End of training
Browse files- all_results.json +13 -0
- eval_results.json +8 -0
- runs/Dec21_21-01-53_3daf2584bc44/events.out.tfevents.1703212113.3daf2584bc44.599.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +2632 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.6788575409265064,
|
| 4 |
+
"eval_loss": 0.8812496662139893,
|
| 5 |
+
"eval_runtime": 39.1877,
|
| 6 |
+
"eval_samples_per_second": 73.263,
|
| 7 |
+
"eval_steps_per_second": 2.297,
|
| 8 |
+
"total_flos": 4.0029439499828675e+19,
|
| 9 |
+
"train_loss": 1.115002474690428,
|
| 10 |
+
"train_runtime": 19507.0594,
|
| 11 |
+
"train_samples_per_second": 26.491,
|
| 12 |
+
"train_steps_per_second": 0.207
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.6788575409265064,
|
| 4 |
+
"eval_loss": 0.8812496662139893,
|
| 5 |
+
"eval_runtime": 39.1877,
|
| 6 |
+
"eval_samples_per_second": 73.263,
|
| 7 |
+
"eval_steps_per_second": 2.297
|
| 8 |
+
}
|
runs/Dec21_21-01-53_3daf2584bc44/events.out.tfevents.1703212113.3daf2584bc44.599.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:34162f0f29b219056f1b29e911d14bf5d47842270eb70a4604414c8a03990eac
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 20.0,
|
| 3 |
+
"total_flos": 4.0029439499828675e+19,
|
| 4 |
+
"train_loss": 1.115002474690428,
|
| 5 |
+
"train_runtime": 19507.0594,
|
| 6 |
+
"train_samples_per_second": 26.491,
|
| 7 |
+
"train_steps_per_second": 0.207
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2632 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.6788575409265064,
|
| 3 |
+
"best_model_checkpoint": "finetuned-FER2013/checkpoint-4040",
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4040,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05,
|
| 13 |
+
"learning_rate": 1.2376237623762377e-07,
|
| 14 |
+
"loss": 2.073,
|
| 15 |
+
"step": 10
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.1,
|
| 19 |
+
"learning_rate": 2.4752475247524754e-07,
|
| 20 |
+
"loss": 2.1018,
|
| 21 |
+
"step": 20
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.15,
|
| 25 |
+
"learning_rate": 3.712871287128713e-07,
|
| 26 |
+
"loss": 2.0777,
|
| 27 |
+
"step": 30
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.2,
|
| 31 |
+
"learning_rate": 4.950495049504951e-07,
|
| 32 |
+
"loss": 2.0483,
|
| 33 |
+
"step": 40
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.25,
|
| 37 |
+
"learning_rate": 6.188118811881188e-07,
|
| 38 |
+
"loss": 2.0325,
|
| 39 |
+
"step": 50
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.3,
|
| 43 |
+
"learning_rate": 7.425742574257426e-07,
|
| 44 |
+
"loss": 1.9965,
|
| 45 |
+
"step": 60
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.35,
|
| 49 |
+
"learning_rate": 8.663366336633663e-07,
|
| 50 |
+
"loss": 1.9529,
|
| 51 |
+
"step": 70
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.4,
|
| 55 |
+
"learning_rate": 9.900990099009902e-07,
|
| 56 |
+
"loss": 1.9238,
|
| 57 |
+
"step": 80
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.45,
|
| 61 |
+
"learning_rate": 1.113861386138614e-06,
|
| 62 |
+
"loss": 1.9034,
|
| 63 |
+
"step": 90
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.5,
|
| 67 |
+
"learning_rate": 1.2376237623762377e-06,
|
| 68 |
+
"loss": 1.8819,
|
| 69 |
+
"step": 100
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.54,
|
| 73 |
+
"learning_rate": 1.3613861386138616e-06,
|
| 74 |
+
"loss": 1.8242,
|
| 75 |
+
"step": 110
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.59,
|
| 79 |
+
"learning_rate": 1.4851485148514852e-06,
|
| 80 |
+
"loss": 1.8252,
|
| 81 |
+
"step": 120
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.64,
|
| 85 |
+
"learning_rate": 1.6089108910891091e-06,
|
| 86 |
+
"loss": 1.7892,
|
| 87 |
+
"step": 130
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.69,
|
| 91 |
+
"learning_rate": 1.7326732673267326e-06,
|
| 92 |
+
"loss": 1.7766,
|
| 93 |
+
"step": 140
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.74,
|
| 97 |
+
"learning_rate": 1.8564356435643566e-06,
|
| 98 |
+
"loss": 1.7322,
|
| 99 |
+
"step": 150
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.79,
|
| 103 |
+
"learning_rate": 1.9801980198019803e-06,
|
| 104 |
+
"loss": 1.6977,
|
| 105 |
+
"step": 160
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.84,
|
| 109 |
+
"learning_rate": 2.103960396039604e-06,
|
| 110 |
+
"loss": 1.6751,
|
| 111 |
+
"step": 170
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.89,
|
| 115 |
+
"learning_rate": 2.227722772277228e-06,
|
| 116 |
+
"loss": 1.6337,
|
| 117 |
+
"step": 180
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.94,
|
| 121 |
+
"learning_rate": 2.3514851485148514e-06,
|
| 122 |
+
"loss": 1.5947,
|
| 123 |
+
"step": 190
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.99,
|
| 127 |
+
"learning_rate": 2.4752475247524753e-06,
|
| 128 |
+
"loss": 1.5466,
|
| 129 |
+
"step": 200
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.0,
|
| 133 |
+
"eval_accuracy": 0.4500174155346569,
|
| 134 |
+
"eval_loss": 1.502170443534851,
|
| 135 |
+
"eval_runtime": 38.4158,
|
| 136 |
+
"eval_samples_per_second": 74.735,
|
| 137 |
+
"eval_steps_per_second": 2.343,
|
| 138 |
+
"step": 202
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 1.04,
|
| 142 |
+
"learning_rate": 2.5990099009900993e-06,
|
| 143 |
+
"loss": 1.5667,
|
| 144 |
+
"step": 210
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.09,
|
| 148 |
+
"learning_rate": 2.7227722772277232e-06,
|
| 149 |
+
"loss": 1.551,
|
| 150 |
+
"step": 220
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.14,
|
| 154 |
+
"learning_rate": 2.8465346534653464e-06,
|
| 155 |
+
"loss": 1.5325,
|
| 156 |
+
"step": 230
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.19,
|
| 160 |
+
"learning_rate": 2.9702970297029703e-06,
|
| 161 |
+
"loss": 1.5274,
|
| 162 |
+
"step": 240
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 1.24,
|
| 166 |
+
"learning_rate": 3.0940594059405943e-06,
|
| 167 |
+
"loss": 1.4687,
|
| 168 |
+
"step": 250
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.29,
|
| 172 |
+
"learning_rate": 3.2178217821782182e-06,
|
| 173 |
+
"loss": 1.4933,
|
| 174 |
+
"step": 260
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.34,
|
| 178 |
+
"learning_rate": 3.341584158415842e-06,
|
| 179 |
+
"loss": 1.4407,
|
| 180 |
+
"step": 270
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 1.39,
|
| 184 |
+
"learning_rate": 3.4653465346534653e-06,
|
| 185 |
+
"loss": 1.4817,
|
| 186 |
+
"step": 280
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 1.44,
|
| 190 |
+
"learning_rate": 3.5891089108910892e-06,
|
| 191 |
+
"loss": 1.4594,
|
| 192 |
+
"step": 290
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.49,
|
| 196 |
+
"learning_rate": 3.712871287128713e-06,
|
| 197 |
+
"loss": 1.4494,
|
| 198 |
+
"step": 300
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.53,
|
| 202 |
+
"learning_rate": 3.836633663366337e-06,
|
| 203 |
+
"loss": 1.4094,
|
| 204 |
+
"step": 310
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.58,
|
| 208 |
+
"learning_rate": 3.960396039603961e-06,
|
| 209 |
+
"loss": 1.4515,
|
| 210 |
+
"step": 320
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.63,
|
| 214 |
+
"learning_rate": 4.084158415841584e-06,
|
| 215 |
+
"loss": 1.3378,
|
| 216 |
+
"step": 330
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 1.68,
|
| 220 |
+
"learning_rate": 4.207920792079208e-06,
|
| 221 |
+
"loss": 1.3925,
|
| 222 |
+
"step": 340
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 1.73,
|
| 226 |
+
"learning_rate": 4.331683168316832e-06,
|
| 227 |
+
"loss": 1.3663,
|
| 228 |
+
"step": 350
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 1.78,
|
| 232 |
+
"learning_rate": 4.455445544554456e-06,
|
| 233 |
+
"loss": 1.3544,
|
| 234 |
+
"step": 360
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.83,
|
| 238 |
+
"learning_rate": 4.57920792079208e-06,
|
| 239 |
+
"loss": 1.3758,
|
| 240 |
+
"step": 370
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.88,
|
| 244 |
+
"learning_rate": 4.702970297029703e-06,
|
| 245 |
+
"loss": 1.3395,
|
| 246 |
+
"step": 380
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.93,
|
| 250 |
+
"learning_rate": 4.826732673267327e-06,
|
| 251 |
+
"loss": 1.328,
|
| 252 |
+
"step": 390
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.98,
|
| 256 |
+
"learning_rate": 4.950495049504951e-06,
|
| 257 |
+
"loss": 1.3372,
|
| 258 |
+
"step": 400
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.0,
|
| 262 |
+
"eval_accuracy": 0.5632183908045977,
|
| 263 |
+
"eval_loss": 1.1727259159088135,
|
| 264 |
+
"eval_runtime": 38.6928,
|
| 265 |
+
"eval_samples_per_second": 74.2,
|
| 266 |
+
"eval_steps_per_second": 2.326,
|
| 267 |
+
"step": 404
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 2.03,
|
| 271 |
+
"learning_rate": 4.991749174917492e-06,
|
| 272 |
+
"loss": 1.313,
|
| 273 |
+
"step": 410
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 2.08,
|
| 277 |
+
"learning_rate": 4.977997799779978e-06,
|
| 278 |
+
"loss": 1.3004,
|
| 279 |
+
"step": 420
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 2.13,
|
| 283 |
+
"learning_rate": 4.964246424642465e-06,
|
| 284 |
+
"loss": 1.2099,
|
| 285 |
+
"step": 430
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 2.18,
|
| 289 |
+
"learning_rate": 4.950495049504951e-06,
|
| 290 |
+
"loss": 1.311,
|
| 291 |
+
"step": 440
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.23,
|
| 295 |
+
"learning_rate": 4.936743674367437e-06,
|
| 296 |
+
"loss": 1.2735,
|
| 297 |
+
"step": 450
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.28,
|
| 301 |
+
"learning_rate": 4.922992299229923e-06,
|
| 302 |
+
"loss": 1.2598,
|
| 303 |
+
"step": 460
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 2.33,
|
| 307 |
+
"learning_rate": 4.90924092409241e-06,
|
| 308 |
+
"loss": 1.2941,
|
| 309 |
+
"step": 470
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.38,
|
| 313 |
+
"learning_rate": 4.8954895489548956e-06,
|
| 314 |
+
"loss": 1.2459,
|
| 315 |
+
"step": 480
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 2.43,
|
| 319 |
+
"learning_rate": 4.881738173817382e-06,
|
| 320 |
+
"loss": 1.2597,
|
| 321 |
+
"step": 490
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.48,
|
| 325 |
+
"learning_rate": 4.867986798679869e-06,
|
| 326 |
+
"loss": 1.2726,
|
| 327 |
+
"step": 500
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 2.52,
|
| 331 |
+
"learning_rate": 4.854235423542355e-06,
|
| 332 |
+
"loss": 1.3236,
|
| 333 |
+
"step": 510
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.57,
|
| 337 |
+
"learning_rate": 4.840484048404841e-06,
|
| 338 |
+
"loss": 1.2716,
|
| 339 |
+
"step": 520
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.62,
|
| 343 |
+
"learning_rate": 4.826732673267327e-06,
|
| 344 |
+
"loss": 1.2441,
|
| 345 |
+
"step": 530
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.67,
|
| 349 |
+
"learning_rate": 4.812981298129814e-06,
|
| 350 |
+
"loss": 1.2843,
|
| 351 |
+
"step": 540
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 2.72,
|
| 355 |
+
"learning_rate": 4.7992299229923e-06,
|
| 356 |
+
"loss": 1.1991,
|
| 357 |
+
"step": 550
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 2.77,
|
| 361 |
+
"learning_rate": 4.785478547854786e-06,
|
| 362 |
+
"loss": 1.2845,
|
| 363 |
+
"step": 560
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 2.82,
|
| 367 |
+
"learning_rate": 4.771727172717272e-06,
|
| 368 |
+
"loss": 1.2553,
|
| 369 |
+
"step": 570
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.87,
|
| 373 |
+
"learning_rate": 4.757975797579759e-06,
|
| 374 |
+
"loss": 1.2146,
|
| 375 |
+
"step": 580
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.92,
|
| 379 |
+
"learning_rate": 4.7442244224422445e-06,
|
| 380 |
+
"loss": 1.2662,
|
| 381 |
+
"step": 590
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.97,
|
| 385 |
+
"learning_rate": 4.730473047304731e-06,
|
| 386 |
+
"loss": 1.2372,
|
| 387 |
+
"step": 600
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 3.0,
|
| 391 |
+
"eval_accuracy": 0.6074538488331592,
|
| 392 |
+
"eval_loss": 1.0635906457901,
|
| 393 |
+
"eval_runtime": 38.5764,
|
| 394 |
+
"eval_samples_per_second": 74.424,
|
| 395 |
+
"eval_steps_per_second": 2.333,
|
| 396 |
+
"step": 606
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 3.02,
|
| 400 |
+
"learning_rate": 4.716721672167217e-06,
|
| 401 |
+
"loss": 1.2289,
|
| 402 |
+
"step": 610
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 3.07,
|
| 406 |
+
"learning_rate": 4.702970297029703e-06,
|
| 407 |
+
"loss": 1.2239,
|
| 408 |
+
"step": 620
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 3.12,
|
| 412 |
+
"learning_rate": 4.689218921892189e-06,
|
| 413 |
+
"loss": 1.2212,
|
| 414 |
+
"step": 630
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.17,
|
| 418 |
+
"learning_rate": 4.675467546754676e-06,
|
| 419 |
+
"loss": 1.2033,
|
| 420 |
+
"step": 640
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 3.22,
|
| 424 |
+
"learning_rate": 4.661716171617162e-06,
|
| 425 |
+
"loss": 1.2295,
|
| 426 |
+
"step": 650
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 3.27,
|
| 430 |
+
"learning_rate": 4.6479647964796485e-06,
|
| 431 |
+
"loss": 1.2271,
|
| 432 |
+
"step": 660
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 3.32,
|
| 436 |
+
"learning_rate": 4.634213421342134e-06,
|
| 437 |
+
"loss": 1.2193,
|
| 438 |
+
"step": 670
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 3.37,
|
| 442 |
+
"learning_rate": 4.620462046204621e-06,
|
| 443 |
+
"loss": 1.2324,
|
| 444 |
+
"step": 680
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 3.42,
|
| 448 |
+
"learning_rate": 4.606710671067108e-06,
|
| 449 |
+
"loss": 1.1368,
|
| 450 |
+
"step": 690
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 3.47,
|
| 454 |
+
"learning_rate": 4.592959295929593e-06,
|
| 455 |
+
"loss": 1.1451,
|
| 456 |
+
"step": 700
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 3.51,
|
| 460 |
+
"learning_rate": 4.57920792079208e-06,
|
| 461 |
+
"loss": 1.1696,
|
| 462 |
+
"step": 710
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 3.56,
|
| 466 |
+
"learning_rate": 4.565456545654566e-06,
|
| 467 |
+
"loss": 1.2064,
|
| 468 |
+
"step": 720
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 3.61,
|
| 472 |
+
"learning_rate": 4.5517051705170525e-06,
|
| 473 |
+
"loss": 1.2161,
|
| 474 |
+
"step": 730
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 3.66,
|
| 478 |
+
"learning_rate": 4.537953795379538e-06,
|
| 479 |
+
"loss": 1.1665,
|
| 480 |
+
"step": 740
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 3.71,
|
| 484 |
+
"learning_rate": 4.524202420242025e-06,
|
| 485 |
+
"loss": 1.165,
|
| 486 |
+
"step": 750
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 3.76,
|
| 490 |
+
"learning_rate": 4.510451045104511e-06,
|
| 491 |
+
"loss": 1.232,
|
| 492 |
+
"step": 760
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 3.81,
|
| 496 |
+
"learning_rate": 4.4966996699669966e-06,
|
| 497 |
+
"loss": 1.2113,
|
| 498 |
+
"step": 770
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 3.86,
|
| 502 |
+
"learning_rate": 4.482948294829483e-06,
|
| 503 |
+
"loss": 1.1504,
|
| 504 |
+
"step": 780
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 3.91,
|
| 508 |
+
"learning_rate": 4.469196919691969e-06,
|
| 509 |
+
"loss": 1.1573,
|
| 510 |
+
"step": 790
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 3.96,
|
| 514 |
+
"learning_rate": 4.455445544554456e-06,
|
| 515 |
+
"loss": 1.2096,
|
| 516 |
+
"step": 800
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 4.0,
|
| 520 |
+
"eval_accuracy": 0.6116335771508186,
|
| 521 |
+
"eval_loss": 1.019976019859314,
|
| 522 |
+
"eval_runtime": 38.619,
|
| 523 |
+
"eval_samples_per_second": 74.342,
|
| 524 |
+
"eval_steps_per_second": 2.33,
|
| 525 |
+
"step": 808
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 4.01,
|
| 529 |
+
"learning_rate": 4.4416941694169415e-06,
|
| 530 |
+
"loss": 1.1387,
|
| 531 |
+
"step": 810
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 4.06,
|
| 535 |
+
"learning_rate": 4.427942794279428e-06,
|
| 536 |
+
"loss": 1.1161,
|
| 537 |
+
"step": 820
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.11,
|
| 541 |
+
"learning_rate": 4.414191419141915e-06,
|
| 542 |
+
"loss": 1.1883,
|
| 543 |
+
"step": 830
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 4.16,
|
| 547 |
+
"learning_rate": 4.400440044004401e-06,
|
| 548 |
+
"loss": 1.1215,
|
| 549 |
+
"step": 840
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 4.21,
|
| 553 |
+
"learning_rate": 4.386688668866887e-06,
|
| 554 |
+
"loss": 1.1195,
|
| 555 |
+
"step": 850
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 4.26,
|
| 559 |
+
"learning_rate": 4.372937293729374e-06,
|
| 560 |
+
"loss": 1.1991,
|
| 561 |
+
"step": 860
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 4.31,
|
| 565 |
+
"learning_rate": 4.35918591859186e-06,
|
| 566 |
+
"loss": 1.1645,
|
| 567 |
+
"step": 870
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 4.36,
|
| 571 |
+
"learning_rate": 4.345434543454346e-06,
|
| 572 |
+
"loss": 1.1446,
|
| 573 |
+
"step": 880
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 4.41,
|
| 577 |
+
"learning_rate": 4.331683168316832e-06,
|
| 578 |
+
"loss": 1.1308,
|
| 579 |
+
"step": 890
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 4.46,
|
| 583 |
+
"learning_rate": 4.317931793179319e-06,
|
| 584 |
+
"loss": 1.1758,
|
| 585 |
+
"step": 900
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 4.5,
|
| 589 |
+
"learning_rate": 4.304180418041805e-06,
|
| 590 |
+
"loss": 1.1254,
|
| 591 |
+
"step": 910
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 4.55,
|
| 595 |
+
"learning_rate": 4.29042904290429e-06,
|
| 596 |
+
"loss": 1.1217,
|
| 597 |
+
"step": 920
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 4.6,
|
| 601 |
+
"learning_rate": 4.276677667766777e-06,
|
| 602 |
+
"loss": 1.1633,
|
| 603 |
+
"step": 930
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 4.65,
|
| 607 |
+
"learning_rate": 4.262926292629263e-06,
|
| 608 |
+
"loss": 1.1553,
|
| 609 |
+
"step": 940
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 4.7,
|
| 613 |
+
"learning_rate": 4.2491749174917495e-06,
|
| 614 |
+
"loss": 1.1764,
|
| 615 |
+
"step": 950
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 4.75,
|
| 619 |
+
"learning_rate": 4.235423542354235e-06,
|
| 620 |
+
"loss": 1.1058,
|
| 621 |
+
"step": 960
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 4.8,
|
| 625 |
+
"learning_rate": 4.221672167216722e-06,
|
| 626 |
+
"loss": 1.1587,
|
| 627 |
+
"step": 970
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 4.85,
|
| 631 |
+
"learning_rate": 4.207920792079208e-06,
|
| 632 |
+
"loss": 1.1889,
|
| 633 |
+
"step": 980
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 4.9,
|
| 637 |
+
"learning_rate": 4.194169416941694e-06,
|
| 638 |
+
"loss": 1.1176,
|
| 639 |
+
"step": 990
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 4.95,
|
| 643 |
+
"learning_rate": 4.180418041804181e-06,
|
| 644 |
+
"loss": 1.1319,
|
| 645 |
+
"step": 1000
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 5.0,
|
| 649 |
+
"learning_rate": 4.166666666666667e-06,
|
| 650 |
+
"loss": 1.145,
|
| 651 |
+
"step": 1010
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 5.0,
|
| 655 |
+
"eval_accuracy": 0.6325322187391152,
|
| 656 |
+
"eval_loss": 0.9768889546394348,
|
| 657 |
+
"eval_runtime": 38.6786,
|
| 658 |
+
"eval_samples_per_second": 74.227,
|
| 659 |
+
"eval_steps_per_second": 2.327,
|
| 660 |
+
"step": 1010
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 5.05,
|
| 664 |
+
"learning_rate": 4.1529152915291535e-06,
|
| 665 |
+
"loss": 1.1321,
|
| 666 |
+
"step": 1020
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 5.1,
|
| 670 |
+
"learning_rate": 4.139163916391639e-06,
|
| 671 |
+
"loss": 1.1268,
|
| 672 |
+
"step": 1030
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 5.15,
|
| 676 |
+
"learning_rate": 4.125412541254126e-06,
|
| 677 |
+
"loss": 1.1032,
|
| 678 |
+
"step": 1040
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 5.2,
|
| 682 |
+
"learning_rate": 4.111661166116613e-06,
|
| 683 |
+
"loss": 1.1327,
|
| 684 |
+
"step": 1050
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 5.25,
|
| 688 |
+
"learning_rate": 4.0979097909790984e-06,
|
| 689 |
+
"loss": 1.141,
|
| 690 |
+
"step": 1060
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 5.3,
|
| 694 |
+
"learning_rate": 4.084158415841584e-06,
|
| 695 |
+
"loss": 1.1025,
|
| 696 |
+
"step": 1070
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 5.35,
|
| 700 |
+
"learning_rate": 4.070407040704071e-06,
|
| 701 |
+
"loss": 1.1138,
|
| 702 |
+
"step": 1080
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 5.4,
|
| 706 |
+
"learning_rate": 4.056655665566557e-06,
|
| 707 |
+
"loss": 1.122,
|
| 708 |
+
"step": 1090
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 5.45,
|
| 712 |
+
"learning_rate": 4.042904290429043e-06,
|
| 713 |
+
"loss": 1.0642,
|
| 714 |
+
"step": 1100
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 5.5,
|
| 718 |
+
"learning_rate": 4.029152915291529e-06,
|
| 719 |
+
"loss": 1.1162,
|
| 720 |
+
"step": 1110
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 5.54,
|
| 724 |
+
"learning_rate": 4.015401540154016e-06,
|
| 725 |
+
"loss": 1.0873,
|
| 726 |
+
"step": 1120
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 5.59,
|
| 730 |
+
"learning_rate": 4.001650165016502e-06,
|
| 731 |
+
"loss": 1.1188,
|
| 732 |
+
"step": 1130
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 5.64,
|
| 736 |
+
"learning_rate": 3.987898789878988e-06,
|
| 737 |
+
"loss": 1.0996,
|
| 738 |
+
"step": 1140
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 5.69,
|
| 742 |
+
"learning_rate": 3.974147414741474e-06,
|
| 743 |
+
"loss": 1.1377,
|
| 744 |
+
"step": 1150
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 5.74,
|
| 748 |
+
"learning_rate": 3.960396039603961e-06,
|
| 749 |
+
"loss": 1.0857,
|
| 750 |
+
"step": 1160
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 5.79,
|
| 754 |
+
"learning_rate": 3.9466446644664465e-06,
|
| 755 |
+
"loss": 1.0875,
|
| 756 |
+
"step": 1170
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 5.84,
|
| 760 |
+
"learning_rate": 3.932893289328933e-06,
|
| 761 |
+
"loss": 1.0532,
|
| 762 |
+
"step": 1180
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 5.89,
|
| 766 |
+
"learning_rate": 3.91914191419142e-06,
|
| 767 |
+
"loss": 1.1319,
|
| 768 |
+
"step": 1190
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 5.94,
|
| 772 |
+
"learning_rate": 3.905390539053906e-06,
|
| 773 |
+
"loss": 1.1125,
|
| 774 |
+
"step": 1200
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 5.99,
|
| 778 |
+
"learning_rate": 3.891639163916392e-06,
|
| 779 |
+
"loss": 1.1589,
|
| 780 |
+
"step": 1210
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 6.0,
|
| 784 |
+
"eval_accuracy": 0.6405433646812957,
|
| 785 |
+
"eval_loss": 0.9515116214752197,
|
| 786 |
+
"eval_runtime": 38.8551,
|
| 787 |
+
"eval_samples_per_second": 73.89,
|
| 788 |
+
"eval_steps_per_second": 2.316,
|
| 789 |
+
"step": 1212
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 6.04,
|
| 793 |
+
"learning_rate": 3.877887788778878e-06,
|
| 794 |
+
"loss": 1.1213,
|
| 795 |
+
"step": 1220
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 6.09,
|
| 799 |
+
"learning_rate": 3.864136413641365e-06,
|
| 800 |
+
"loss": 1.0874,
|
| 801 |
+
"step": 1230
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 6.14,
|
| 805 |
+
"learning_rate": 3.8503850385038505e-06,
|
| 806 |
+
"loss": 1.0901,
|
| 807 |
+
"step": 1240
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 6.19,
|
| 811 |
+
"learning_rate": 3.836633663366337e-06,
|
| 812 |
+
"loss": 1.0979,
|
| 813 |
+
"step": 1250
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 6.24,
|
| 817 |
+
"learning_rate": 3.822882288228823e-06,
|
| 818 |
+
"loss": 1.074,
|
| 819 |
+
"step": 1260
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 6.29,
|
| 823 |
+
"learning_rate": 3.8091309130913096e-06,
|
| 824 |
+
"loss": 1.0936,
|
| 825 |
+
"step": 1270
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 6.34,
|
| 829 |
+
"learning_rate": 3.7953795379537954e-06,
|
| 830 |
+
"loss": 1.071,
|
| 831 |
+
"step": 1280
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 6.39,
|
| 835 |
+
"learning_rate": 3.781628162816282e-06,
|
| 836 |
+
"loss": 1.1341,
|
| 837 |
+
"step": 1290
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 6.44,
|
| 841 |
+
"learning_rate": 3.767876787678768e-06,
|
| 842 |
+
"loss": 1.0803,
|
| 843 |
+
"step": 1300
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 6.49,
|
| 847 |
+
"learning_rate": 3.7541254125412545e-06,
|
| 848 |
+
"loss": 1.1026,
|
| 849 |
+
"step": 1310
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 6.53,
|
| 853 |
+
"learning_rate": 3.7403740374037408e-06,
|
| 854 |
+
"loss": 1.0736,
|
| 855 |
+
"step": 1320
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 6.58,
|
| 859 |
+
"learning_rate": 3.726622662266227e-06,
|
| 860 |
+
"loss": 1.1181,
|
| 861 |
+
"step": 1330
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 6.63,
|
| 865 |
+
"learning_rate": 3.712871287128713e-06,
|
| 866 |
+
"loss": 1.0721,
|
| 867 |
+
"step": 1340
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 6.68,
|
| 871 |
+
"learning_rate": 3.6991199119912e-06,
|
| 872 |
+
"loss": 1.0743,
|
| 873 |
+
"step": 1350
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 6.73,
|
| 877 |
+
"learning_rate": 3.6853685368536857e-06,
|
| 878 |
+
"loss": 1.1077,
|
| 879 |
+
"step": 1360
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 6.78,
|
| 883 |
+
"learning_rate": 3.6716171617161715e-06,
|
| 884 |
+
"loss": 1.0815,
|
| 885 |
+
"step": 1370
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 6.83,
|
| 889 |
+
"learning_rate": 3.657865786578658e-06,
|
| 890 |
+
"loss": 1.0834,
|
| 891 |
+
"step": 1380
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 6.88,
|
| 895 |
+
"learning_rate": 3.6441144114411443e-06,
|
| 896 |
+
"loss": 1.1362,
|
| 897 |
+
"step": 1390
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 6.93,
|
| 901 |
+
"learning_rate": 3.6303630363036306e-06,
|
| 902 |
+
"loss": 1.0713,
|
| 903 |
+
"step": 1400
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 6.98,
|
| 907 |
+
"learning_rate": 3.616611661166117e-06,
|
| 908 |
+
"loss": 1.0752,
|
| 909 |
+
"step": 1410
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 7.0,
|
| 913 |
+
"eval_accuracy": 0.64576802507837,
|
| 914 |
+
"eval_loss": 0.9395281076431274,
|
| 915 |
+
"eval_runtime": 38.3561,
|
| 916 |
+
"eval_samples_per_second": 74.851,
|
| 917 |
+
"eval_steps_per_second": 2.346,
|
| 918 |
+
"step": 1414
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 7.03,
|
| 922 |
+
"learning_rate": 3.6028602860286034e-06,
|
| 923 |
+
"loss": 1.0504,
|
| 924 |
+
"step": 1420
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 7.08,
|
| 928 |
+
"learning_rate": 3.5891089108910892e-06,
|
| 929 |
+
"loss": 1.1083,
|
| 930 |
+
"step": 1430
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 7.13,
|
| 934 |
+
"learning_rate": 3.575357535753576e-06,
|
| 935 |
+
"loss": 1.0908,
|
| 936 |
+
"step": 1440
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 7.18,
|
| 940 |
+
"learning_rate": 3.5616061606160617e-06,
|
| 941 |
+
"loss": 1.1186,
|
| 942 |
+
"step": 1450
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 7.23,
|
| 946 |
+
"learning_rate": 3.5478547854785484e-06,
|
| 947 |
+
"loss": 1.096,
|
| 948 |
+
"step": 1460
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 7.28,
|
| 952 |
+
"learning_rate": 3.534103410341034e-06,
|
| 953 |
+
"loss": 1.1121,
|
| 954 |
+
"step": 1470
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 7.33,
|
| 958 |
+
"learning_rate": 3.520352035203521e-06,
|
| 959 |
+
"loss": 1.0535,
|
| 960 |
+
"step": 1480
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 7.38,
|
| 964 |
+
"learning_rate": 3.506600660066007e-06,
|
| 965 |
+
"loss": 1.0403,
|
| 966 |
+
"step": 1490
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 7.43,
|
| 970 |
+
"learning_rate": 3.4928492849284933e-06,
|
| 971 |
+
"loss": 1.0904,
|
| 972 |
+
"step": 1500
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 7.48,
|
| 976 |
+
"learning_rate": 3.4790979097909795e-06,
|
| 977 |
+
"loss": 1.0987,
|
| 978 |
+
"step": 1510
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 7.52,
|
| 982 |
+
"learning_rate": 3.4653465346534653e-06,
|
| 983 |
+
"loss": 1.0786,
|
| 984 |
+
"step": 1520
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 7.57,
|
| 988 |
+
"learning_rate": 3.451595159515952e-06,
|
| 989 |
+
"loss": 1.1133,
|
| 990 |
+
"step": 1530
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 7.62,
|
| 994 |
+
"learning_rate": 3.4378437843784377e-06,
|
| 995 |
+
"loss": 1.0804,
|
| 996 |
+
"step": 1540
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 7.67,
|
| 1000 |
+
"learning_rate": 3.4240924092409244e-06,
|
| 1001 |
+
"loss": 1.0237,
|
| 1002 |
+
"step": 1550
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 7.72,
|
| 1006 |
+
"learning_rate": 3.4103410341034106e-06,
|
| 1007 |
+
"loss": 1.0276,
|
| 1008 |
+
"step": 1560
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 7.77,
|
| 1012 |
+
"learning_rate": 3.396589658965897e-06,
|
| 1013 |
+
"loss": 1.0769,
|
| 1014 |
+
"step": 1570
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 7.82,
|
| 1018 |
+
"learning_rate": 3.382838283828383e-06,
|
| 1019 |
+
"loss": 1.0962,
|
| 1020 |
+
"step": 1580
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 7.87,
|
| 1024 |
+
"learning_rate": 3.3690869086908697e-06,
|
| 1025 |
+
"loss": 1.0586,
|
| 1026 |
+
"step": 1590
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 7.92,
|
| 1030 |
+
"learning_rate": 3.3553355335533555e-06,
|
| 1031 |
+
"loss": 1.0786,
|
| 1032 |
+
"step": 1600
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 7.97,
|
| 1036 |
+
"learning_rate": 3.341584158415842e-06,
|
| 1037 |
+
"loss": 1.0524,
|
| 1038 |
+
"step": 1610
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 8.0,
|
| 1042 |
+
"eval_accuracy": 0.64576802507837,
|
| 1043 |
+
"eval_loss": 0.9330951571464539,
|
| 1044 |
+
"eval_runtime": 38.8204,
|
| 1045 |
+
"eval_samples_per_second": 73.956,
|
| 1046 |
+
"eval_steps_per_second": 2.318,
|
| 1047 |
+
"step": 1616
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 8.02,
|
| 1051 |
+
"learning_rate": 3.327832783278328e-06,
|
| 1052 |
+
"loss": 1.0425,
|
| 1053 |
+
"step": 1620
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 8.07,
|
| 1057 |
+
"learning_rate": 3.3140814081408146e-06,
|
| 1058 |
+
"loss": 1.0729,
|
| 1059 |
+
"step": 1630
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 8.12,
|
| 1063 |
+
"learning_rate": 3.3003300330033004e-06,
|
| 1064 |
+
"loss": 1.0748,
|
| 1065 |
+
"step": 1640
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 8.17,
|
| 1069 |
+
"learning_rate": 3.286578657865787e-06,
|
| 1070 |
+
"loss": 1.0766,
|
| 1071 |
+
"step": 1650
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 8.22,
|
| 1075 |
+
"learning_rate": 3.272827282728273e-06,
|
| 1076 |
+
"loss": 1.0779,
|
| 1077 |
+
"step": 1660
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 8.27,
|
| 1081 |
+
"learning_rate": 3.259075907590759e-06,
|
| 1082 |
+
"loss": 1.0591,
|
| 1083 |
+
"step": 1670
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 8.32,
|
| 1087 |
+
"learning_rate": 3.2453245324532458e-06,
|
| 1088 |
+
"loss": 1.0551,
|
| 1089 |
+
"step": 1680
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 8.37,
|
| 1093 |
+
"learning_rate": 3.2315731573157316e-06,
|
| 1094 |
+
"loss": 1.031,
|
| 1095 |
+
"step": 1690
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 8.42,
|
| 1099 |
+
"learning_rate": 3.2178217821782182e-06,
|
| 1100 |
+
"loss": 1.0348,
|
| 1101 |
+
"step": 1700
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 8.47,
|
| 1105 |
+
"learning_rate": 3.204070407040704e-06,
|
| 1106 |
+
"loss": 1.064,
|
| 1107 |
+
"step": 1710
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 8.51,
|
| 1111 |
+
"learning_rate": 3.1903190319031907e-06,
|
| 1112 |
+
"loss": 1.0142,
|
| 1113 |
+
"step": 1720
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 8.56,
|
| 1117 |
+
"learning_rate": 3.1765676567656765e-06,
|
| 1118 |
+
"loss": 1.0871,
|
| 1119 |
+
"step": 1730
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 8.61,
|
| 1123 |
+
"learning_rate": 3.162816281628163e-06,
|
| 1124 |
+
"loss": 1.0586,
|
| 1125 |
+
"step": 1740
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 8.66,
|
| 1129 |
+
"learning_rate": 3.1490649064906494e-06,
|
| 1130 |
+
"loss": 1.05,
|
| 1131 |
+
"step": 1750
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 8.71,
|
| 1135 |
+
"learning_rate": 3.1353135313531356e-06,
|
| 1136 |
+
"loss": 1.0503,
|
| 1137 |
+
"step": 1760
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 8.76,
|
| 1141 |
+
"learning_rate": 3.121562156215622e-06,
|
| 1142 |
+
"loss": 1.0318,
|
| 1143 |
+
"step": 1770
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 8.81,
|
| 1147 |
+
"learning_rate": 3.1078107810781085e-06,
|
| 1148 |
+
"loss": 1.087,
|
| 1149 |
+
"step": 1780
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 8.86,
|
| 1153 |
+
"learning_rate": 3.0940594059405943e-06,
|
| 1154 |
+
"loss": 1.0563,
|
| 1155 |
+
"step": 1790
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 8.91,
|
| 1159 |
+
"learning_rate": 3.080308030803081e-06,
|
| 1160 |
+
"loss": 1.0416,
|
| 1161 |
+
"step": 1800
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 8.96,
|
| 1165 |
+
"learning_rate": 3.0665566556655667e-06,
|
| 1166 |
+
"loss": 1.0829,
|
| 1167 |
+
"step": 1810
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 9.0,
|
| 1171 |
+
"eval_accuracy": 0.6572622779519331,
|
| 1172 |
+
"eval_loss": 0.9173280000686646,
|
| 1173 |
+
"eval_runtime": 38.5578,
|
| 1174 |
+
"eval_samples_per_second": 74.46,
|
| 1175 |
+
"eval_steps_per_second": 2.334,
|
| 1176 |
+
"step": 1818
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 9.01,
|
| 1180 |
+
"learning_rate": 3.052805280528053e-06,
|
| 1181 |
+
"loss": 1.0785,
|
| 1182 |
+
"step": 1820
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 9.06,
|
| 1186 |
+
"learning_rate": 3.039053905390539e-06,
|
| 1187 |
+
"loss": 1.0458,
|
| 1188 |
+
"step": 1830
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 9.11,
|
| 1192 |
+
"learning_rate": 3.0253025302530254e-06,
|
| 1193 |
+
"loss": 1.0499,
|
| 1194 |
+
"step": 1840
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 9.16,
|
| 1198 |
+
"learning_rate": 3.011551155115512e-06,
|
| 1199 |
+
"loss": 1.0528,
|
| 1200 |
+
"step": 1850
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 9.21,
|
| 1204 |
+
"learning_rate": 2.997799779977998e-06,
|
| 1205 |
+
"loss": 1.0891,
|
| 1206 |
+
"step": 1860
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 9.26,
|
| 1210 |
+
"learning_rate": 2.9840484048404845e-06,
|
| 1211 |
+
"loss": 1.0229,
|
| 1212 |
+
"step": 1870
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 9.31,
|
| 1216 |
+
"learning_rate": 2.9702970297029703e-06,
|
| 1217 |
+
"loss": 1.0544,
|
| 1218 |
+
"step": 1880
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 9.36,
|
| 1222 |
+
"learning_rate": 2.956545654565457e-06,
|
| 1223 |
+
"loss": 1.0305,
|
| 1224 |
+
"step": 1890
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 9.41,
|
| 1228 |
+
"learning_rate": 2.9427942794279428e-06,
|
| 1229 |
+
"loss": 1.036,
|
| 1230 |
+
"step": 1900
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 9.46,
|
| 1234 |
+
"learning_rate": 2.9290429042904294e-06,
|
| 1235 |
+
"loss": 1.0191,
|
| 1236 |
+
"step": 1910
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 9.5,
|
| 1240 |
+
"learning_rate": 2.9152915291529156e-06,
|
| 1241 |
+
"loss": 1.0612,
|
| 1242 |
+
"step": 1920
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 9.55,
|
| 1246 |
+
"learning_rate": 2.901540154015402e-06,
|
| 1247 |
+
"loss": 1.0258,
|
| 1248 |
+
"step": 1930
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 9.6,
|
| 1252 |
+
"learning_rate": 2.887788778877888e-06,
|
| 1253 |
+
"loss": 0.9895,
|
| 1254 |
+
"step": 1940
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 9.65,
|
| 1258 |
+
"learning_rate": 2.8740374037403747e-06,
|
| 1259 |
+
"loss": 1.0716,
|
| 1260 |
+
"step": 1950
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 9.7,
|
| 1264 |
+
"learning_rate": 2.8602860286028605e-06,
|
| 1265 |
+
"loss": 1.0605,
|
| 1266 |
+
"step": 1960
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 9.75,
|
| 1270 |
+
"learning_rate": 2.8465346534653464e-06,
|
| 1271 |
+
"loss": 1.0462,
|
| 1272 |
+
"step": 1970
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 9.8,
|
| 1276 |
+
"learning_rate": 2.832783278327833e-06,
|
| 1277 |
+
"loss": 1.0582,
|
| 1278 |
+
"step": 1980
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 9.85,
|
| 1282 |
+
"learning_rate": 2.8190319031903192e-06,
|
| 1283 |
+
"loss": 1.0484,
|
| 1284 |
+
"step": 1990
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 9.9,
|
| 1288 |
+
"learning_rate": 2.8052805280528055e-06,
|
| 1289 |
+
"loss": 0.9998,
|
| 1290 |
+
"step": 2000
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 9.95,
|
| 1294 |
+
"learning_rate": 2.7915291529152917e-06,
|
| 1295 |
+
"loss": 1.0437,
|
| 1296 |
+
"step": 2010
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 10.0,
|
| 1300 |
+
"learning_rate": 2.7777777777777783e-06,
|
| 1301 |
+
"loss": 1.0219,
|
| 1302 |
+
"step": 2020
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 10.0,
|
| 1306 |
+
"eval_accuracy": 0.6597004528039011,
|
| 1307 |
+
"eval_loss": 0.9113705158233643,
|
| 1308 |
+
"eval_runtime": 38.5342,
|
| 1309 |
+
"eval_samples_per_second": 74.505,
|
| 1310 |
+
"eval_steps_per_second": 2.336,
|
| 1311 |
+
"step": 2020
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 10.05,
|
| 1315 |
+
"learning_rate": 2.764026402640264e-06,
|
| 1316 |
+
"loss": 1.0749,
|
| 1317 |
+
"step": 2030
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 10.1,
|
| 1321 |
+
"learning_rate": 2.750275027502751e-06,
|
| 1322 |
+
"loss": 1.0419,
|
| 1323 |
+
"step": 2040
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 10.15,
|
| 1327 |
+
"learning_rate": 2.7365236523652366e-06,
|
| 1328 |
+
"loss": 1.0407,
|
| 1329 |
+
"step": 2050
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 10.2,
|
| 1333 |
+
"learning_rate": 2.7227722772277232e-06,
|
| 1334 |
+
"loss": 0.9984,
|
| 1335 |
+
"step": 2060
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 10.25,
|
| 1339 |
+
"learning_rate": 2.709020902090209e-06,
|
| 1340 |
+
"loss": 0.9575,
|
| 1341 |
+
"step": 2070
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 10.3,
|
| 1345 |
+
"learning_rate": 2.6952695269526957e-06,
|
| 1346 |
+
"loss": 1.021,
|
| 1347 |
+
"step": 2080
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 10.35,
|
| 1351 |
+
"learning_rate": 2.681518151815182e-06,
|
| 1352 |
+
"loss": 1.0191,
|
| 1353 |
+
"step": 2090
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 10.4,
|
| 1357 |
+
"learning_rate": 2.667766776677668e-06,
|
| 1358 |
+
"loss": 1.0502,
|
| 1359 |
+
"step": 2100
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 10.45,
|
| 1363 |
+
"learning_rate": 2.6540154015401544e-06,
|
| 1364 |
+
"loss": 1.0292,
|
| 1365 |
+
"step": 2110
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 10.5,
|
| 1369 |
+
"learning_rate": 2.64026402640264e-06,
|
| 1370 |
+
"loss": 0.9986,
|
| 1371 |
+
"step": 2120
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 10.54,
|
| 1375 |
+
"learning_rate": 2.626512651265127e-06,
|
| 1376 |
+
"loss": 0.9604,
|
| 1377 |
+
"step": 2130
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 10.59,
|
| 1381 |
+
"learning_rate": 2.6127612761276126e-06,
|
| 1382 |
+
"loss": 1.0127,
|
| 1383 |
+
"step": 2140
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 10.64,
|
| 1387 |
+
"learning_rate": 2.5990099009900993e-06,
|
| 1388 |
+
"loss": 1.048,
|
| 1389 |
+
"step": 2150
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 10.69,
|
| 1393 |
+
"learning_rate": 2.585258525852585e-06,
|
| 1394 |
+
"loss": 1.0475,
|
| 1395 |
+
"step": 2160
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 10.74,
|
| 1399 |
+
"learning_rate": 2.5715071507150717e-06,
|
| 1400 |
+
"loss": 1.0552,
|
| 1401 |
+
"step": 2170
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 10.79,
|
| 1405 |
+
"learning_rate": 2.557755775577558e-06,
|
| 1406 |
+
"loss": 1.0351,
|
| 1407 |
+
"step": 2180
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 10.84,
|
| 1411 |
+
"learning_rate": 2.544004400440044e-06,
|
| 1412 |
+
"loss": 1.0465,
|
| 1413 |
+
"step": 2190
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 10.89,
|
| 1417 |
+
"learning_rate": 2.5302530253025304e-06,
|
| 1418 |
+
"loss": 1.0807,
|
| 1419 |
+
"step": 2200
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 10.94,
|
| 1423 |
+
"learning_rate": 2.516501650165017e-06,
|
| 1424 |
+
"loss": 1.0255,
|
| 1425 |
+
"step": 2210
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 10.99,
|
| 1429 |
+
"learning_rate": 2.502750275027503e-06,
|
| 1430 |
+
"loss": 0.9986,
|
| 1431 |
+
"step": 2220
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 11.0,
|
| 1435 |
+
"eval_accuracy": 0.6579588993382097,
|
| 1436 |
+
"eval_loss": 0.9033769369125366,
|
| 1437 |
+
"eval_runtime": 38.6552,
|
| 1438 |
+
"eval_samples_per_second": 74.272,
|
| 1439 |
+
"eval_steps_per_second": 2.328,
|
| 1440 |
+
"step": 2222
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 11.04,
|
| 1444 |
+
"learning_rate": 2.488998899889989e-06,
|
| 1445 |
+
"loss": 1.0336,
|
| 1446 |
+
"step": 2230
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 11.09,
|
| 1450 |
+
"learning_rate": 2.4752475247524753e-06,
|
| 1451 |
+
"loss": 1.0238,
|
| 1452 |
+
"step": 2240
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 11.14,
|
| 1456 |
+
"learning_rate": 2.4614961496149616e-06,
|
| 1457 |
+
"loss": 1.04,
|
| 1458 |
+
"step": 2250
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 11.19,
|
| 1462 |
+
"learning_rate": 2.4477447744774478e-06,
|
| 1463 |
+
"loss": 1.0165,
|
| 1464 |
+
"step": 2260
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 11.24,
|
| 1468 |
+
"learning_rate": 2.4339933993399344e-06,
|
| 1469 |
+
"loss": 1.0155,
|
| 1470 |
+
"step": 2270
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 11.29,
|
| 1474 |
+
"learning_rate": 2.4202420242024207e-06,
|
| 1475 |
+
"loss": 1.0273,
|
| 1476 |
+
"step": 2280
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 11.34,
|
| 1480 |
+
"learning_rate": 2.406490649064907e-06,
|
| 1481 |
+
"loss": 1.0048,
|
| 1482 |
+
"step": 2290
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 11.39,
|
| 1486 |
+
"learning_rate": 2.392739273927393e-06,
|
| 1487 |
+
"loss": 0.988,
|
| 1488 |
+
"step": 2300
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 11.44,
|
| 1492 |
+
"learning_rate": 2.3789878987898793e-06,
|
| 1493 |
+
"loss": 1.0388,
|
| 1494 |
+
"step": 2310
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 11.49,
|
| 1498 |
+
"learning_rate": 2.3652365236523656e-06,
|
| 1499 |
+
"loss": 1.0389,
|
| 1500 |
+
"step": 2320
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 11.53,
|
| 1504 |
+
"learning_rate": 2.3514851485148514e-06,
|
| 1505 |
+
"loss": 1.0651,
|
| 1506 |
+
"step": 2330
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 11.58,
|
| 1510 |
+
"learning_rate": 2.337733773377338e-06,
|
| 1511 |
+
"loss": 1.0789,
|
| 1512 |
+
"step": 2340
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 11.63,
|
| 1516 |
+
"learning_rate": 2.3239823982398242e-06,
|
| 1517 |
+
"loss": 0.9802,
|
| 1518 |
+
"step": 2350
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 11.68,
|
| 1522 |
+
"learning_rate": 2.3102310231023105e-06,
|
| 1523 |
+
"loss": 1.054,
|
| 1524 |
+
"step": 2360
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 11.73,
|
| 1528 |
+
"learning_rate": 2.2964796479647967e-06,
|
| 1529 |
+
"loss": 1.0351,
|
| 1530 |
+
"step": 2370
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 11.78,
|
| 1534 |
+
"learning_rate": 2.282728272827283e-06,
|
| 1535 |
+
"loss": 1.0146,
|
| 1536 |
+
"step": 2380
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 11.83,
|
| 1540 |
+
"learning_rate": 2.268976897689769e-06,
|
| 1541 |
+
"loss": 1.0293,
|
| 1542 |
+
"step": 2390
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 11.88,
|
| 1546 |
+
"learning_rate": 2.2552255225522554e-06,
|
| 1547 |
+
"loss": 0.9827,
|
| 1548 |
+
"step": 2400
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 11.93,
|
| 1552 |
+
"learning_rate": 2.2414741474147416e-06,
|
| 1553 |
+
"loss": 0.9807,
|
| 1554 |
+
"step": 2410
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 11.98,
|
| 1558 |
+
"learning_rate": 2.227722772277228e-06,
|
| 1559 |
+
"loss": 1.013,
|
| 1560 |
+
"step": 2420
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 12.0,
|
| 1564 |
+
"eval_accuracy": 0.6656217345872518,
|
| 1565 |
+
"eval_loss": 0.9003686904907227,
|
| 1566 |
+
"eval_runtime": 38.455,
|
| 1567 |
+
"eval_samples_per_second": 74.659,
|
| 1568 |
+
"eval_steps_per_second": 2.34,
|
| 1569 |
+
"step": 2424
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 12.03,
|
| 1573 |
+
"learning_rate": 2.213971397139714e-06,
|
| 1574 |
+
"loss": 1.0289,
|
| 1575 |
+
"step": 2430
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 12.08,
|
| 1579 |
+
"learning_rate": 2.2002200220022003e-06,
|
| 1580 |
+
"loss": 1.0303,
|
| 1581 |
+
"step": 2440
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 12.13,
|
| 1585 |
+
"learning_rate": 2.186468646864687e-06,
|
| 1586 |
+
"loss": 1.0304,
|
| 1587 |
+
"step": 2450
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 12.18,
|
| 1591 |
+
"learning_rate": 2.172717271727173e-06,
|
| 1592 |
+
"loss": 1.0318,
|
| 1593 |
+
"step": 2460
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 12.23,
|
| 1597 |
+
"learning_rate": 2.1589658965896594e-06,
|
| 1598 |
+
"loss": 1.0566,
|
| 1599 |
+
"step": 2470
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 12.28,
|
| 1603 |
+
"learning_rate": 2.145214521452145e-06,
|
| 1604 |
+
"loss": 1.0052,
|
| 1605 |
+
"step": 2480
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 12.33,
|
| 1609 |
+
"learning_rate": 2.1314631463146314e-06,
|
| 1610 |
+
"loss": 0.9798,
|
| 1611 |
+
"step": 2490
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 12.38,
|
| 1615 |
+
"learning_rate": 2.1177117711771177e-06,
|
| 1616 |
+
"loss": 1.0139,
|
| 1617 |
+
"step": 2500
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 12.43,
|
| 1621 |
+
"learning_rate": 2.103960396039604e-06,
|
| 1622 |
+
"loss": 0.9932,
|
| 1623 |
+
"step": 2510
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 12.48,
|
| 1627 |
+
"learning_rate": 2.0902090209020905e-06,
|
| 1628 |
+
"loss": 1.0196,
|
| 1629 |
+
"step": 2520
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 12.52,
|
| 1633 |
+
"learning_rate": 2.0764576457645768e-06,
|
| 1634 |
+
"loss": 1.0261,
|
| 1635 |
+
"step": 2530
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 12.57,
|
| 1639 |
+
"learning_rate": 2.062706270627063e-06,
|
| 1640 |
+
"loss": 1.0209,
|
| 1641 |
+
"step": 2540
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 12.62,
|
| 1645 |
+
"learning_rate": 2.0489548954895492e-06,
|
| 1646 |
+
"loss": 0.9964,
|
| 1647 |
+
"step": 2550
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 12.67,
|
| 1651 |
+
"learning_rate": 2.0352035203520354e-06,
|
| 1652 |
+
"loss": 1.0248,
|
| 1653 |
+
"step": 2560
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 12.72,
|
| 1657 |
+
"learning_rate": 2.0214521452145217e-06,
|
| 1658 |
+
"loss": 0.9682,
|
| 1659 |
+
"step": 2570
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 12.77,
|
| 1663 |
+
"learning_rate": 2.007700770077008e-06,
|
| 1664 |
+
"loss": 1.0485,
|
| 1665 |
+
"step": 2580
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 12.82,
|
| 1669 |
+
"learning_rate": 1.993949394939494e-06,
|
| 1670 |
+
"loss": 0.996,
|
| 1671 |
+
"step": 2590
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 12.87,
|
| 1675 |
+
"learning_rate": 1.9801980198019803e-06,
|
| 1676 |
+
"loss": 1.0432,
|
| 1677 |
+
"step": 2600
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 12.92,
|
| 1681 |
+
"learning_rate": 1.9664466446644666e-06,
|
| 1682 |
+
"loss": 1.0554,
|
| 1683 |
+
"step": 2610
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 12.97,
|
| 1687 |
+
"learning_rate": 1.952695269526953e-06,
|
| 1688 |
+
"loss": 1.0133,
|
| 1689 |
+
"step": 2620
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 13.0,
|
| 1693 |
+
"eval_accuracy": 0.6628352490421456,
|
| 1694 |
+
"eval_loss": 0.8939908146858215,
|
| 1695 |
+
"eval_runtime": 38.8578,
|
| 1696 |
+
"eval_samples_per_second": 73.885,
|
| 1697 |
+
"eval_steps_per_second": 2.316,
|
| 1698 |
+
"step": 2626
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 13.02,
|
| 1702 |
+
"learning_rate": 1.938943894389439e-06,
|
| 1703 |
+
"loss": 0.9908,
|
| 1704 |
+
"step": 2630
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 13.07,
|
| 1708 |
+
"learning_rate": 1.9251925192519253e-06,
|
| 1709 |
+
"loss": 0.9901,
|
| 1710 |
+
"step": 2640
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 13.12,
|
| 1714 |
+
"learning_rate": 1.9114411441144115e-06,
|
| 1715 |
+
"loss": 0.9948,
|
| 1716 |
+
"step": 2650
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 13.17,
|
| 1720 |
+
"learning_rate": 1.8976897689768977e-06,
|
| 1721 |
+
"loss": 1.0064,
|
| 1722 |
+
"step": 2660
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 13.22,
|
| 1726 |
+
"learning_rate": 1.883938393839384e-06,
|
| 1727 |
+
"loss": 1.0087,
|
| 1728 |
+
"step": 2670
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 13.27,
|
| 1732 |
+
"learning_rate": 1.8701870187018704e-06,
|
| 1733 |
+
"loss": 0.9959,
|
| 1734 |
+
"step": 2680
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 13.32,
|
| 1738 |
+
"learning_rate": 1.8564356435643566e-06,
|
| 1739 |
+
"loss": 0.9834,
|
| 1740 |
+
"step": 2690
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 13.37,
|
| 1744 |
+
"learning_rate": 1.8426842684268428e-06,
|
| 1745 |
+
"loss": 1.0346,
|
| 1746 |
+
"step": 2700
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 13.42,
|
| 1750 |
+
"learning_rate": 1.828932893289329e-06,
|
| 1751 |
+
"loss": 0.9487,
|
| 1752 |
+
"step": 2710
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 13.47,
|
| 1756 |
+
"learning_rate": 1.8151815181518153e-06,
|
| 1757 |
+
"loss": 0.9952,
|
| 1758 |
+
"step": 2720
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 13.51,
|
| 1762 |
+
"learning_rate": 1.8014301430143017e-06,
|
| 1763 |
+
"loss": 1.0243,
|
| 1764 |
+
"step": 2730
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 13.56,
|
| 1768 |
+
"learning_rate": 1.787678767876788e-06,
|
| 1769 |
+
"loss": 0.9498,
|
| 1770 |
+
"step": 2740
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 13.61,
|
| 1774 |
+
"learning_rate": 1.7739273927392742e-06,
|
| 1775 |
+
"loss": 0.9573,
|
| 1776 |
+
"step": 2750
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 13.66,
|
| 1780 |
+
"learning_rate": 1.7601760176017604e-06,
|
| 1781 |
+
"loss": 0.9838,
|
| 1782 |
+
"step": 2760
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 13.71,
|
| 1786 |
+
"learning_rate": 1.7464246424642466e-06,
|
| 1787 |
+
"loss": 1.0254,
|
| 1788 |
+
"step": 2770
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 13.76,
|
| 1792 |
+
"learning_rate": 1.7326732673267326e-06,
|
| 1793 |
+
"loss": 1.0186,
|
| 1794 |
+
"step": 2780
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 13.81,
|
| 1798 |
+
"learning_rate": 1.7189218921892189e-06,
|
| 1799 |
+
"loss": 1.032,
|
| 1800 |
+
"step": 2790
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 13.86,
|
| 1804 |
+
"learning_rate": 1.7051705170517053e-06,
|
| 1805 |
+
"loss": 1.0525,
|
| 1806 |
+
"step": 2800
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 13.91,
|
| 1810 |
+
"learning_rate": 1.6914191419141915e-06,
|
| 1811 |
+
"loss": 1.0422,
|
| 1812 |
+
"step": 2810
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 13.96,
|
| 1816 |
+
"learning_rate": 1.6776677667766778e-06,
|
| 1817 |
+
"loss": 1.0064,
|
| 1818 |
+
"step": 2820
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 14.0,
|
| 1822 |
+
"eval_accuracy": 0.6649251132009752,
|
| 1823 |
+
"eval_loss": 0.8916007876396179,
|
| 1824 |
+
"eval_runtime": 38.5634,
|
| 1825 |
+
"eval_samples_per_second": 74.449,
|
| 1826 |
+
"eval_steps_per_second": 2.334,
|
| 1827 |
+
"step": 2828
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 14.01,
|
| 1831 |
+
"learning_rate": 1.663916391639164e-06,
|
| 1832 |
+
"loss": 1.0109,
|
| 1833 |
+
"step": 2830
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 14.06,
|
| 1837 |
+
"learning_rate": 1.6501650165016502e-06,
|
| 1838 |
+
"loss": 1.016,
|
| 1839 |
+
"step": 2840
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 14.11,
|
| 1843 |
+
"learning_rate": 1.6364136413641364e-06,
|
| 1844 |
+
"loss": 1.0287,
|
| 1845 |
+
"step": 2850
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 14.16,
|
| 1849 |
+
"learning_rate": 1.6226622662266229e-06,
|
| 1850 |
+
"loss": 1.0048,
|
| 1851 |
+
"step": 2860
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 14.21,
|
| 1855 |
+
"learning_rate": 1.6089108910891091e-06,
|
| 1856 |
+
"loss": 1.0639,
|
| 1857 |
+
"step": 2870
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 14.26,
|
| 1861 |
+
"learning_rate": 1.5951595159515953e-06,
|
| 1862 |
+
"loss": 0.9976,
|
| 1863 |
+
"step": 2880
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 14.31,
|
| 1867 |
+
"learning_rate": 1.5814081408140816e-06,
|
| 1868 |
+
"loss": 0.9993,
|
| 1869 |
+
"step": 2890
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 14.36,
|
| 1873 |
+
"learning_rate": 1.5676567656765678e-06,
|
| 1874 |
+
"loss": 1.045,
|
| 1875 |
+
"step": 2900
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 14.41,
|
| 1879 |
+
"learning_rate": 1.5539053905390542e-06,
|
| 1880 |
+
"loss": 1.0057,
|
| 1881 |
+
"step": 2910
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 14.46,
|
| 1885 |
+
"learning_rate": 1.5401540154015405e-06,
|
| 1886 |
+
"loss": 1.0304,
|
| 1887 |
+
"step": 2920
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 14.5,
|
| 1891 |
+
"learning_rate": 1.5264026402640265e-06,
|
| 1892 |
+
"loss": 1.0412,
|
| 1893 |
+
"step": 2930
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 14.55,
|
| 1897 |
+
"learning_rate": 1.5126512651265127e-06,
|
| 1898 |
+
"loss": 1.0259,
|
| 1899 |
+
"step": 2940
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 14.6,
|
| 1903 |
+
"learning_rate": 1.498899889988999e-06,
|
| 1904 |
+
"loss": 1.0207,
|
| 1905 |
+
"step": 2950
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 14.65,
|
| 1909 |
+
"learning_rate": 1.4851485148514852e-06,
|
| 1910 |
+
"loss": 0.9956,
|
| 1911 |
+
"step": 2960
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 14.7,
|
| 1915 |
+
"learning_rate": 1.4713971397139714e-06,
|
| 1916 |
+
"loss": 0.9838,
|
| 1917 |
+
"step": 2970
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 14.75,
|
| 1921 |
+
"learning_rate": 1.4576457645764578e-06,
|
| 1922 |
+
"loss": 1.0084,
|
| 1923 |
+
"step": 2980
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 14.8,
|
| 1927 |
+
"learning_rate": 1.443894389438944e-06,
|
| 1928 |
+
"loss": 0.9956,
|
| 1929 |
+
"step": 2990
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 14.85,
|
| 1933 |
+
"learning_rate": 1.4301430143014303e-06,
|
| 1934 |
+
"loss": 0.9595,
|
| 1935 |
+
"step": 3000
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 14.9,
|
| 1939 |
+
"learning_rate": 1.4163916391639165e-06,
|
| 1940 |
+
"loss": 0.9391,
|
| 1941 |
+
"step": 3010
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 14.95,
|
| 1945 |
+
"learning_rate": 1.4026402640264027e-06,
|
| 1946 |
+
"loss": 1.0213,
|
| 1947 |
+
"step": 3020
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 15.0,
|
| 1951 |
+
"learning_rate": 1.3888888888888892e-06,
|
| 1952 |
+
"loss": 0.9858,
|
| 1953 |
+
"step": 3030
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 15.0,
|
| 1957 |
+
"eval_accuracy": 0.6732845698362939,
|
| 1958 |
+
"eval_loss": 0.8882250785827637,
|
| 1959 |
+
"eval_runtime": 38.5965,
|
| 1960 |
+
"eval_samples_per_second": 74.385,
|
| 1961 |
+
"eval_steps_per_second": 2.332,
|
| 1962 |
+
"step": 3030
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 15.05,
|
| 1966 |
+
"learning_rate": 1.3751375137513754e-06,
|
| 1967 |
+
"loss": 0.9837,
|
| 1968 |
+
"step": 3040
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 15.1,
|
| 1972 |
+
"learning_rate": 1.3613861386138616e-06,
|
| 1973 |
+
"loss": 0.9723,
|
| 1974 |
+
"step": 3050
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 15.15,
|
| 1978 |
+
"learning_rate": 1.3476347634763478e-06,
|
| 1979 |
+
"loss": 1.018,
|
| 1980 |
+
"step": 3060
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 15.2,
|
| 1984 |
+
"learning_rate": 1.333883388338834e-06,
|
| 1985 |
+
"loss": 1.0612,
|
| 1986 |
+
"step": 3070
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 15.25,
|
| 1990 |
+
"learning_rate": 1.32013201320132e-06,
|
| 1991 |
+
"loss": 1.0081,
|
| 1992 |
+
"step": 3080
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 15.3,
|
| 1996 |
+
"learning_rate": 1.3063806380638063e-06,
|
| 1997 |
+
"loss": 1.0096,
|
| 1998 |
+
"step": 3090
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 15.35,
|
| 2002 |
+
"learning_rate": 1.2926292629262925e-06,
|
| 2003 |
+
"loss": 0.9442,
|
| 2004 |
+
"step": 3100
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 15.4,
|
| 2008 |
+
"learning_rate": 1.278877887788779e-06,
|
| 2009 |
+
"loss": 1.0434,
|
| 2010 |
+
"step": 3110
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 15.45,
|
| 2014 |
+
"learning_rate": 1.2651265126512652e-06,
|
| 2015 |
+
"loss": 1.025,
|
| 2016 |
+
"step": 3120
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 15.5,
|
| 2020 |
+
"learning_rate": 1.2513751375137514e-06,
|
| 2021 |
+
"loss": 1.0102,
|
| 2022 |
+
"step": 3130
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 15.54,
|
| 2026 |
+
"learning_rate": 1.2376237623762377e-06,
|
| 2027 |
+
"loss": 0.9965,
|
| 2028 |
+
"step": 3140
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 15.59,
|
| 2032 |
+
"learning_rate": 1.2238723872387239e-06,
|
| 2033 |
+
"loss": 1.0295,
|
| 2034 |
+
"step": 3150
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 15.64,
|
| 2038 |
+
"learning_rate": 1.2101210121012103e-06,
|
| 2039 |
+
"loss": 0.988,
|
| 2040 |
+
"step": 3160
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 15.69,
|
| 2044 |
+
"learning_rate": 1.1963696369636966e-06,
|
| 2045 |
+
"loss": 1.001,
|
| 2046 |
+
"step": 3170
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 15.74,
|
| 2050 |
+
"learning_rate": 1.1826182618261828e-06,
|
| 2051 |
+
"loss": 1.0232,
|
| 2052 |
+
"step": 3180
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 15.79,
|
| 2056 |
+
"learning_rate": 1.168866886688669e-06,
|
| 2057 |
+
"loss": 1.0148,
|
| 2058 |
+
"step": 3190
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 15.84,
|
| 2062 |
+
"learning_rate": 1.1551155115511552e-06,
|
| 2063 |
+
"loss": 0.9523,
|
| 2064 |
+
"step": 3200
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 15.89,
|
| 2068 |
+
"learning_rate": 1.1413641364136415e-06,
|
| 2069 |
+
"loss": 1.0571,
|
| 2070 |
+
"step": 3210
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 15.94,
|
| 2074 |
+
"learning_rate": 1.1276127612761277e-06,
|
| 2075 |
+
"loss": 0.973,
|
| 2076 |
+
"step": 3220
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 15.99,
|
| 2080 |
+
"learning_rate": 1.113861386138614e-06,
|
| 2081 |
+
"loss": 0.9863,
|
| 2082 |
+
"step": 3230
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 16.0,
|
| 2086 |
+
"eval_accuracy": 0.6739811912225705,
|
| 2087 |
+
"eval_loss": 0.8849846720695496,
|
| 2088 |
+
"eval_runtime": 38.9067,
|
| 2089 |
+
"eval_samples_per_second": 73.792,
|
| 2090 |
+
"eval_steps_per_second": 2.313,
|
| 2091 |
+
"step": 3232
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 16.04,
|
| 2095 |
+
"learning_rate": 1.1001100110011001e-06,
|
| 2096 |
+
"loss": 0.9616,
|
| 2097 |
+
"step": 3240
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 16.09,
|
| 2101 |
+
"learning_rate": 1.0863586358635866e-06,
|
| 2102 |
+
"loss": 1.0094,
|
| 2103 |
+
"step": 3250
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 16.14,
|
| 2107 |
+
"learning_rate": 1.0726072607260726e-06,
|
| 2108 |
+
"loss": 0.9672,
|
| 2109 |
+
"step": 3260
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 16.19,
|
| 2113 |
+
"learning_rate": 1.0588558855885588e-06,
|
| 2114 |
+
"loss": 1.0321,
|
| 2115 |
+
"step": 3270
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 16.24,
|
| 2119 |
+
"learning_rate": 1.0451045104510453e-06,
|
| 2120 |
+
"loss": 1.0313,
|
| 2121 |
+
"step": 3280
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 16.29,
|
| 2125 |
+
"learning_rate": 1.0313531353135315e-06,
|
| 2126 |
+
"loss": 1.0116,
|
| 2127 |
+
"step": 3290
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 16.34,
|
| 2131 |
+
"learning_rate": 1.0176017601760177e-06,
|
| 2132 |
+
"loss": 0.9632,
|
| 2133 |
+
"step": 3300
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 16.39,
|
| 2137 |
+
"learning_rate": 1.003850385038504e-06,
|
| 2138 |
+
"loss": 1.0021,
|
| 2139 |
+
"step": 3310
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 16.44,
|
| 2143 |
+
"learning_rate": 9.900990099009902e-07,
|
| 2144 |
+
"loss": 0.9827,
|
| 2145 |
+
"step": 3320
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 16.49,
|
| 2149 |
+
"learning_rate": 9.763476347634764e-07,
|
| 2150 |
+
"loss": 0.9916,
|
| 2151 |
+
"step": 3330
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 16.53,
|
| 2155 |
+
"learning_rate": 9.625962596259626e-07,
|
| 2156 |
+
"loss": 0.9863,
|
| 2157 |
+
"step": 3340
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 16.58,
|
| 2161 |
+
"learning_rate": 9.488448844884489e-07,
|
| 2162 |
+
"loss": 0.9428,
|
| 2163 |
+
"step": 3350
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 16.63,
|
| 2167 |
+
"learning_rate": 9.350935093509352e-07,
|
| 2168 |
+
"loss": 0.9882,
|
| 2169 |
+
"step": 3360
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 16.68,
|
| 2173 |
+
"learning_rate": 9.213421342134214e-07,
|
| 2174 |
+
"loss": 1.0206,
|
| 2175 |
+
"step": 3370
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 16.73,
|
| 2179 |
+
"learning_rate": 9.075907590759076e-07,
|
| 2180 |
+
"loss": 0.9906,
|
| 2181 |
+
"step": 3380
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 16.78,
|
| 2185 |
+
"learning_rate": 8.93839383938394e-07,
|
| 2186 |
+
"loss": 0.9814,
|
| 2187 |
+
"step": 3390
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 16.83,
|
| 2191 |
+
"learning_rate": 8.800880088008802e-07,
|
| 2192 |
+
"loss": 0.9964,
|
| 2193 |
+
"step": 3400
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 16.88,
|
| 2197 |
+
"learning_rate": 8.663366336633663e-07,
|
| 2198 |
+
"loss": 0.9854,
|
| 2199 |
+
"step": 3410
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 16.93,
|
| 2203 |
+
"learning_rate": 8.525852585258527e-07,
|
| 2204 |
+
"loss": 1.0075,
|
| 2205 |
+
"step": 3420
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 16.98,
|
| 2209 |
+
"learning_rate": 8.388338833883389e-07,
|
| 2210 |
+
"loss": 1.0058,
|
| 2211 |
+
"step": 3430
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 17.0,
|
| 2215 |
+
"eval_accuracy": 0.6746778126088471,
|
| 2216 |
+
"eval_loss": 0.8856182098388672,
|
| 2217 |
+
"eval_runtime": 38.4783,
|
| 2218 |
+
"eval_samples_per_second": 74.614,
|
| 2219 |
+
"eval_steps_per_second": 2.339,
|
| 2220 |
+
"step": 3434
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 17.03,
|
| 2224 |
+
"learning_rate": 8.250825082508251e-07,
|
| 2225 |
+
"loss": 1.0082,
|
| 2226 |
+
"step": 3440
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 17.08,
|
| 2230 |
+
"learning_rate": 8.113311331133114e-07,
|
| 2231 |
+
"loss": 0.9692,
|
| 2232 |
+
"step": 3450
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"epoch": 17.13,
|
| 2236 |
+
"learning_rate": 7.975797579757977e-07,
|
| 2237 |
+
"loss": 0.9329,
|
| 2238 |
+
"step": 3460
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 17.18,
|
| 2242 |
+
"learning_rate": 7.838283828382839e-07,
|
| 2243 |
+
"loss": 0.9938,
|
| 2244 |
+
"step": 3470
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 17.23,
|
| 2248 |
+
"learning_rate": 7.700770077007702e-07,
|
| 2249 |
+
"loss": 0.9911,
|
| 2250 |
+
"step": 3480
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 17.28,
|
| 2254 |
+
"learning_rate": 7.563256325632564e-07,
|
| 2255 |
+
"loss": 1.0087,
|
| 2256 |
+
"step": 3490
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 17.33,
|
| 2260 |
+
"learning_rate": 7.425742574257426e-07,
|
| 2261 |
+
"loss": 0.9697,
|
| 2262 |
+
"step": 3500
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 17.38,
|
| 2266 |
+
"learning_rate": 7.288228822882289e-07,
|
| 2267 |
+
"loss": 0.973,
|
| 2268 |
+
"step": 3510
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 17.43,
|
| 2272 |
+
"learning_rate": 7.150715071507151e-07,
|
| 2273 |
+
"loss": 1.0056,
|
| 2274 |
+
"step": 3520
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 17.48,
|
| 2278 |
+
"learning_rate": 7.013201320132014e-07,
|
| 2279 |
+
"loss": 0.9723,
|
| 2280 |
+
"step": 3530
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 17.52,
|
| 2284 |
+
"learning_rate": 6.875687568756877e-07,
|
| 2285 |
+
"loss": 0.9905,
|
| 2286 |
+
"step": 3540
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 17.57,
|
| 2290 |
+
"learning_rate": 6.738173817381739e-07,
|
| 2291 |
+
"loss": 1.0209,
|
| 2292 |
+
"step": 3550
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 17.62,
|
| 2296 |
+
"learning_rate": 6.6006600660066e-07,
|
| 2297 |
+
"loss": 0.9991,
|
| 2298 |
+
"step": 3560
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 17.67,
|
| 2302 |
+
"learning_rate": 6.463146314631463e-07,
|
| 2303 |
+
"loss": 1.0039,
|
| 2304 |
+
"step": 3570
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 17.72,
|
| 2308 |
+
"learning_rate": 6.325632563256326e-07,
|
| 2309 |
+
"loss": 0.946,
|
| 2310 |
+
"step": 3580
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 17.77,
|
| 2314 |
+
"learning_rate": 6.188118811881188e-07,
|
| 2315 |
+
"loss": 0.9709,
|
| 2316 |
+
"step": 3590
|
| 2317 |
+
},
|
| 2318 |
+
{
|
| 2319 |
+
"epoch": 17.82,
|
| 2320 |
+
"learning_rate": 6.050605060506052e-07,
|
| 2321 |
+
"loss": 1.0213,
|
| 2322 |
+
"step": 3600
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 17.87,
|
| 2326 |
+
"learning_rate": 5.913091309130914e-07,
|
| 2327 |
+
"loss": 1.0016,
|
| 2328 |
+
"step": 3610
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 17.92,
|
| 2332 |
+
"learning_rate": 5.775577557755776e-07,
|
| 2333 |
+
"loss": 0.9889,
|
| 2334 |
+
"step": 3620
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 17.97,
|
| 2338 |
+
"learning_rate": 5.638063806380638e-07,
|
| 2339 |
+
"loss": 0.9637,
|
| 2340 |
+
"step": 3630
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 18.0,
|
| 2344 |
+
"eval_accuracy": 0.6722396377568791,
|
| 2345 |
+
"eval_loss": 0.8851516842842102,
|
| 2346 |
+
"eval_runtime": 38.5349,
|
| 2347 |
+
"eval_samples_per_second": 74.504,
|
| 2348 |
+
"eval_steps_per_second": 2.336,
|
| 2349 |
+
"step": 3636
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 18.02,
|
| 2353 |
+
"learning_rate": 5.500550055005501e-07,
|
| 2354 |
+
"loss": 0.9832,
|
| 2355 |
+
"step": 3640
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 18.07,
|
| 2359 |
+
"learning_rate": 5.363036303630363e-07,
|
| 2360 |
+
"loss": 0.9492,
|
| 2361 |
+
"step": 3650
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 18.12,
|
| 2365 |
+
"learning_rate": 5.225522552255226e-07,
|
| 2366 |
+
"loss": 0.9676,
|
| 2367 |
+
"step": 3660
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 18.17,
|
| 2371 |
+
"learning_rate": 5.088008800880089e-07,
|
| 2372 |
+
"loss": 0.991,
|
| 2373 |
+
"step": 3670
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 18.22,
|
| 2377 |
+
"learning_rate": 4.950495049504951e-07,
|
| 2378 |
+
"loss": 1.0016,
|
| 2379 |
+
"step": 3680
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 18.27,
|
| 2383 |
+
"learning_rate": 4.812981298129813e-07,
|
| 2384 |
+
"loss": 1.0346,
|
| 2385 |
+
"step": 3690
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 18.32,
|
| 2389 |
+
"learning_rate": 4.675467546754676e-07,
|
| 2390 |
+
"loss": 1.0101,
|
| 2391 |
+
"step": 3700
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 18.37,
|
| 2395 |
+
"learning_rate": 4.537953795379538e-07,
|
| 2396 |
+
"loss": 1.0226,
|
| 2397 |
+
"step": 3710
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 18.42,
|
| 2401 |
+
"learning_rate": 4.400440044004401e-07,
|
| 2402 |
+
"loss": 1.0054,
|
| 2403 |
+
"step": 3720
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 18.47,
|
| 2407 |
+
"learning_rate": 4.2629262926292633e-07,
|
| 2408 |
+
"loss": 0.994,
|
| 2409 |
+
"step": 3730
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 18.51,
|
| 2413 |
+
"learning_rate": 4.1254125412541255e-07,
|
| 2414 |
+
"loss": 1.0177,
|
| 2415 |
+
"step": 3740
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 18.56,
|
| 2419 |
+
"learning_rate": 3.9878987898789883e-07,
|
| 2420 |
+
"loss": 0.97,
|
| 2421 |
+
"step": 3750
|
| 2422 |
+
},
|
| 2423 |
+
{
|
| 2424 |
+
"epoch": 18.61,
|
| 2425 |
+
"learning_rate": 3.850385038503851e-07,
|
| 2426 |
+
"loss": 0.9796,
|
| 2427 |
+
"step": 3760
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 18.66,
|
| 2431 |
+
"learning_rate": 3.712871287128713e-07,
|
| 2432 |
+
"loss": 1.0034,
|
| 2433 |
+
"step": 3770
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 18.71,
|
| 2437 |
+
"learning_rate": 3.5753575357535757e-07,
|
| 2438 |
+
"loss": 0.9773,
|
| 2439 |
+
"step": 3780
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 18.76,
|
| 2443 |
+
"learning_rate": 3.4378437843784385e-07,
|
| 2444 |
+
"loss": 0.984,
|
| 2445 |
+
"step": 3790
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 18.81,
|
| 2449 |
+
"learning_rate": 3.3003300330033e-07,
|
| 2450 |
+
"loss": 0.9513,
|
| 2451 |
+
"step": 3800
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 18.86,
|
| 2455 |
+
"learning_rate": 3.162816281628163e-07,
|
| 2456 |
+
"loss": 0.9628,
|
| 2457 |
+
"step": 3810
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 18.91,
|
| 2461 |
+
"learning_rate": 3.025302530253026e-07,
|
| 2462 |
+
"loss": 0.9573,
|
| 2463 |
+
"step": 3820
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 18.96,
|
| 2467 |
+
"learning_rate": 2.887788778877888e-07,
|
| 2468 |
+
"loss": 0.9803,
|
| 2469 |
+
"step": 3830
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 19.0,
|
| 2473 |
+
"eval_accuracy": 0.6753744339951236,
|
| 2474 |
+
"eval_loss": 0.8829284906387329,
|
| 2475 |
+
"eval_runtime": 38.4857,
|
| 2476 |
+
"eval_samples_per_second": 74.599,
|
| 2477 |
+
"eval_steps_per_second": 2.339,
|
| 2478 |
+
"step": 3838
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 19.01,
|
| 2482 |
+
"learning_rate": 2.7502750275027504e-07,
|
| 2483 |
+
"loss": 0.9945,
|
| 2484 |
+
"step": 3840
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 19.06,
|
| 2488 |
+
"learning_rate": 2.612761276127613e-07,
|
| 2489 |
+
"loss": 1.0255,
|
| 2490 |
+
"step": 3850
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 19.11,
|
| 2494 |
+
"learning_rate": 2.4752475247524754e-07,
|
| 2495 |
+
"loss": 0.966,
|
| 2496 |
+
"step": 3860
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 19.16,
|
| 2500 |
+
"learning_rate": 2.337733773377338e-07,
|
| 2501 |
+
"loss": 0.9895,
|
| 2502 |
+
"step": 3870
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 19.21,
|
| 2506 |
+
"learning_rate": 2.2002200220022005e-07,
|
| 2507 |
+
"loss": 0.9694,
|
| 2508 |
+
"step": 3880
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 19.26,
|
| 2512 |
+
"learning_rate": 2.0627062706270628e-07,
|
| 2513 |
+
"loss": 0.9622,
|
| 2514 |
+
"step": 3890
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 19.31,
|
| 2518 |
+
"learning_rate": 1.9251925192519256e-07,
|
| 2519 |
+
"loss": 0.9951,
|
| 2520 |
+
"step": 3900
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 19.36,
|
| 2524 |
+
"learning_rate": 1.7876787678767878e-07,
|
| 2525 |
+
"loss": 0.9759,
|
| 2526 |
+
"step": 3910
|
| 2527 |
+
},
|
| 2528 |
+
{
|
| 2529 |
+
"epoch": 19.41,
|
| 2530 |
+
"learning_rate": 1.65016501650165e-07,
|
| 2531 |
+
"loss": 1.019,
|
| 2532 |
+
"step": 3920
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 19.46,
|
| 2536 |
+
"learning_rate": 1.512651265126513e-07,
|
| 2537 |
+
"loss": 1.0533,
|
| 2538 |
+
"step": 3930
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 19.5,
|
| 2542 |
+
"learning_rate": 1.3751375137513752e-07,
|
| 2543 |
+
"loss": 0.9828,
|
| 2544 |
+
"step": 3940
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 19.55,
|
| 2548 |
+
"learning_rate": 1.2376237623762377e-07,
|
| 2549 |
+
"loss": 1.0073,
|
| 2550 |
+
"step": 3950
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 19.6,
|
| 2554 |
+
"learning_rate": 1.1001100110011003e-07,
|
| 2555 |
+
"loss": 0.9639,
|
| 2556 |
+
"step": 3960
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 19.65,
|
| 2560 |
+
"learning_rate": 9.625962596259628e-08,
|
| 2561 |
+
"loss": 0.9632,
|
| 2562 |
+
"step": 3970
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 19.7,
|
| 2566 |
+
"learning_rate": 8.25082508250825e-08,
|
| 2567 |
+
"loss": 0.9732,
|
| 2568 |
+
"step": 3980
|
| 2569 |
+
},
|
| 2570 |
+
{
|
| 2571 |
+
"epoch": 19.75,
|
| 2572 |
+
"learning_rate": 6.875687568756876e-08,
|
| 2573 |
+
"loss": 0.9832,
|
| 2574 |
+
"step": 3990
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 19.8,
|
| 2578 |
+
"learning_rate": 5.500550055005501e-08,
|
| 2579 |
+
"loss": 1.0138,
|
| 2580 |
+
"step": 4000
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 19.85,
|
| 2584 |
+
"learning_rate": 4.125412541254125e-08,
|
| 2585 |
+
"loss": 0.9895,
|
| 2586 |
+
"step": 4010
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 19.9,
|
| 2590 |
+
"learning_rate": 2.7502750275027506e-08,
|
| 2591 |
+
"loss": 0.9244,
|
| 2592 |
+
"step": 4020
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 19.95,
|
| 2596 |
+
"learning_rate": 1.3751375137513753e-08,
|
| 2597 |
+
"loss": 0.9557,
|
| 2598 |
+
"step": 4030
|
| 2599 |
+
},
|
| 2600 |
+
{
|
| 2601 |
+
"epoch": 20.0,
|
| 2602 |
+
"learning_rate": 0.0,
|
| 2603 |
+
"loss": 0.9356,
|
| 2604 |
+
"step": 4040
|
| 2605 |
+
},
|
| 2606 |
+
{
|
| 2607 |
+
"epoch": 20.0,
|
| 2608 |
+
"eval_accuracy": 0.6788575409265064,
|
| 2609 |
+
"eval_loss": 0.8812496662139893,
|
| 2610 |
+
"eval_runtime": 38.3816,
|
| 2611 |
+
"eval_samples_per_second": 74.801,
|
| 2612 |
+
"eval_steps_per_second": 2.345,
|
| 2613 |
+
"step": 4040
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 20.0,
|
| 2617 |
+
"step": 4040,
|
| 2618 |
+
"total_flos": 4.0029439499828675e+19,
|
| 2619 |
+
"train_loss": 1.115002474690428,
|
| 2620 |
+
"train_runtime": 19507.0594,
|
| 2621 |
+
"train_samples_per_second": 26.491,
|
| 2622 |
+
"train_steps_per_second": 0.207
|
| 2623 |
+
}
|
| 2624 |
+
],
|
| 2625 |
+
"logging_steps": 10,
|
| 2626 |
+
"max_steps": 4040,
|
| 2627 |
+
"num_train_epochs": 20,
|
| 2628 |
+
"save_steps": 500,
|
| 2629 |
+
"total_flos": 4.0029439499828675e+19,
|
| 2630 |
+
"trial_name": null,
|
| 2631 |
+
"trial_params": null
|
| 2632 |
+
}
|