+
"learning_rate": 0.00010082266910420475,
|
| 1716 |
+
"loss": 0.0914,
|
| 1717 |
+
"step": 2170
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.9963436928702011,
|
| 1721 |
+
"grad_norm": 4.719986438751221,
|
| 1722 |
+
"learning_rate": 0.00010036563071297988,
|
| 1723 |
+
"loss": 0.137,
|
| 1724 |
+
"step": 2180
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.0009140767824498,
|
| 1728 |
+
"grad_norm": 0.4999386668205261,
|
| 1729 |
+
"learning_rate": 9.990859232175503e-05,
|
| 1730 |
+
"loss": 0.1686,
|
| 1731 |
+
"step": 2190
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.0054844606946984,
|
| 1735 |
+
"grad_norm": 0.2870078980922699,
|
| 1736 |
+
"learning_rate": 9.945155393053017e-05,
|
| 1737 |
+
"loss": 0.1088,
|
| 1738 |
+
"step": 2200
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.0054844606946984,
|
| 1742 |
+
"eval_accuracy": 0.9577333333333333,
|
| 1743 |
+
"eval_loss": 0.1361219733953476,
|
| 1744 |
+
"eval_runtime": 70.2423,
|
| 1745 |
+
"eval_samples_per_second": 213.547,
|
| 1746 |
+
"eval_steps_per_second": 13.354,
|
| 1747 |
+
"step": 2200
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 1.010054844606947,
|
| 1751 |
+
"grad_norm": 0.8209459781646729,
|
| 1752 |
+
"learning_rate": 9.89945155393053e-05,
|
| 1753 |
+
"loss": 0.1142,
|
| 1754 |
+
"step": 2210
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 1.0146252285191957,
|
| 1758 |
+
"grad_norm": 9.744668006896973,
|
| 1759 |
+
"learning_rate": 9.853747714808045e-05,
|
| 1760 |
+
"loss": 0.0546,
|
| 1761 |
+
"step": 2220
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 1.0191956124314443,
|
| 1765 |
+
"grad_norm": 0.15120500326156616,
|
| 1766 |
+
"learning_rate": 9.808043875685559e-05,
|
| 1767 |
+
"loss": 0.0184,
|
| 1768 |
+
"step": 2230
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 1.023765996343693,
|
| 1772 |
+
"grad_norm": 3.5596585273742676,
|
| 1773 |
+
"learning_rate": 9.762340036563071e-05,
|
| 1774 |
+
"loss": 0.1219,
|
| 1775 |
+
"step": 2240
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 1.0283363802559415,
|
| 1779 |
+
"grad_norm": 4.676599025726318,
|
| 1780 |
+
"learning_rate": 9.716636197440585e-05,
|
| 1781 |
+
"loss": 0.0767,
|
| 1782 |
+
"step": 2250
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 1.0329067641681902,
|
| 1786 |
+
"grad_norm": 0.02072470262646675,
|
| 1787 |
+
"learning_rate": 9.670932358318099e-05,
|
| 1788 |
+
"loss": 0.0152,
|
| 1789 |
+
"step": 2260
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 1.0374771480804388,
|
| 1793 |
+
"grad_norm": 0.09274908900260925,
|
| 1794 |
+
"learning_rate": 9.625228519195612e-05,
|
| 1795 |
+
"loss": 0.0423,
|
| 1796 |
+
"step": 2270
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 1.0420475319926874,
|
| 1800 |
+
"grad_norm": 2.4434385299682617,
|
| 1801 |
+
"learning_rate": 9.579524680073126e-05,
|
| 1802 |
+
"loss": 0.0651,
|
| 1803 |
+
"step": 2280
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 1.046617915904936,
|
| 1807 |
+
"grad_norm": 0.028794042766094208,
|
| 1808 |
+
"learning_rate": 9.53382084095064e-05,
|
| 1809 |
+
"loss": 0.0635,
|
| 1810 |
+
"step": 2290
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 1.0511882998171846,
|
| 1814 |
+
"grad_norm": 1.619289755821228,
|
| 1815 |
+
"learning_rate": 9.488117001828155e-05,
|
| 1816 |
+
"loss": 0.0916,
|
| 1817 |
+
"step": 2300
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 1.0511882998171846,
|
| 1821 |
+
"eval_accuracy": 0.9596666666666667,
|
| 1822 |
+
"eval_loss": 0.13929586112499237,
|
| 1823 |
+
"eval_runtime": 70.231,
|
| 1824 |
+
"eval_samples_per_second": 213.581,
|
| 1825 |
+
"eval_steps_per_second": 13.356,
|
| 1826 |
+
"step": 2300
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 1.0557586837294333,
|
| 1830 |
+
"grad_norm": 4.361185073852539,
|
| 1831 |
+
"learning_rate": 9.442413162705668e-05,
|
| 1832 |
+
"loss": 0.039,
|
| 1833 |
+
"step": 2310
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 1.0603290676416819,
|
| 1837 |
+
"grad_norm": 0.0270086620002985,
|
| 1838 |
+
"learning_rate": 9.396709323583182e-05,
|
| 1839 |
+
"loss": 0.087,
|
| 1840 |
+
"step": 2320
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 1.0648994515539305,
|
| 1844 |
+
"grad_norm": 0.09628736972808838,
|
| 1845 |
+
"learning_rate": 9.351005484460696e-05,
|
| 1846 |
+
"loss": 0.0222,
|
| 1847 |
+
"step": 2330
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 1.0694698354661791,
|
| 1851 |
+
"grad_norm": 4.285031318664551,
|
| 1852 |
+
"learning_rate": 9.305301645338209e-05,
|
| 1853 |
+
"loss": 0.0478,
|
| 1854 |
+
"step": 2340
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 1.0740402193784278,
|
| 1858 |
+
"grad_norm": 0.06926529854536057,
|
| 1859 |
+
"learning_rate": 9.259597806215723e-05,
|
| 1860 |
+
"loss": 0.0531,
|
| 1861 |
+
"step": 2350
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 1.0786106032906764,
|
| 1865 |
+
"grad_norm": 0.01788182742893696,
|
| 1866 |
+
"learning_rate": 9.213893967093236e-05,
|
| 1867 |
+
"loss": 0.0723,
|
| 1868 |
+
"step": 2360
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 1.083180987202925,
|
| 1872 |
+
"grad_norm": 0.1285097748041153,
|
| 1873 |
+
"learning_rate": 9.16819012797075e-05,
|
| 1874 |
+
"loss": 0.0565,
|
| 1875 |
+
"step": 2370
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 1.0877513711151736,
|
| 1879 |
+
"grad_norm": 0.013244764879345894,
|
| 1880 |
+
"learning_rate": 9.122486288848264e-05,
|
| 1881 |
+
"loss": 0.0363,
|
| 1882 |
+
"step": 2380
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 1.0923217550274222,
|
| 1886 |
+
"grad_norm": 2.3318049907684326,
|
| 1887 |
+
"learning_rate": 9.076782449725777e-05,
|
| 1888 |
+
"loss": 0.0584,
|
| 1889 |
+
"step": 2390
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 1.0968921389396709,
|
| 1893 |
+
"grad_norm": 0.3406066298484802,
|
| 1894 |
+
"learning_rate": 9.031078610603291e-05,
|
| 1895 |
+
"loss": 0.005,
|
| 1896 |
+
"step": 2400
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 1.0968921389396709,
|
| 1900 |
+
"eval_accuracy": 0.9620666666666666,
|
| 1901 |
+
"eval_loss": 0.12949973344802856,
|
| 1902 |
+
"eval_runtime": 70.9675,
|
| 1903 |
+
"eval_samples_per_second": 211.364,
|
| 1904 |
+
"eval_steps_per_second": 13.217,
|
| 1905 |
+
"step": 2400
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 1.1014625228519195,
|
| 1909 |
+
"grad_norm": 0.09992707520723343,
|
| 1910 |
+
"learning_rate": 8.985374771480805e-05,
|
| 1911 |
+
"loss": 0.006,
|
| 1912 |
+
"step": 2410
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 1.106032906764168,
|
| 1916 |
+
"grad_norm": 3.9798166751861572,
|
| 1917 |
+
"learning_rate": 8.939670932358318e-05,
|
| 1918 |
+
"loss": 0.0458,
|
| 1919 |
+
"step": 2420
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 1.1106032906764167,
|
| 1923 |
+
"grad_norm": 0.20624032616615295,
|
| 1924 |
+
"learning_rate": 8.893967093235832e-05,
|
| 1925 |
+
"loss": 0.0366,
|
| 1926 |
+
"step": 2430
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 1.1151736745886653,
|
| 1930 |
+
"grad_norm": 0.03891080617904663,
|
| 1931 |
+
"learning_rate": 8.848263254113346e-05,
|
| 1932 |
+
"loss": 0.0118,
|
| 1933 |
+
"step": 2440
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 1.1197440585009142,
|
| 1937 |
+
"grad_norm": 7.250652313232422,
|
| 1938 |
+
"learning_rate": 8.802559414990859e-05,
|
| 1939 |
+
"loss": 0.0622,
|
| 1940 |
+
"step": 2450
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 1.1243144424131628,
|
| 1944 |
+
"grad_norm": 2.0701119899749756,
|
| 1945 |
+
"learning_rate": 8.756855575868373e-05,
|
| 1946 |
+
"loss": 0.0504,
|
| 1947 |
+
"step": 2460
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 1.1288848263254114,
|
| 1951 |
+
"grad_norm": 4.752568244934082,
|
| 1952 |
+
"learning_rate": 8.711151736745888e-05,
|
| 1953 |
+
"loss": 0.0581,
|
| 1954 |
+
"step": 2470
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 1.13345521023766,
|
| 1958 |
+
"grad_norm": 0.023835673928260803,
|
| 1959 |
+
"learning_rate": 8.665447897623402e-05,
|
| 1960 |
+
"loss": 0.0401,
|
| 1961 |
+
"step": 2480
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 1.1380255941499087,
|
| 1965 |
+
"grad_norm": 0.009058034047484398,
|
| 1966 |
+
"learning_rate": 8.619744058500915e-05,
|
| 1967 |
+
"loss": 0.0406,
|
| 1968 |
+
"step": 2490
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 1.1425959780621573,
|
| 1972 |
+
"grad_norm": 0.2688920795917511,
|
| 1973 |
+
"learning_rate": 8.574040219378429e-05,
|
| 1974 |
+
"loss": 0.0294,
|
| 1975 |
+
"step": 2500
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 1.1425959780621573,
|
| 1979 |
+
"eval_accuracy": 0.9639333333333333,
|
| 1980 |
+
"eval_loss": 0.1327054649591446,
|
| 1981 |
+
"eval_runtime": 70.9468,
|
| 1982 |
+
"eval_samples_per_second": 211.426,
|
| 1983 |
+
"eval_steps_per_second": 13.221,
|
| 1984 |
+
"step": 2500
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.147166361974406,
|
| 1988 |
+
"grad_norm": 0.007774589583277702,
|
| 1989 |
+
"learning_rate": 8.528336380255942e-05,
|
| 1990 |
+
"loss": 0.0427,
|
| 1991 |
+
"step": 2510
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 1.1517367458866545,
|
| 1995 |
+
"grad_norm": 5.866430759429932,
|
| 1996 |
+
"learning_rate": 8.482632541133455e-05,
|
| 1997 |
+
"loss": 0.0595,
|
| 1998 |
+
"step": 2520
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.1563071297989032,
|
| 2002 |
+
"grad_norm": 0.11866763979196548,
|
| 2003 |
+
"learning_rate": 8.436928702010968e-05,
|
| 2004 |
+
"loss": 0.0673,
|
| 2005 |
+
"step": 2530
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 1.1608775137111518,
|
| 2009 |
+
"grad_norm": 5.5359978675842285,
|
| 2010 |
+
"learning_rate": 8.391224862888482e-05,
|
| 2011 |
+
"loss": 0.1446,
|
| 2012 |
+
"step": 2540
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 1.1654478976234004,
|
| 2016 |
+
"grad_norm": 0.016985343769192696,
|
| 2017 |
+
"learning_rate": 8.345521023765997e-05,
|
| 2018 |
+
"loss": 0.0559,
|
| 2019 |
+
"step": 2550
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 1.170018281535649,
|
| 2023 |
+
"grad_norm": 0.7032074928283691,
|
| 2024 |
+
"learning_rate": 8.29981718464351e-05,
|
| 2025 |
+
"loss": 0.0093,
|
| 2026 |
+
"step": 2560
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 1.1745886654478976,
|
| 2030 |
+
"grad_norm": 0.14500297605991364,
|
| 2031 |
+
"learning_rate": 8.254113345521024e-05,
|
| 2032 |
+
"loss": 0.0591,
|
| 2033 |
+
"step": 2570
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 1.1791590493601463,
|
| 2037 |
+
"grad_norm": 4.615384578704834,
|
| 2038 |
+
"learning_rate": 8.208409506398538e-05,
|
| 2039 |
+
"loss": 0.0167,
|
| 2040 |
+
"step": 2580
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.1837294332723949,
|
| 2044 |
+
"grad_norm": 3.747305154800415,
|
| 2045 |
+
"learning_rate": 8.162705667276052e-05,
|
| 2046 |
+
"loss": 0.1315,
|
| 2047 |
+
"step": 2590
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 1.1882998171846435,
|
| 2051 |
+
"grad_norm": 6.55547571182251,
|
| 2052 |
+
"learning_rate": 8.117001828153565e-05,
|
| 2053 |
+
"loss": 0.0939,
|
| 2054 |
+
"step": 2600
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 1.1882998171846435,
|
| 2058 |
+
"eval_accuracy": 0.9621333333333333,
|
| 2059 |
+
"eval_loss": 0.1408853828907013,
|
| 2060 |
+
"eval_runtime": 70.3628,
|
| 2061 |
+
"eval_samples_per_second": 213.181,
|
| 2062 |
+
"eval_steps_per_second": 13.331,
|
| 2063 |
+
"step": 2600
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 1.1928702010968921,
|
| 2067 |
+
"grad_norm": 0.018332751467823982,
|
| 2068 |
+
"learning_rate": 8.071297989031079e-05,
|
| 2069 |
+
"loss": 0.0422,
|
| 2070 |
+
"step": 2610
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 1.1974405850091407,
|
| 2074 |
+
"grad_norm": 8.07509708404541,
|
| 2075 |
+
"learning_rate": 8.025594149908592e-05,
|
| 2076 |
+
"loss": 0.0548,
|
| 2077 |
+
"step": 2620
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 1.2020109689213894,
|
| 2081 |
+
"grad_norm": 0.015664540231227875,
|
| 2082 |
+
"learning_rate": 7.979890310786106e-05,
|
| 2083 |
+
"loss": 0.0201,
|
| 2084 |
+
"step": 2630
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 1.206581352833638,
|
| 2088 |
+
"grad_norm": 0.34166190028190613,
|
| 2089 |
+
"learning_rate": 7.934186471663621e-05,
|
| 2090 |
+
"loss": 0.0063,
|
| 2091 |
+
"step": 2640
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 1.2111517367458866,
|
| 2095 |
+
"grad_norm": 0.024543585255742073,
|
| 2096 |
+
"learning_rate": 7.888482632541135e-05,
|
| 2097 |
+
"loss": 0.067,
|
| 2098 |
+
"step": 2650
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 1.2157221206581352,
|
| 2102 |
+
"grad_norm": 5.0015788078308105,
|
| 2103 |
+
"learning_rate": 7.842778793418648e-05,
|
| 2104 |
+
"loss": 0.0677,
|
| 2105 |
+
"step": 2660
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 1.2202925045703839,
|
| 2109 |
+
"grad_norm": 0.3825192153453827,
|
| 2110 |
+
"learning_rate": 7.79707495429616e-05,
|
| 2111 |
+
"loss": 0.0447,
|
| 2112 |
+
"step": 2670
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 1.2248628884826325,
|
| 2116 |
+
"grad_norm": 1.0526628494262695,
|
| 2117 |
+
"learning_rate": 7.751371115173674e-05,
|
| 2118 |
+
"loss": 0.0264,
|
| 2119 |
+
"step": 2680
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 1.229433272394881,
|
| 2123 |
+
"grad_norm": 0.015143281780183315,
|
| 2124 |
+
"learning_rate": 7.705667276051188e-05,
|
| 2125 |
+
"loss": 0.0182,
|
| 2126 |
+
"step": 2690
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 1.2340036563071297,
|
| 2130 |
+
"grad_norm": 6.874438285827637,
|
| 2131 |
+
"learning_rate": 7.659963436928702e-05,
|
| 2132 |
+
"loss": 0.0756,
|
| 2133 |
+
"step": 2700
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 1.2340036563071297,
|
| 2137 |
+
"eval_accuracy": 0.9682,
|
| 2138 |
+
"eval_loss": 0.1201971173286438,
|
| 2139 |
+
"eval_runtime": 70.6459,
|
| 2140 |
+
"eval_samples_per_second": 212.327,
|
| 2141 |
+
"eval_steps_per_second": 13.277,
|
| 2142 |
+
"step": 2700
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 1.2385740402193783,
|
| 2146 |
+
"grad_norm": 0.02307463437318802,
|
| 2147 |
+
"learning_rate": 7.614259597806215e-05,
|
| 2148 |
+
"loss": 0.033,
|
| 2149 |
+
"step": 2710
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 1.2431444241316272,
|
| 2153 |
+
"grad_norm": 0.01136768702417612,
|
| 2154 |
+
"learning_rate": 7.56855575868373e-05,
|
| 2155 |
+
"loss": 0.0762,
|
| 2156 |
+
"step": 2720
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 1.2477148080438756,
|
| 2160 |
+
"grad_norm": 5.031988620758057,
|
| 2161 |
+
"learning_rate": 7.522851919561244e-05,
|
| 2162 |
+
"loss": 0.0692,
|
| 2163 |
+
"step": 2730
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 1.2522851919561244,
|
| 2167 |
+
"grad_norm": 0.028815852478146553,
|
| 2168 |
+
"learning_rate": 7.477148080438758e-05,
|
| 2169 |
+
"loss": 0.008,
|
| 2170 |
+
"step": 2740
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 1.2568555758683728,
|
| 2174 |
+
"grad_norm": 10.7840576171875,
|
| 2175 |
+
"learning_rate": 7.431444241316271e-05,
|
| 2176 |
+
"loss": 0.0711,
|
| 2177 |
+
"step": 2750
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 1.2614259597806217,
|
| 2181 |
+
"grad_norm": 3.8280370235443115,
|
| 2182 |
+
"learning_rate": 7.385740402193785e-05,
|
| 2183 |
+
"loss": 0.0581,
|
| 2184 |
+
"step": 2760
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 1.26599634369287,
|
| 2188 |
+
"grad_norm": 0.03191199526190758,
|
| 2189 |
+
"learning_rate": 7.340036563071298e-05,
|
| 2190 |
+
"loss": 0.0882,
|
| 2191 |
+
"step": 2770
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 1.270566727605119,
|
| 2195 |
+
"grad_norm": 0.010684626176953316,
|
| 2196 |
+
"learning_rate": 7.294332723948812e-05,
|
| 2197 |
+
"loss": 0.0228,
|
| 2198 |
+
"step": 2780
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 1.2751371115173675,
|
| 2202 |
+
"grad_norm": 0.01165696233510971,
|
| 2203 |
+
"learning_rate": 7.248628884826326e-05,
|
| 2204 |
+
"loss": 0.0364,
|
| 2205 |
+
"step": 2790
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 1.2797074954296161,
|
| 2209 |
+
"grad_norm": 5.020371913909912,
|
| 2210 |
+
"learning_rate": 7.20292504570384e-05,
|
| 2211 |
+
"loss": 0.0466,
|
| 2212 |
+
"step": 2800
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 1.2797074954296161,
|
| 2216 |
+
"eval_accuracy": 0.964,
|
| 2217 |
+
"eval_loss": 0.1273525506258011,
|
| 2218 |
+
"eval_runtime": 70.5245,
|
| 2219 |
+
"eval_samples_per_second": 212.692,
|
| 2220 |
+
"eval_steps_per_second": 13.3,
|
| 2221 |
+
"step": 2800
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 1.2842778793418648,
|
| 2225 |
+
"grad_norm": 3.339799165725708,
|
| 2226 |
+
"learning_rate": 7.157221206581353e-05,
|
| 2227 |
+
"loss": 0.0331,
|
| 2228 |
+
"step": 2810
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 1.2888482632541134,
|
| 2232 |
+
"grad_norm": 2.2458271980285645,
|
| 2233 |
+
"learning_rate": 7.111517367458867e-05,
|
| 2234 |
+
"loss": 0.0139,
|
| 2235 |
+
"step": 2820
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 1.293418647166362,
|
| 2239 |
+
"grad_norm": 0.03158143162727356,
|
| 2240 |
+
"learning_rate": 7.06581352833638e-05,
|
| 2241 |
+
"loss": 0.0316,
|
| 2242 |
+
"step": 2830
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 1.2979890310786106,
|
| 2246 |
+
"grad_norm": 0.012245237827301025,
|
| 2247 |
+
"learning_rate": 7.020109689213894e-05,
|
| 2248 |
+
"loss": 0.0501,
|
| 2249 |
+
"step": 2840
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 1.3025594149908593,
|
| 2253 |
+
"grad_norm": 6.688177585601807,
|
| 2254 |
+
"learning_rate": 6.974405850091408e-05,
|
| 2255 |
+
"loss": 0.0349,
|
| 2256 |
+
"step": 2850
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 1.3071297989031079,
|
| 2260 |
+
"grad_norm": 0.01068816240876913,
|
| 2261 |
+
"learning_rate": 6.928702010968921e-05,
|
| 2262 |
+
"loss": 0.0373,
|
| 2263 |
+
"step": 2860
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 1.3117001828153565,
|
| 2267 |
+
"grad_norm": 0.017882896587252617,
|
| 2268 |
+
"learning_rate": 6.882998171846435e-05,
|
| 2269 |
+
"loss": 0.0272,
|
| 2270 |
+
"step": 2870
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 1.3162705667276051,
|
| 2274 |
+
"grad_norm": 0.009448254480957985,
|
| 2275 |
+
"learning_rate": 6.837294332723948e-05,
|
| 2276 |
+
"loss": 0.045,
|
| 2277 |
+
"step": 2880
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 1.3208409506398537,
|
| 2281 |
+
"grad_norm": 8.753164291381836,
|
| 2282 |
+
"learning_rate": 6.791590493601463e-05,
|
| 2283 |
+
"loss": 0.1252,
|
| 2284 |
+
"step": 2890
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 1.3254113345521024,
|
| 2288 |
+
"grad_norm": 0.02937444858253002,
|
| 2289 |
+
"learning_rate": 6.745886654478977e-05,
|
| 2290 |
+
"loss": 0.0565,
|
| 2291 |
+
"step": 2900
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 1.3254113345521024,
|
| 2295 |
+
"eval_accuracy": 0.9662666666666667,
|
| 2296 |
+
"eval_loss": 0.12496425956487656,
|
| 2297 |
+
"eval_runtime": 70.9263,
|
| 2298 |
+
"eval_samples_per_second": 211.487,
|
| 2299 |
+
"eval_steps_per_second": 13.225,
|
| 2300 |
+
"step": 2900
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 1.329981718464351,
|
| 2304 |
+
"grad_norm": 0.7451736927032471,
|
| 2305 |
+
"learning_rate": 6.700182815356491e-05,
|
| 2306 |
+
"loss": 0.0298,
|
| 2307 |
+
"step": 2910
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 1.3345521023765996,
|
| 2311 |
+
"grad_norm": 0.014505515806376934,
|
| 2312 |
+
"learning_rate": 6.654478976234004e-05,
|
| 2313 |
+
"loss": 0.0954,
|
| 2314 |
+
"step": 2920
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 1.3391224862888482,
|
| 2318 |
+
"grad_norm": 6.800475597381592,
|
| 2319 |
+
"learning_rate": 6.608775137111518e-05,
|
| 2320 |
+
"loss": 0.0584,
|
| 2321 |
+
"step": 2930
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 1.3436928702010968,
|
| 2325 |
+
"grad_norm": 0.08044274151325226,
|
| 2326 |
+
"learning_rate": 6.563071297989032e-05,
|
| 2327 |
+
"loss": 0.0944,
|
| 2328 |
+
"step": 2940
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 1.3482632541133455,
|
| 2332 |
+
"grad_norm": 0.051912058144807816,
|
| 2333 |
+
"learning_rate": 6.517367458866545e-05,
|
| 2334 |
+
"loss": 0.0128,
|
| 2335 |
+
"step": 2950
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 1.352833638025594,
|
| 2339 |
+
"grad_norm": 0.8656260967254639,
|
| 2340 |
+
"learning_rate": 6.471663619744059e-05,
|
| 2341 |
+
"loss": 0.1187,
|
| 2342 |
+
"step": 2960
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 1.3574040219378427,
|
| 2346 |
+
"grad_norm": 0.004978422075510025,
|
| 2347 |
+
"learning_rate": 6.425959780621573e-05,
|
| 2348 |
+
"loss": 0.0228,
|
| 2349 |
+
"step": 2970
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 1.3619744058500913,
|
| 2353 |
+
"grad_norm": 0.5934199094772339,
|
| 2354 |
+
"learning_rate": 6.380255941499086e-05,
|
| 2355 |
+
"loss": 0.0038,
|
| 2356 |
+
"step": 2980
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 1.3665447897623402,
|
| 2360 |
+
"grad_norm": 3.7717771530151367,
|
| 2361 |
+
"learning_rate": 6.3345521023766e-05,
|
| 2362 |
+
"loss": 0.087,
|
| 2363 |
+
"step": 2990
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 1.3711151736745886,
|
| 2367 |
+
"grad_norm": 0.01153448224067688,
|
| 2368 |
+
"learning_rate": 6.288848263254114e-05,
|
| 2369 |
+
"loss": 0.0609,
|
| 2370 |
+
"step": 3000
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 1.3711151736745886,
|
| 2374 |
+
"eval_accuracy": 0.9656666666666667,
|
| 2375 |
+
"eval_loss": 0.12994171679019928,
|
| 2376 |
+
"eval_runtime": 71.2707,
|
| 2377 |
+
"eval_samples_per_second": 210.465,
|
| 2378 |
+
"eval_steps_per_second": 13.161,
|
| 2379 |
+
"step": 3000
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 1.3756855575868374,
|
| 2383 |
+
"grad_norm": 0.17517925798892975,
|
| 2384 |
+
"learning_rate": 6.243144424131627e-05,
|
| 2385 |
+
"loss": 0.026,
|
| 2386 |
+
"step": 3010
|
| 2387 |
+
},
|
| 2388 |
+
{
|
| 2389 |
+
"epoch": 1.3802559414990858,
|
| 2390 |
+
"grad_norm": 0.11696294695138931,
|
| 2391 |
+
"learning_rate": 6.197440585009141e-05,
|
| 2392 |
+
"loss": 0.025,
|
| 2393 |
+
"step": 3020
|
| 2394 |
+
},
|
| 2395 |
+
{
|
| 2396 |
+
"epoch": 1.3848263254113347,
|
| 2397 |
+
"grad_norm": 0.007365319412201643,
|
| 2398 |
+
"learning_rate": 6.151736745886654e-05,
|
| 2399 |
+
"loss": 0.0251,
|
| 2400 |
+
"step": 3030
|
| 2401 |
+
},
|
| 2402 |
+
{
|
| 2403 |
+
"epoch": 1.389396709323583,
|
| 2404 |
+
"grad_norm": 0.3844846189022064,
|
| 2405 |
+
"learning_rate": 6.106032906764168e-05,
|
| 2406 |
+
"loss": 0.0252,
|
| 2407 |
+
"step": 3040
|
| 2408 |
+
},
|
| 2409 |
+
{
|
| 2410 |
+
"epoch": 1.393967093235832,
|
| 2411 |
+
"grad_norm": 0.014660494402050972,
|
| 2412 |
+
"learning_rate": 6.0603290676416824e-05,
|
| 2413 |
+
"loss": 0.0274,
|
| 2414 |
+
"step": 3050
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 1.3985374771480805,
|
| 2418 |
+
"grad_norm": 4.684697151184082,
|
| 2419 |
+
"learning_rate": 6.014625228519196e-05,
|
| 2420 |
+
"loss": 0.0294,
|
| 2421 |
+
"step": 3060
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 1.4031078610603291,
|
| 2425 |
+
"grad_norm": 0.05706701800227165,
|
| 2426 |
+
"learning_rate": 5.96892138939671e-05,
|
| 2427 |
+
"loss": 0.0053,
|
| 2428 |
+
"step": 3070
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 1.4076782449725778,
|
| 2432 |
+
"grad_norm": 0.010988248512148857,
|
| 2433 |
+
"learning_rate": 5.923217550274224e-05,
|
| 2434 |
+
"loss": 0.0227,
|
| 2435 |
+
"step": 3080
|
| 2436 |
+
},
|
| 2437 |
+
{
|
| 2438 |
+
"epoch": 1.4122486288848264,
|
| 2439 |
+
"grad_norm": 0.009499771520495415,
|
| 2440 |
+
"learning_rate": 5.8775137111517377e-05,
|
| 2441 |
+
"loss": 0.0061,
|
| 2442 |
+
"step": 3090
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 1.416819012797075,
|
| 2446 |
+
"grad_norm": 0.35233938694000244,
|
| 2447 |
+
"learning_rate": 5.83180987202925e-05,
|
| 2448 |
+
"loss": 0.0201,
|
| 2449 |
+
"step": 3100
|
| 2450 |
+
},
|
| 2451 |
+
{
|
| 2452 |
+
"epoch": 1.416819012797075,
|
| 2453 |
+
"eval_accuracy": 0.9685333333333334,
|
| 2454 |
+
"eval_loss": 0.12030760943889618,
|
| 2455 |
+
"eval_runtime": 70.6066,
|
| 2456 |
+
"eval_samples_per_second": 212.445,
|
| 2457 |
+
"eval_steps_per_second": 13.285,
|
| 2458 |
+
"step": 3100
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 1.4213893967093236,
|
| 2462 |
+
"grad_norm": 0.680586576461792,
|
| 2463 |
+
"learning_rate": 5.786106032906764e-05,
|
| 2464 |
+
"loss": 0.0825,
|
| 2465 |
+
"step": 3110
|
| 2466 |
+
},
|
| 2467 |
+
{
|
| 2468 |
+
"epoch": 1.4259597806215722,
|
| 2469 |
+
"grad_norm": 1.1647635698318481,
|
| 2470 |
+
"learning_rate": 5.740402193784278e-05,
|
| 2471 |
+
"loss": 0.0668,
|
| 2472 |
+
"step": 3120
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 1.4305301645338209,
|
| 2476 |
+
"grad_norm": 0.3994854986667633,
|
| 2477 |
+
"learning_rate": 5.6946983546617915e-05,
|
| 2478 |
+
"loss": 0.054,
|
| 2479 |
+
"step": 3130
|
| 2480 |
+
},
|
| 2481 |
+
{
|
| 2482 |
+
"epoch": 1.4351005484460695,
|
| 2483 |
+
"grad_norm": 0.012723923660814762,
|
| 2484 |
+
"learning_rate": 5.648994515539305e-05,
|
| 2485 |
+
"loss": 0.1109,
|
| 2486 |
+
"step": 3140
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 1.4396709323583181,
|
| 2490 |
+
"grad_norm": 0.02893258072435856,
|
| 2491 |
+
"learning_rate": 5.603290676416819e-05,
|
| 2492 |
+
"loss": 0.0447,
|
| 2493 |
+
"step": 3150
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 1.4442413162705667,
|
| 2497 |
+
"grad_norm": 2.296046495437622,
|
| 2498 |
+
"learning_rate": 5.557586837294333e-05,
|
| 2499 |
+
"loss": 0.0139,
|
| 2500 |
+
"step": 3160
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 1.4488117001828154,
|
| 2504 |
+
"grad_norm": 0.0075446791015565395,
|
| 2505 |
+
"learning_rate": 5.511882998171847e-05,
|
| 2506 |
+
"loss": 0.0254,
|
| 2507 |
+
"step": 3170
|
| 2508 |
+
},
|
| 2509 |
+
{
|
| 2510 |
+
"epoch": 1.453382084095064,
|
| 2511 |
+
"grad_norm": 4.601187705993652,
|
| 2512 |
+
"learning_rate": 5.47074954296161e-05,
|
| 2513 |
+
"loss": 0.0619,
|
| 2514 |
+
"step": 3180
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 1.4579524680073126,
|
| 2518 |
+
"grad_norm": 0.028089461848139763,
|
| 2519 |
+
"learning_rate": 5.425045703839122e-05,
|
| 2520 |
+
"loss": 0.0324,
|
| 2521 |
+
"step": 3190
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 1.4625228519195612,
|
| 2525 |
+
"grad_norm": 0.00567116541787982,
|
| 2526 |
+
"learning_rate": 5.3793418647166363e-05,
|
| 2527 |
+
"loss": 0.0258,
|
| 2528 |
+
"step": 3200
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 1.4625228519195612,
|
| 2532 |
+
"eval_accuracy": 0.9692666666666667,
|
| 2533 |
+
"eval_loss": 0.11664163321256638,
|
| 2534 |
+
"eval_runtime": 70.1343,
|
| 2535 |
+
"eval_samples_per_second": 213.875,
|
| 2536 |
+
"eval_steps_per_second": 13.374,
|
| 2537 |
+
"step": 3200
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 1.4670932358318098,
|
| 2541 |
+
"grad_norm": 0.008565380237996578,
|
| 2542 |
+
"learning_rate": 5.33363802559415e-05,
|
| 2543 |
+
"loss": 0.1077,
|
| 2544 |
+
"step": 3210
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 1.4716636197440585,
|
| 2548 |
+
"grad_norm": 6.626660346984863,
|
| 2549 |
+
"learning_rate": 5.2879341864716636e-05,
|
| 2550 |
+
"loss": 0.0685,
|
| 2551 |
+
"step": 3220
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 1.476234003656307,
|
| 2555 |
+
"grad_norm": 5.336280822753906,
|
| 2556 |
+
"learning_rate": 5.242230347349177e-05,
|
| 2557 |
+
"loss": 0.0734,
|
| 2558 |
+
"step": 3230
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 1.4808043875685557,
|
| 2562 |
+
"grad_norm": 2.9951882362365723,
|
| 2563 |
+
"learning_rate": 5.196526508226691e-05,
|
| 2564 |
+
"loss": 0.0216,
|
| 2565 |
+
"step": 3240
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 1.4853747714808043,
|
| 2569 |
+
"grad_norm": 0.44102242588996887,
|
| 2570 |
+
"learning_rate": 5.150822669104205e-05,
|
| 2571 |
+
"loss": 0.0036,
|
| 2572 |
+
"step": 3250
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 1.489945155393053,
|
| 2576 |
+
"grad_norm": 3.0561587810516357,
|
| 2577 |
+
"learning_rate": 5.105118829981719e-05,
|
| 2578 |
+
"loss": 0.0808,
|
| 2579 |
+
"step": 3260
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 1.4945155393053016,
|
| 2583 |
+
"grad_norm": 0.09664315730333328,
|
| 2584 |
+
"learning_rate": 5.0594149908592325e-05,
|
| 2585 |
+
"loss": 0.0796,
|
| 2586 |
+
"step": 3270
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 1.4990859232175504,
|
| 2590 |
+
"grad_norm": 0.01629328727722168,
|
| 2591 |
+
"learning_rate": 5.013711151736746e-05,
|
| 2592 |
+
"loss": 0.003,
|
| 2593 |
+
"step": 3280
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 1.5036563071297988,
|
| 2597 |
+
"grad_norm": 0.008465313352644444,
|
| 2598 |
+
"learning_rate": 4.96800731261426e-05,
|
| 2599 |
+
"loss": 0.1065,
|
| 2600 |
+
"step": 3290
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 1.5082266910420477,
|
| 2604 |
+
"grad_norm": 2.8709588050842285,
|
| 2605 |
+
"learning_rate": 4.9223034734917734e-05,
|
| 2606 |
+
"loss": 0.0913,
|
| 2607 |
+
"step": 3300
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 1.5082266910420477,
|
| 2611 |
+
"eval_accuracy": 0.9736,
|
| 2612 |
+
"eval_loss": 0.10086899250745773,
|
| 2613 |
+
"eval_runtime": 70.4855,
|
| 2614 |
+
"eval_samples_per_second": 212.81,
|
| 2615 |
+
"eval_steps_per_second": 13.308,
|
| 2616 |
+
"step": 3300
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 1.512797074954296,
|
| 2620 |
+
"grad_norm": 0.006857722532004118,
|
| 2621 |
+
"learning_rate": 4.876599634369287e-05,
|
| 2622 |
+
"loss": 0.0035,
|
| 2623 |
+
"step": 3310
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 1.517367458866545,
|
| 2627 |
+
"grad_norm": 0.046820204704999924,
|
| 2628 |
+
"learning_rate": 4.830895795246801e-05,
|
| 2629 |
+
"loss": 0.0234,
|
| 2630 |
+
"step": 3320
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 1.5219378427787933,
|
| 2634 |
+
"grad_norm": 0.009667345322668552,
|
| 2635 |
+
"learning_rate": 4.785191956124315e-05,
|
| 2636 |
+
"loss": 0.0304,
|
| 2637 |
+
"step": 3330
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 1.5265082266910421,
|
| 2641 |
+
"grad_norm": 5.041330814361572,
|
| 2642 |
+
"learning_rate": 4.739488117001829e-05,
|
| 2643 |
+
"loss": 0.0812,
|
| 2644 |
+
"step": 3340
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 1.5310786106032905,
|
| 2648 |
+
"grad_norm": 0.12366422265768051,
|
| 2649 |
+
"learning_rate": 4.693784277879342e-05,
|
| 2650 |
+
"loss": 0.0913,
|
| 2651 |
+
"step": 3350
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 1.5356489945155394,
|
| 2655 |
+
"grad_norm": 0.19020313024520874,
|
| 2656 |
+
"learning_rate": 4.648080438756856e-05,
|
| 2657 |
+
"loss": 0.0027,
|
| 2658 |
+
"step": 3360
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 1.5402193784277878,
|
| 2662 |
+
"grad_norm": 0.008719071745872498,
|
| 2663 |
+
"learning_rate": 4.6023765996343696e-05,
|
| 2664 |
+
"loss": 0.0087,
|
| 2665 |
+
"step": 3370
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 1.5447897623400366,
|
| 2669 |
+
"grad_norm": 5.164638996124268,
|
| 2670 |
+
"learning_rate": 4.556672760511883e-05,
|
| 2671 |
+
"loss": 0.0509,
|
| 2672 |
+
"step": 3380
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 1.5493601462522852,
|
| 2676 |
+
"grad_norm": 1.9091380834579468,
|
| 2677 |
+
"learning_rate": 4.510968921389397e-05,
|
| 2678 |
+
"loss": 0.0622,
|
| 2679 |
+
"step": 3390
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 1.5539305301645339,
|
| 2683 |
+
"grad_norm": 0.015776393935084343,
|
| 2684 |
+
"learning_rate": 4.4652650822669105e-05,
|
| 2685 |
+
"loss": 0.0235,
|
| 2686 |
+
"step": 3400
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 1.5539305301645339,
|
| 2690 |
+
"eval_accuracy": 0.9732,
|
| 2691 |
+
"eval_loss": 0.0964307188987732,
|
| 2692 |
+
"eval_runtime": 71.1476,
|
| 2693 |
+
"eval_samples_per_second": 210.829,
|
| 2694 |
+
"eval_steps_per_second": 13.184,
|
| 2695 |
+
"step": 3400
|
| 2696 |
+
},
|
| 2697 |
+
{
|
| 2698 |
+
"epoch": 1.5585009140767825,
|
| 2699 |
+
"grad_norm": 0.04088925942778587,
|
| 2700 |
+
"learning_rate": 4.419561243144424e-05,
|
| 2701 |
+
"loss": 0.0579,
|
| 2702 |
+
"step": 3410
|
| 2703 |
+
},
|
| 2704 |
+
{
|
| 2705 |
+
"epoch": 1.563071297989031,
|
| 2706 |
+
"grad_norm": 0.04248817265033722,
|
| 2707 |
+
"learning_rate": 4.3738574040219385e-05,
|
| 2708 |
+
"loss": 0.0368,
|
| 2709 |
+
"step": 3420
|
| 2710 |
+
},
|
| 2711 |
+
{
|
| 2712 |
+
"epoch": 1.5676416819012797,
|
| 2713 |
+
"grad_norm": 0.06178814917802811,
|
| 2714 |
+
"learning_rate": 4.328153564899452e-05,
|
| 2715 |
+
"loss": 0.0356,
|
| 2716 |
+
"step": 3430
|
| 2717 |
+
},
|
| 2718 |
+
{
|
| 2719 |
+
"epoch": 1.5722120658135283,
|
| 2720 |
+
"grad_norm": 0.014863620512187481,
|
| 2721 |
+
"learning_rate": 4.282449725776965e-05,
|
| 2722 |
+
"loss": 0.0173,
|
| 2723 |
+
"step": 3440
|
| 2724 |
+
},
|
| 2725 |
+
{
|
| 2726 |
+
"epoch": 1.576782449725777,
|
| 2727 |
+
"grad_norm": 0.0053153312765061855,
|
| 2728 |
+
"learning_rate": 4.236745886654479e-05,
|
| 2729 |
+
"loss": 0.0295,
|
| 2730 |
+
"step": 3450
|
| 2731 |
+
},
|
| 2732 |
+
{
|
| 2733 |
+
"epoch": 1.5813528336380256,
|
| 2734 |
+
"grad_norm": 0.03949157893657684,
|
| 2735 |
+
"learning_rate": 4.191042047531993e-05,
|
| 2736 |
+
"loss": 0.0751,
|
| 2737 |
+
"step": 3460
|
| 2738 |
+
},
|
| 2739 |
+
{
|
| 2740 |
+
"epoch": 1.5859232175502742,
|
| 2741 |
+
"grad_norm": 0.01701487973332405,
|
| 2742 |
+
"learning_rate": 4.145338208409507e-05,
|
| 2743 |
+
"loss": 0.0311,
|
| 2744 |
+
"step": 3470
|
| 2745 |
+
},
|
| 2746 |
+
{
|
| 2747 |
+
"epoch": 1.5904936014625228,
|
| 2748 |
+
"grad_norm": 0.06288379430770874,
|
| 2749 |
+
"learning_rate": 4.09963436928702e-05,
|
| 2750 |
+
"loss": 0.0779,
|
| 2751 |
+
"step": 3480
|
| 2752 |
+
},
|
| 2753 |
+
{
|
| 2754 |
+
"epoch": 1.5950639853747715,
|
| 2755 |
+
"grad_norm": 12.93021297454834,
|
| 2756 |
+
"learning_rate": 4.053930530164534e-05,
|
| 2757 |
+
"loss": 0.0274,
|
| 2758 |
+
"step": 3490
|
| 2759 |
+
},
|
| 2760 |
+
{
|
| 2761 |
+
"epoch": 1.59963436928702,
|
| 2762 |
+
"grad_norm": 0.03469611704349518,
|
| 2763 |
+
"learning_rate": 4.008226691042048e-05,
|
| 2764 |
+
"loss": 0.0089,
|
| 2765 |
+
"step": 3500
|
| 2766 |
+
},
|
| 2767 |
+
{
|
| 2768 |
+
"epoch": 1.59963436928702,
|
| 2769 |
+
"eval_accuracy": 0.9747333333333333,
|
| 2770 |
+
"eval_loss": 0.09657016396522522,
|
| 2771 |
+
"eval_runtime": 70.2849,
|
| 2772 |
+
"eval_samples_per_second": 213.417,
|
| 2773 |
+
"eval_steps_per_second": 13.346,
|
| 2774 |
+
"step": 3500
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 1.6042047531992687,
|
| 2778 |
+
"grad_norm": 0.01145760528743267,
|
| 2779 |
+
"learning_rate": 3.962522851919561e-05,
|
| 2780 |
+
"loss": 0.0226,
|
| 2781 |
+
"step": 3510
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 1.6087751371115173,
|
| 2785 |
+
"grad_norm": 0.006313066463917494,
|
| 2786 |
+
"learning_rate": 3.916819012797075e-05,
|
| 2787 |
+
"loss": 0.1061,
|
| 2788 |
+
"step": 3520
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 1.6133455210237662,
|
| 2792 |
+
"grad_norm": 0.006270520854741335,
|
| 2793 |
+
"learning_rate": 3.8711151736745885e-05,
|
| 2794 |
+
"loss": 0.0439,
|
| 2795 |
+
"step": 3530
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 1.6179159049360146,
|
| 2799 |
+
"grad_norm": 2.423236608505249,
|
| 2800 |
+
"learning_rate": 3.825411334552103e-05,
|
| 2801 |
+
"loss": 0.0273,
|
| 2802 |
+
"step": 3540
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 1.6224862888482634,
|
| 2806 |
+
"grad_norm": 0.025253351777791977,
|
| 2807 |
+
"learning_rate": 3.7797074954296165e-05,
|
| 2808 |
+
"loss": 0.0265,
|
| 2809 |
+
"step": 3550
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 1.6270566727605118,
|
| 2813 |
+
"grad_norm": 0.016615109518170357,
|
| 2814 |
+
"learning_rate": 3.73400365630713e-05,
|
| 2815 |
+
"loss": 0.0226,
|
| 2816 |
+
"step": 3560
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 1.6316270566727606,
|
| 2820 |
+
"grad_norm": 0.006486339028924704,
|
| 2821 |
+
"learning_rate": 3.688299817184644e-05,
|
| 2822 |
+
"loss": 0.0108,
|
| 2823 |
+
"step": 3570
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 1.636197440585009,
|
| 2827 |
+
"grad_norm": 0.019093792885541916,
|
| 2828 |
+
"learning_rate": 3.6425959780621574e-05,
|
| 2829 |
+
"loss": 0.0916,
|
| 2830 |
+
"step": 3580
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 1.6407678244972579,
|
| 2834 |
+
"grad_norm": 0.014140899293124676,
|
| 2835 |
+
"learning_rate": 3.596892138939671e-05,
|
| 2836 |
+
"loss": 0.1538,
|
| 2837 |
+
"step": 3590
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 1.6453382084095063,
|
| 2841 |
+
"grad_norm": 0.009632795117795467,
|
| 2842 |
+
"learning_rate": 3.551188299817185e-05,
|
| 2843 |
+
"loss": 0.0455,
|
| 2844 |
+
"step": 3600
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 1.6453382084095063,
|
| 2848 |
+
"eval_accuracy": 0.9748,
|
| 2849 |
+
"eval_loss": 0.09634851664304733,
|
| 2850 |
+
"eval_runtime": 70.4832,
|
| 2851 |
+
"eval_samples_per_second": 212.817,
|
| 2852 |
+
"eval_steps_per_second": 13.308,
|
| 2853 |
+
"step": 3600
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 1.6499085923217551,
|
| 2857 |
+
"grad_norm": 0.538037896156311,
|
| 2858 |
+
"learning_rate": 3.505484460694698e-05,
|
| 2859 |
+
"loss": 0.0632,
|
| 2860 |
+
"step": 3610
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 1.6544789762340035,
|
| 2864 |
+
"grad_norm": 2.536642551422119,
|
| 2865 |
+
"learning_rate": 3.459780621572212e-05,
|
| 2866 |
+
"loss": 0.0215,
|
| 2867 |
+
"step": 3620
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 1.6590493601462524,
|
| 2871 |
+
"grad_norm": 0.014462544582784176,
|
| 2872 |
+
"learning_rate": 3.414076782449726e-05,
|
| 2873 |
+
"loss": 0.0783,
|
| 2874 |
+
"step": 3630
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 1.6636197440585008,
|
| 2878 |
+
"grad_norm": 0.04104587808251381,
|
| 2879 |
+
"learning_rate": 3.36837294332724e-05,
|
| 2880 |
+
"loss": 0.0468,
|
| 2881 |
+
"step": 3640
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 1.6681901279707496,
|
| 2885 |
+
"grad_norm": 12.753653526306152,
|
| 2886 |
+
"learning_rate": 3.3226691042047536e-05,
|
| 2887 |
+
"loss": 0.096,
|
| 2888 |
+
"step": 3650
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 1.672760511882998,
|
| 2892 |
+
"grad_norm": 0.6275530457496643,
|
| 2893 |
+
"learning_rate": 3.2769652650822665e-05,
|
| 2894 |
+
"loss": 0.0459,
|
| 2895 |
+
"step": 3660
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 1.6773308957952469,
|
| 2899 |
+
"grad_norm": 0.028450943529605865,
|
| 2900 |
+
"learning_rate": 3.231261425959781e-05,
|
| 2901 |
+
"loss": 0.0129,
|
| 2902 |
+
"step": 3670
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 1.6819012797074955,
|
| 2906 |
+
"grad_norm": 3.766301393508911,
|
| 2907 |
+
"learning_rate": 3.1855575868372945e-05,
|
| 2908 |
+
"loss": 0.01,
|
| 2909 |
+
"step": 3680
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 1.686471663619744,
|
| 2913 |
+
"grad_norm": 1.72735595703125,
|
| 2914 |
+
"learning_rate": 3.139853747714808e-05,
|
| 2915 |
+
"loss": 0.0162,
|
| 2916 |
+
"step": 3690
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 1.6910420475319927,
|
| 2920 |
+
"grad_norm": 0.012926338240504265,
|
| 2921 |
+
"learning_rate": 3.094149908592322e-05,
|
| 2922 |
+
"loss": 0.0271,
|
| 2923 |
+
"step": 3700
|
| 2924 |
+
},
|
| 2925 |
+
{
|
| 2926 |
+
"epoch": 1.6910420475319927,
|
| 2927 |
+
"eval_accuracy": 0.9762666666666666,
|
| 2928 |
+
"eval_loss": 0.0874376893043518,
|
| 2929 |
+
"eval_runtime": 71.1478,
|
| 2930 |
+
"eval_samples_per_second": 210.829,
|
| 2931 |
+
"eval_steps_per_second": 13.184,
|
| 2932 |
+
"step": 3700
|
| 2933 |
+
},
|
| 2934 |
+
{
|
| 2935 |
+
"epoch": 1.6956124314442413,
|
| 2936 |
+
"grad_norm": 0.20961987972259521,
|
| 2937 |
+
"learning_rate": 3.0484460694698358e-05,
|
| 2938 |
+
"loss": 0.0269,
|
| 2939 |
+
"step": 3710
|
| 2940 |
+
},
|
| 2941 |
+
{
|
| 2942 |
+
"epoch": 1.70018281535649,
|
| 2943 |
+
"grad_norm": 5.752171039581299,
|
| 2944 |
+
"learning_rate": 3.0027422303473497e-05,
|
| 2945 |
+
"loss": 0.0117,
|
| 2946 |
+
"step": 3720
|
| 2947 |
+
},
|
| 2948 |
+
{
|
| 2949 |
+
"epoch": 1.7047531992687386,
|
| 2950 |
+
"grad_norm": 0.02492084540426731,
|
| 2951 |
+
"learning_rate": 2.9570383912248627e-05,
|
| 2952 |
+
"loss": 0.0494,
|
| 2953 |
+
"step": 3730
|
| 2954 |
+
},
|
| 2955 |
+
{
|
| 2956 |
+
"epoch": 1.7093235831809872,
|
| 2957 |
+
"grad_norm": 1.6408967971801758,
|
| 2958 |
+
"learning_rate": 2.9113345521023767e-05,
|
| 2959 |
+
"loss": 0.0079,
|
| 2960 |
+
"step": 3740
|
| 2961 |
+
},
|
| 2962 |
+
{
|
| 2963 |
+
"epoch": 1.7138939670932358,
|
| 2964 |
+
"grad_norm": 0.010781402699649334,
|
| 2965 |
+
"learning_rate": 2.8656307129798903e-05,
|
| 2966 |
+
"loss": 0.0288,
|
| 2967 |
+
"step": 3750
|
| 2968 |
+
},
|
| 2969 |
+
{
|
| 2970 |
+
"epoch": 1.7184643510054844,
|
| 2971 |
+
"grad_norm": 0.007589300163090229,
|
| 2972 |
+
"learning_rate": 2.8199268738574043e-05,
|
| 2973 |
+
"loss": 0.0304,
|
| 2974 |
+
"step": 3760
|
| 2975 |
+
},
|
| 2976 |
+
{
|
| 2977 |
+
"epoch": 1.723034734917733,
|
| 2978 |
+
"grad_norm": 0.0105056157335639,
|
| 2979 |
+
"learning_rate": 2.774223034734918e-05,
|
| 2980 |
+
"loss": 0.0023,
|
| 2981 |
+
"step": 3770
|
| 2982 |
+
},
|
| 2983 |
+
{
|
| 2984 |
+
"epoch": 1.7276051188299817,
|
| 2985 |
+
"grad_norm": 0.028248343616724014,
|
| 2986 |
+
"learning_rate": 2.7285191956124316e-05,
|
| 2987 |
+
"loss": 0.0156,
|
| 2988 |
+
"step": 3780
|
| 2989 |
+
},
|
| 2990 |
+
{
|
| 2991 |
+
"epoch": 1.7321755027422303,
|
| 2992 |
+
"grad_norm": 0.004776041954755783,
|
| 2993 |
+
"learning_rate": 2.6828153564899456e-05,
|
| 2994 |
+
"loss": 0.0113,
|
| 2995 |
+
"step": 3790
|
| 2996 |
+
},
|
| 2997 |
+
{
|
| 2998 |
+
"epoch": 1.736745886654479,
|
| 2999 |
+
"grad_norm": 7.163562297821045,
|
| 3000 |
+
"learning_rate": 2.637111517367459e-05,
|
| 3001 |
+
"loss": 0.0407,
|
| 3002 |
+
"step": 3800
|
| 3003 |
+
},
|
| 3004 |
+
{
|
| 3005 |
+
"epoch": 1.736745886654479,
|
| 3006 |
+
"eval_accuracy": 0.9761333333333333,
|
| 3007 |
+
"eval_loss": 0.08977096527814865,
|
| 3008 |
+
"eval_runtime": 70.6674,
|
| 3009 |
+
"eval_samples_per_second": 212.262,
|
| 3010 |
+
"eval_steps_per_second": 13.273,
|
| 3011 |
+
"step": 3800
|
| 3012 |
+
},
|
| 3013 |
+
{
|
| 3014 |
+
"epoch": 1.7413162705667276,
|
| 3015 |
+
"grad_norm": 1.1980034112930298,
|
| 3016 |
+
"learning_rate": 2.5914076782449725e-05,
|
| 3017 |
+
"loss": 0.008,
|
| 3018 |
+
"step": 3810
|
| 3019 |
+
},
|
| 3020 |
+
{
|
| 3021 |
+
"epoch": 1.7458866544789764,
|
| 3022 |
+
"grad_norm": 5.438980579376221,
|
| 3023 |
+
"learning_rate": 2.5457038391224865e-05,
|
| 3024 |
+
"loss": 0.0558,
|
| 3025 |
+
"step": 3820
|
| 3026 |
+
},
|
| 3027 |
+
{
|
| 3028 |
+
"epoch": 1.7504570383912248,
|
| 3029 |
+
"grad_norm": 0.02217746712267399,
|
| 3030 |
+
"learning_rate": 2.5e-05,
|
| 3031 |
+
"loss": 0.0057,
|
| 3032 |
+
"step": 3830
|
| 3033 |
+
},
|
| 3034 |
+
{
|
| 3035 |
+
"epoch": 1.7550274223034736,
|
| 3036 |
+
"grad_norm": 0.595504641532898,
|
| 3037 |
+
"learning_rate": 2.4542961608775138e-05,
|
| 3038 |
+
"loss": 0.0551,
|
| 3039 |
+
"step": 3840
|
| 3040 |
+
},
|
| 3041 |
+
{
|
| 3042 |
+
"epoch": 1.759597806215722,
|
| 3043 |
+
"grad_norm": 0.09388578683137894,
|
| 3044 |
+
"learning_rate": 2.4085923217550274e-05,
|
| 3045 |
+
"loss": 0.0744,
|
| 3046 |
+
"step": 3850
|
| 3047 |
+
},
|
| 3048 |
+
{
|
| 3049 |
+
"epoch": 1.7641681901279709,
|
| 3050 |
+
"grad_norm": 2.807389736175537,
|
| 3051 |
+
"learning_rate": 2.362888482632541e-05,
|
| 3052 |
+
"loss": 0.0051,
|
| 3053 |
+
"step": 3860
|
| 3054 |
+
},
|
| 3055 |
+
{
|
| 3056 |
+
"epoch": 1.7687385740402193,
|
| 3057 |
+
"grad_norm": 0.006797166541218758,
|
| 3058 |
+
"learning_rate": 2.317184643510055e-05,
|
| 3059 |
+
"loss": 0.0069,
|
| 3060 |
+
"step": 3870
|
| 3061 |
+
},
|
| 3062 |
+
{
|
| 3063 |
+
"epoch": 1.7733089579524681,
|
| 3064 |
+
"grad_norm": 0.0043932488188147545,
|
| 3065 |
+
"learning_rate": 2.2714808043875687e-05,
|
| 3066 |
+
"loss": 0.0435,
|
| 3067 |
+
"step": 3880
|
| 3068 |
+
},
|
| 3069 |
+
{
|
| 3070 |
+
"epoch": 1.7778793418647165,
|
| 3071 |
+
"grad_norm": 0.009305426850914955,
|
| 3072 |
+
"learning_rate": 2.2257769652650823e-05,
|
| 3073 |
+
"loss": 0.0189,
|
| 3074 |
+
"step": 3890
|
| 3075 |
+
},
|
| 3076 |
+
{
|
| 3077 |
+
"epoch": 1.7824497257769654,
|
| 3078 |
+
"grad_norm": 0.003490304574370384,
|
| 3079 |
+
"learning_rate": 2.180073126142596e-05,
|
| 3080 |
+
"loss": 0.1095,
|
| 3081 |
+
"step": 3900
|
| 3082 |
+
},
|
| 3083 |
+
{
|
| 3084 |
+
"epoch": 1.7824497257769654,
|
| 3085 |
+
"eval_accuracy": 0.976,
|
| 3086 |
+
"eval_loss": 0.08494840562343597,
|
| 3087 |
+
"eval_runtime": 71.5803,
|
| 3088 |
+
"eval_samples_per_second": 209.555,
|
| 3089 |
+
"eval_steps_per_second": 13.104,
|
| 3090 |
+
"step": 3900
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 1.7870201096892138,
|
| 3094 |
+
"grad_norm": 0.00673332205042243,
|
| 3095 |
+
"learning_rate": 2.13436928702011e-05,
|
| 3096 |
+
"loss": 0.0653,
|
| 3097 |
+
"step": 3910
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 1.7915904936014626,
|
| 3101 |
+
"grad_norm": 0.9469023942947388,
|
| 3102 |
+
"learning_rate": 2.0886654478976232e-05,
|
| 3103 |
+
"loss": 0.0077,
|
| 3104 |
+
"step": 3920
|
| 3105 |
+
},
|
| 3106 |
+
{
|
| 3107 |
+
"epoch": 1.796160877513711,
|
| 3108 |
+
"grad_norm": 0.03154715150594711,
|
| 3109 |
+
"learning_rate": 2.0429616087751372e-05,
|
| 3110 |
+
"loss": 0.0323,
|
| 3111 |
+
"step": 3930
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 1.8007312614259599,
|
| 3115 |
+
"grad_norm": 0.020610906183719635,
|
| 3116 |
+
"learning_rate": 1.997257769652651e-05,
|
| 3117 |
+
"loss": 0.0211,
|
| 3118 |
+
"step": 3940
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 1.8053016453382082,
|
| 3122 |
+
"grad_norm": 0.014532508328557014,
|
| 3123 |
+
"learning_rate": 1.9515539305301648e-05,
|
| 3124 |
+
"loss": 0.0357,
|
| 3125 |
+
"step": 3950
|
| 3126 |
+
},
|
| 3127 |
+
{
|
| 3128 |
+
"epoch": 1.809872029250457,
|
| 3129 |
+
"grad_norm": 0.020481685176491737,
|
| 3130 |
+
"learning_rate": 1.905850091407678e-05,
|
| 3131 |
+
"loss": 0.0498,
|
| 3132 |
+
"step": 3960
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 1.8144424131627057,
|
| 3136 |
+
"grad_norm": 0.018279431387782097,
|
| 3137 |
+
"learning_rate": 1.860146252285192e-05,
|
| 3138 |
+
"loss": 0.0559,
|
| 3139 |
+
"step": 3970
|
| 3140 |
+
},
|
| 3141 |
+
{
|
| 3142 |
+
"epoch": 1.8190127970749543,
|
| 3143 |
+
"grad_norm": 0.03680342435836792,
|
| 3144 |
+
"learning_rate": 1.8144424131627057e-05,
|
| 3145 |
+
"loss": 0.0176,
|
| 3146 |
+
"step": 3980
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 1.823583180987203,
|
| 3150 |
+
"grad_norm": 0.014550072140991688,
|
| 3151 |
+
"learning_rate": 1.7687385740402197e-05,
|
| 3152 |
+
"loss": 0.1098,
|
| 3153 |
+
"step": 3990
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 1.8281535648994516,
|
| 3157 |
+
"grad_norm": 0.01788398250937462,
|
| 3158 |
+
"learning_rate": 1.723034734917733e-05,
|
| 3159 |
+
"loss": 0.0327,
|
| 3160 |
+
"step": 4000
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 1.8281535648994516,
|
| 3164 |
+
"eval_accuracy": 0.9745333333333334,
|
| 3165 |
+
"eval_loss": 0.0925898626446724,
|
| 3166 |
+
"eval_runtime": 71.855,
|
| 3167 |
+
"eval_samples_per_second": 208.754,
|
| 3168 |
+
"eval_steps_per_second": 13.054,
|
| 3169 |
+
"step": 4000
|
| 3170 |
+
},
|
| 3171 |
+
{
|
| 3172 |
+
"epoch": 1.8327239488117002,
|
| 3173 |
+
"grad_norm": 1.894504189491272,
|
| 3174 |
+
"learning_rate": 1.677330895795247e-05,
|
| 3175 |
+
"loss": 0.0373,
|
| 3176 |
+
"step": 4010
|
| 3177 |
+
},
|
| 3178 |
+
{
|
| 3179 |
+
"epoch": 1.8372943327239488,
|
| 3180 |
+
"grad_norm": 0.005655787419527769,
|
| 3181 |
+
"learning_rate": 1.6316270566727607e-05,
|
| 3182 |
+
"loss": 0.0067,
|
| 3183 |
+
"step": 4020
|
| 3184 |
+
},
|
| 3185 |
+
{
|
| 3186 |
+
"epoch": 1.8418647166361974,
|
| 3187 |
+
"grad_norm": 4.612732410430908,
|
| 3188 |
+
"learning_rate": 1.5859232175502743e-05,
|
| 3189 |
+
"loss": 0.0183,
|
| 3190 |
+
"step": 4030
|
| 3191 |
+
},
|
| 3192 |
+
{
|
| 3193 |
+
"epoch": 1.846435100548446,
|
| 3194 |
+
"grad_norm": 0.7349024415016174,
|
| 3195 |
+
"learning_rate": 1.540219378427788e-05,
|
| 3196 |
+
"loss": 0.0021,
|
| 3197 |
+
"step": 4040
|
| 3198 |
+
},
|
| 3199 |
+
{
|
| 3200 |
+
"epoch": 1.8510054844606947,
|
| 3201 |
+
"grad_norm": 0.03837637975811958,
|
| 3202 |
+
"learning_rate": 1.4945155393053017e-05,
|
| 3203 |
+
"loss": 0.0031,
|
| 3204 |
+
"step": 4050
|
| 3205 |
+
},
|
| 3206 |
+
{
|
| 3207 |
+
"epoch": 1.8555758683729433,
|
| 3208 |
+
"grad_norm": 0.0046151746064424515,
|
| 3209 |
+
"learning_rate": 1.4488117001828156e-05,
|
| 3210 |
+
"loss": 0.1252,
|
| 3211 |
+
"step": 4060
|
| 3212 |
+
},
|
| 3213 |
+
{
|
| 3214 |
+
"epoch": 1.860146252285192,
|
| 3215 |
+
"grad_norm": 0.01960400864481926,
|
| 3216 |
+
"learning_rate": 1.403107861060329e-05,
|
| 3217 |
+
"loss": 0.0024,
|
| 3218 |
+
"step": 4070
|
| 3219 |
+
},
|
| 3220 |
+
{
|
| 3221 |
+
"epoch": 1.8647166361974405,
|
| 3222 |
+
"grad_norm": 0.012547549791634083,
|
| 3223 |
+
"learning_rate": 1.3574040219378428e-05,
|
| 3224 |
+
"loss": 0.0207,
|
| 3225 |
+
"step": 4080
|
| 3226 |
+
},
|
| 3227 |
+
{
|
| 3228 |
+
"epoch": 1.8692870201096892,
|
| 3229 |
+
"grad_norm": 0.0962536633014679,
|
| 3230 |
+
"learning_rate": 1.3117001828153566e-05,
|
| 3231 |
+
"loss": 0.0475,
|
| 3232 |
+
"step": 4090
|
| 3233 |
+
},
|
| 3234 |
+
{
|
| 3235 |
+
"epoch": 1.8738574040219378,
|
| 3236 |
+
"grad_norm": 0.012872631661593914,
|
| 3237 |
+
"learning_rate": 1.2659963436928701e-05,
|
| 3238 |
+
"loss": 0.0427,
|
| 3239 |
+
"step": 4100
|
| 3240 |
+
},
|
| 3241 |
+
{
|
| 3242 |
+
"epoch": 1.8738574040219378,
|
| 3243 |
+
"eval_accuracy": 0.9768666666666667,
|
| 3244 |
+
"eval_loss": 0.08114204555749893,
|
| 3245 |
+
"eval_runtime": 70.3536,
|
| 3246 |
+
"eval_samples_per_second": 213.209,
|
| 3247 |
+
"eval_steps_per_second": 13.333,
|
| 3248 |
+
"step": 4100
|
| 3249 |
+
},
|
| 3250 |
+
{
|
| 3251 |
+
"epoch": 1.8784277879341866,
|
| 3252 |
+
"grad_norm": 0.004386584740132093,
|
| 3253 |
+
"learning_rate": 1.220292504570384e-05,
|
| 3254 |
+
"loss": 0.0148,
|
| 3255 |
+
"step": 4110
|
| 3256 |
+
},
|
| 3257 |
+
{
|
| 3258 |
+
"epoch": 1.882998171846435,
|
| 3259 |
+
"grad_norm": 7.613697528839111,
|
| 3260 |
+
"learning_rate": 1.1745886654478977e-05,
|
| 3261 |
+
"loss": 0.0598,
|
| 3262 |
+
"step": 4120
|
| 3263 |
+
},
|
| 3264 |
+
{
|
| 3265 |
+
"epoch": 1.8875685557586839,
|
| 3266 |
+
"grad_norm": 0.19733187556266785,
|
| 3267 |
+
"learning_rate": 1.1288848263254114e-05,
|
| 3268 |
+
"loss": 0.0254,
|
| 3269 |
+
"step": 4130
|
| 3270 |
+
},
|
| 3271 |
+
{
|
| 3272 |
+
"epoch": 1.8921389396709323,
|
| 3273 |
+
"grad_norm": 0.003689356381073594,
|
| 3274 |
+
"learning_rate": 1.0831809872029252e-05,
|
| 3275 |
+
"loss": 0.0044,
|
| 3276 |
+
"step": 4140
|
| 3277 |
+
},
|
| 3278 |
+
{
|
| 3279 |
+
"epoch": 1.8967093235831811,
|
| 3280 |
+
"grad_norm": 1.0105313062667847,
|
| 3281 |
+
"learning_rate": 1.0374771480804388e-05,
|
| 3282 |
+
"loss": 0.03,
|
| 3283 |
+
"step": 4150
|
| 3284 |
+
},
|
| 3285 |
+
{
|
| 3286 |
+
"epoch": 1.9012797074954295,
|
| 3287 |
+
"grad_norm": 0.02574901282787323,
|
| 3288 |
+
"learning_rate": 9.917733089579526e-06,
|
| 3289 |
+
"loss": 0.012,
|
| 3290 |
+
"step": 4160
|
| 3291 |
+
},
|
| 3292 |
+
{
|
| 3293 |
+
"epoch": 1.9058500914076784,
|
| 3294 |
+
"grad_norm": 4.768786907196045,
|
| 3295 |
+
"learning_rate": 9.460694698354663e-06,
|
| 3296 |
+
"loss": 0.0395,
|
| 3297 |
+
"step": 4170
|
| 3298 |
+
},
|
| 3299 |
+
{
|
| 3300 |
+
"epoch": 1.9104204753199268,
|
| 3301 |
+
"grad_norm": 0.004033361561596394,
|
| 3302 |
+
"learning_rate": 9.0036563071298e-06,
|
| 3303 |
+
"loss": 0.012,
|
| 3304 |
+
"step": 4180
|
| 3305 |
+
},
|
| 3306 |
+
{
|
| 3307 |
+
"epoch": 1.9149908592321756,
|
| 3308 |
+
"grad_norm": 0.013113109394907951,
|
| 3309 |
+
"learning_rate": 8.546617915904936e-06,
|
| 3310 |
+
"loss": 0.0672,
|
| 3311 |
+
"step": 4190
|
| 3312 |
+
},
|
| 3313 |
+
{
|
| 3314 |
+
"epoch": 1.919561243144424,
|
| 3315 |
+
"grad_norm": 0.294810950756073,
|
| 3316 |
+
"learning_rate": 8.089579524680074e-06,
|
| 3317 |
+
"loss": 0.003,
|
| 3318 |
+
"step": 4200
|
| 3319 |
+
},
|
| 3320 |
+
{
|
| 3321 |
+
"epoch": 1.919561243144424,
|
| 3322 |
+
"eval_accuracy": 0.9761333333333333,
|
| 3323 |
+
"eval_loss": 0.08205202966928482,
|
| 3324 |
+
"eval_runtime": 70.0255,
|
| 3325 |
+
"eval_samples_per_second": 214.208,
|
| 3326 |
+
"eval_steps_per_second": 13.395,
|
| 3327 |
+
"step": 4200
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 1.9241316270566728,
|
| 3331 |
+
"grad_norm": 0.05837790668010712,
|
| 3332 |
+
"learning_rate": 7.63254113345521e-06,
|
| 3333 |
+
"loss": 0.0415,
|
| 3334 |
+
"step": 4210
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 1.9287020109689212,
|
| 3338 |
+
"grad_norm": 1.440628170967102,
|
| 3339 |
+
"learning_rate": 7.175502742230347e-06,
|
| 3340 |
+
"loss": 0.0267,
|
| 3341 |
+
"step": 4220
|
| 3342 |
+
},
|
| 3343 |
+
{
|
| 3344 |
+
"epoch": 1.93327239488117,
|
| 3345 |
+
"grad_norm": 0.04937027022242546,
|
| 3346 |
+
"learning_rate": 6.7184643510054855e-06,
|
| 3347 |
+
"loss": 0.0385,
|
| 3348 |
+
"step": 4230
|
| 3349 |
+
},
|
| 3350 |
+
{
|
| 3351 |
+
"epoch": 1.9378427787934185,
|
| 3352 |
+
"grad_norm": 0.006680316291749477,
|
| 3353 |
+
"learning_rate": 6.261425959780622e-06,
|
| 3354 |
+
"loss": 0.0025,
|
| 3355 |
+
"step": 4240
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 1.9424131627056673,
|
| 3359 |
+
"grad_norm": 0.004382527898997068,
|
| 3360 |
+
"learning_rate": 5.804387568555759e-06,
|
| 3361 |
+
"loss": 0.0843,
|
| 3362 |
+
"step": 4250
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 1.946983546617916,
|
| 3366 |
+
"grad_norm": 0.06263825297355652,
|
| 3367 |
+
"learning_rate": 5.3473491773308956e-06,
|
| 3368 |
+
"loss": 0.0359,
|
| 3369 |
+
"step": 4260
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 1.9515539305301646,
|
| 3373 |
+
"grad_norm": 0.017276106402277946,
|
| 3374 |
+
"learning_rate": 4.890310786106033e-06,
|
| 3375 |
+
"loss": 0.0264,
|
| 3376 |
+
"step": 4270
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 1.9561243144424132,
|
| 3380 |
+
"grad_norm": 0.7312209606170654,
|
| 3381 |
+
"learning_rate": 4.43327239488117e-06,
|
| 3382 |
+
"loss": 0.0128,
|
| 3383 |
+
"step": 4280
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 1.9606946983546618,
|
| 3387 |
+
"grad_norm": 0.007708389312028885,
|
| 3388 |
+
"learning_rate": 3.976234003656307e-06,
|
| 3389 |
+
"loss": 0.0356,
|
| 3390 |
+
"step": 4290
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 1.9652650822669104,
|
| 3394 |
+
"grad_norm": 0.004229346755892038,
|
| 3395 |
+
"learning_rate": 3.5191956124314446e-06,
|
| 3396 |
+
"loss": 0.0182,
|
| 3397 |
+
"step": 4300
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 1.9652650822669104,
|
| 3401 |
+
"eval_accuracy": 0.9772666666666666,
|
| 3402 |
+
"eval_loss": 0.08026164770126343,
|
| 3403 |
+
"eval_runtime": 69.5428,
|
| 3404 |
+
"eval_samples_per_second": 215.695,
|
| 3405 |
+
"eval_steps_per_second": 13.488,
|
| 3406 |
+
"step": 4300
|
| 3407 |
+
},
|
| 3408 |
+
{
|
| 3409 |
+
"epoch": 1.969835466179159,
|
| 3410 |
+
"grad_norm": 0.0215240940451622,
|
| 3411 |
+
"learning_rate": 3.0621572212065814e-06,
|
| 3412 |
+
"loss": 0.062,
|
| 3413 |
+
"step": 4310
|
| 3414 |
+
},
|
| 3415 |
+
{
|
| 3416 |
+
"epoch": 1.9744058500914077,
|
| 3417 |
+
"grad_norm": 0.022770356386899948,
|
| 3418 |
+
"learning_rate": 2.6051188299817187e-06,
|
| 3419 |
+
"loss": 0.0424,
|
| 3420 |
+
"step": 4320
|
| 3421 |
+
},
|
| 3422 |
+
{
|
| 3423 |
+
"epoch": 1.9789762340036563,
|
| 3424 |
+
"grad_norm": 0.01893909089267254,
|
| 3425 |
+
"learning_rate": 2.148080438756856e-06,
|
| 3426 |
+
"loss": 0.0031,
|
| 3427 |
+
"step": 4330
|
| 3428 |
+
},
|
| 3429 |
+
{
|
| 3430 |
+
"epoch": 1.983546617915905,
|
| 3431 |
+
"grad_norm": 7.610752105712891,
|
| 3432 |
+
"learning_rate": 1.691042047531993e-06,
|
| 3433 |
+
"loss": 0.0827,
|
| 3434 |
+
"step": 4340
|
| 3435 |
+
},
|
| 3436 |
+
{
|
| 3437 |
+
"epoch": 1.9881170018281535,
|
| 3438 |
+
"grad_norm": 0.008086251094937325,
|
| 3439 |
+
"learning_rate": 1.2340036563071298e-06,
|
| 3440 |
+
"loss": 0.0087,
|
| 3441 |
+
"step": 4350
|
| 3442 |
+
},
|
| 3443 |
+
{
|
| 3444 |
+
"epoch": 1.9926873857404022,
|
| 3445 |
+
"grad_norm": 4.746099948883057,
|
| 3446 |
+
"learning_rate": 7.769652650822669e-07,
|
| 3447 |
+
"loss": 0.0487,
|
| 3448 |
+
"step": 4360
|
| 3449 |
+
},
|
| 3450 |
+
{
|
| 3451 |
+
"epoch": 1.9972577696526508,
|
| 3452 |
+
"grad_norm": 0.3847046494483948,
|
| 3453 |
+
"learning_rate": 3.1992687385740404e-07,
|
| 3454 |
+
"loss": 0.0442,
|
| 3455 |
+
"step": 4370
|
| 3456 |
+
},
|
| 3457 |
+
{
|
| 3458 |
+
"epoch": 2.0,
|
| 3459 |
+
"step": 4376,
|
| 3460 |
+
"total_flos": 5.42482821328896e+18,
|
| 3461 |
+
"train_loss": 0.13969962793986365,
|
| 3462 |
+
"train_runtime": 4718.8358,
|
| 3463 |
+
"train_samples_per_second": 14.834,
|
| 3464 |
+
"train_steps_per_second": 0.927
|
| 3465 |
+
},
|
| 3466 |
+
{
|
| 3467 |
+
"epoch": 2.0,
|
| 3468 |
+
"eval_accuracy": 0.9772666666666666,
|
| 3469 |
+
"eval_loss": 0.08026164770126343,
|
| 3470 |
+
"eval_runtime": 73.8075,
|
| 3471 |
+
"eval_samples_per_second": 203.231,
|
| 3472 |
+
"eval_steps_per_second": 12.709,
|
| 3473 |
+
"step": 4376
|
| 3474 |
+
}
|
| 3475 |
+
],
|
| 3476 |
+
"logging_steps": 10,
|
| 3477 |
+
"max_steps": 4376,
|
| 3478 |
+
"num_input_tokens_seen": 0,
|
| 3479 |
+
"num_train_epochs": 2,
|
| 3480 |
+
"save_steps": 100,
|
| 3481 |
+
"stateful_callbacks": {
|
| 3482 |
+
"TrainerControl": {
|
| 3483 |
+
"args": {
|
| 3484 |
+
"should_epoch_stop": false,
|
| 3485 |
+
"should_evaluate": false,
|
| 3486 |
+
"should_log": false,
|
| 3487 |
+
"should_save": true,
|
| 3488 |
+
"should_training_stop": false
|
| 3489 |
+
},
|
| 3490 |
+
"attributes": {}
|
| 3491 |
+
}
|
| 3492 |
+
},
|
| 3493 |
+
"total_flos": 5.42482821328896e+18,
|
| 3494 |
+
"train_batch_size": 16,
|
| 3495 |
+
"trial_name": null,
|
| 3496 |
+
"trial_params": null
|
| 3497 |
+
}
|