End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- logs/events.out.tfevents.1686627545.serv-3317.1603486.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +1333 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.83,
|
| 3 |
+
"eval_accuracy": 0.5855461819146771,
|
| 4 |
+
"eval_loss": 2.197262763977051,
|
| 5 |
+
"eval_runtime": 1891.0124,
|
| 6 |
+
"eval_samples": 308356,
|
| 7 |
+
"eval_samples_per_second": 163.064,
|
| 8 |
+
"eval_steps_per_second": 3.398,
|
| 9 |
+
"perplexity": 9.000343686329554,
|
| 10 |
+
"train_loss": 2.3200887040276985,
|
| 11 |
+
"train_runtime": 107983.587,
|
| 12 |
+
"train_samples": 5858758,
|
| 13 |
+
"train_samples_per_second": 271.28,
|
| 14 |
+
"train_steps_per_second": 5.652
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.83,
|
| 3 |
+
"eval_accuracy": 0.5855461819146771,
|
| 4 |
+
"eval_loss": 2.197262763977051,
|
| 5 |
+
"eval_runtime": 1891.0124,
|
| 6 |
+
"eval_samples": 308356,
|
| 7 |
+
"eval_samples_per_second": 163.064,
|
| 8 |
+
"eval_steps_per_second": 3.398,
|
| 9 |
+
"perplexity": 9.000343686329554
|
| 10 |
+
}
|
logs/events.out.tfevents.1686627545.serv-3317.1603486.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea9713354ebd5d40ae3ea615c30c3d023329b5c56331d45d040c92a57cdbbd2b
|
| 3 |
+
size 369
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.83,
|
| 3 |
+
"train_loss": 2.3200887040276985,
|
| 4 |
+
"train_runtime": 107983.587,
|
| 5 |
+
"train_samples": 5858758,
|
| 6 |
+
"train_samples_per_second": 271.28,
|
| 7 |
+
"train_steps_per_second": 5.652
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1333 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 2.196564197540283,
|
| 3 |
+
"best_model_checkpoint": "bert_12_layer_model_v1_complete_training_new_wt_init_120/checkpoint-100000",
|
| 4 |
+
"epoch": 0.83470153533566,
|
| 5 |
+
"global_step": 101882,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 5.000000000000001e-07,
|
| 13 |
+
"loss": 2.3675,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 19 |
+
"loss": 2.3595,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 1.5e-06,
|
| 25 |
+
"loss": 2.3576,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.02,
|
| 30 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 31 |
+
"loss": 2.3519,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 2.5e-06,
|
| 37 |
+
"loss": 2.3507,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
+
"learning_rate": 3e-06,
|
| 43 |
+
"loss": 2.3571,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03,
|
| 48 |
+
"learning_rate": 3.5e-06,
|
| 49 |
+
"loss": 2.351,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.03,
|
| 54 |
+
"learning_rate": 4.000000000000001e-06,
|
| 55 |
+
"loss": 2.357,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.04,
|
| 60 |
+
"learning_rate": 4.5e-06,
|
| 61 |
+
"loss": 2.3538,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.04,
|
| 66 |
+
"learning_rate": 5e-06,
|
| 67 |
+
"loss": 2.3588,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.05,
|
| 72 |
+
"learning_rate": 5.500000000000001e-06,
|
| 73 |
+
"loss": 2.3534,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.05,
|
| 78 |
+
"learning_rate": 6e-06,
|
| 79 |
+
"loss": 2.3591,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.05,
|
| 84 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 85 |
+
"loss": 2.3618,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.06,
|
| 90 |
+
"learning_rate": 7e-06,
|
| 91 |
+
"loss": 2.3547,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.06,
|
| 96 |
+
"learning_rate": 7.500000000000001e-06,
|
| 97 |
+
"loss": 2.3576,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.07,
|
| 102 |
+
"learning_rate": 8.000000000000001e-06,
|
| 103 |
+
"loss": 2.363,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.07,
|
| 108 |
+
"learning_rate": 8.5e-06,
|
| 109 |
+
"loss": 2.3594,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.07,
|
| 114 |
+
"learning_rate": 9e-06,
|
| 115 |
+
"loss": 2.3598,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.08,
|
| 120 |
+
"learning_rate": 9.5e-06,
|
| 121 |
+
"loss": 2.3633,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.08,
|
| 126 |
+
"learning_rate": 1e-05,
|
| 127 |
+
"loss": 2.3673,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.08,
|
| 132 |
+
"eval_accuracy": 0.5731936070718191,
|
| 133 |
+
"eval_loss": 2.2852351665496826,
|
| 134 |
+
"eval_runtime": 1891.0124,
|
| 135 |
+
"eval_samples_per_second": 163.064,
|
| 136 |
+
"eval_steps_per_second": 3.398,
|
| 137 |
+
"step": 10000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.09,
|
| 141 |
+
"learning_rate": 9.991670692498626e-06,
|
| 142 |
+
"loss": 2.3655,
|
| 143 |
+
"step": 10500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.09,
|
| 147 |
+
"learning_rate": 9.983341384997253e-06,
|
| 148 |
+
"loss": 2.3768,
|
| 149 |
+
"step": 11000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.09,
|
| 153 |
+
"learning_rate": 9.975012077495878e-06,
|
| 154 |
+
"loss": 2.3679,
|
| 155 |
+
"step": 11500
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1,
|
| 159 |
+
"learning_rate": 9.966682769994504e-06,
|
| 160 |
+
"loss": 2.367,
|
| 161 |
+
"step": 12000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.1,
|
| 165 |
+
"learning_rate": 9.95835346249313e-06,
|
| 166 |
+
"loss": 2.3626,
|
| 167 |
+
"step": 12500
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.11,
|
| 171 |
+
"learning_rate": 9.950024154991756e-06,
|
| 172 |
+
"loss": 2.3604,
|
| 173 |
+
"step": 13000
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.11,
|
| 177 |
+
"learning_rate": 9.941694847490381e-06,
|
| 178 |
+
"loss": 2.3657,
|
| 179 |
+
"step": 13500
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.11,
|
| 183 |
+
"learning_rate": 9.933365539989006e-06,
|
| 184 |
+
"loss": 2.366,
|
| 185 |
+
"step": 14000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.12,
|
| 189 |
+
"learning_rate": 9.925036232487631e-06,
|
| 190 |
+
"loss": 2.3569,
|
| 191 |
+
"step": 14500
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.12,
|
| 195 |
+
"learning_rate": 9.916706924986258e-06,
|
| 196 |
+
"loss": 2.3627,
|
| 197 |
+
"step": 15000
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.13,
|
| 201 |
+
"learning_rate": 9.908377617484883e-06,
|
| 202 |
+
"loss": 2.3573,
|
| 203 |
+
"step": 15500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.13,
|
| 207 |
+
"learning_rate": 9.90004830998351e-06,
|
| 208 |
+
"loss": 2.3618,
|
| 209 |
+
"step": 16000
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.14,
|
| 213 |
+
"learning_rate": 9.891719002482134e-06,
|
| 214 |
+
"loss": 2.3621,
|
| 215 |
+
"step": 16500
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.14,
|
| 219 |
+
"learning_rate": 9.883389694980761e-06,
|
| 220 |
+
"loss": 2.3579,
|
| 221 |
+
"step": 17000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.14,
|
| 225 |
+
"learning_rate": 9.875060387479386e-06,
|
| 226 |
+
"loss": 2.3607,
|
| 227 |
+
"step": 17500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.15,
|
| 231 |
+
"learning_rate": 9.866731079978011e-06,
|
| 232 |
+
"loss": 2.3618,
|
| 233 |
+
"step": 18000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.15,
|
| 237 |
+
"learning_rate": 9.858401772476636e-06,
|
| 238 |
+
"loss": 2.3637,
|
| 239 |
+
"step": 18500
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.16,
|
| 243 |
+
"learning_rate": 9.850072464975263e-06,
|
| 244 |
+
"loss": 2.3576,
|
| 245 |
+
"step": 19000
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.16,
|
| 249 |
+
"learning_rate": 9.84174315747389e-06,
|
| 250 |
+
"loss": 2.3575,
|
| 251 |
+
"step": 19500
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.16,
|
| 255 |
+
"learning_rate": 9.833413849972514e-06,
|
| 256 |
+
"loss": 2.356,
|
| 257 |
+
"step": 20000
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.16,
|
| 261 |
+
"eval_accuracy": 0.5743537529447622,
|
| 262 |
+
"eval_loss": 2.2771806716918945,
|
| 263 |
+
"eval_runtime": 1892.4941,
|
| 264 |
+
"eval_samples_per_second": 162.936,
|
| 265 |
+
"eval_steps_per_second": 3.395,
|
| 266 |
+
"step": 20000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.17,
|
| 270 |
+
"learning_rate": 9.82508454247114e-06,
|
| 271 |
+
"loss": 2.3515,
|
| 272 |
+
"step": 20500
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.17,
|
| 276 |
+
"learning_rate": 9.816755234969766e-06,
|
| 277 |
+
"loss": 2.358,
|
| 278 |
+
"step": 21000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.18,
|
| 282 |
+
"learning_rate": 9.808425927468391e-06,
|
| 283 |
+
"loss": 2.3538,
|
| 284 |
+
"step": 21500
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.18,
|
| 288 |
+
"learning_rate": 9.800096619967016e-06,
|
| 289 |
+
"loss": 2.3531,
|
| 290 |
+
"step": 22000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.18,
|
| 294 |
+
"learning_rate": 9.791767312465643e-06,
|
| 295 |
+
"loss": 2.3464,
|
| 296 |
+
"step": 22500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.19,
|
| 300 |
+
"learning_rate": 9.783438004964268e-06,
|
| 301 |
+
"loss": 2.3455,
|
| 302 |
+
"step": 23000
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.19,
|
| 306 |
+
"learning_rate": 9.775108697462895e-06,
|
| 307 |
+
"loss": 2.3533,
|
| 308 |
+
"step": 23500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.2,
|
| 312 |
+
"learning_rate": 9.76677938996152e-06,
|
| 313 |
+
"loss": 2.3474,
|
| 314 |
+
"step": 24000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2,
|
| 318 |
+
"learning_rate": 9.758450082460145e-06,
|
| 319 |
+
"loss": 2.3466,
|
| 320 |
+
"step": 24500
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.2,
|
| 324 |
+
"learning_rate": 9.750120774958771e-06,
|
| 325 |
+
"loss": 2.3418,
|
| 326 |
+
"step": 25000
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.21,
|
| 330 |
+
"learning_rate": 9.741791467457396e-06,
|
| 331 |
+
"loss": 2.3468,
|
| 332 |
+
"step": 25500
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.21,
|
| 336 |
+
"learning_rate": 9.733462159956021e-06,
|
| 337 |
+
"loss": 2.3472,
|
| 338 |
+
"step": 26000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.22,
|
| 342 |
+
"learning_rate": 9.725132852454648e-06,
|
| 343 |
+
"loss": 2.3502,
|
| 344 |
+
"step": 26500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.22,
|
| 348 |
+
"learning_rate": 9.716803544953275e-06,
|
| 349 |
+
"loss": 2.3481,
|
| 350 |
+
"step": 27000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.23,
|
| 354 |
+
"learning_rate": 9.7084742374519e-06,
|
| 355 |
+
"loss": 2.3465,
|
| 356 |
+
"step": 27500
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.23,
|
| 360 |
+
"learning_rate": 9.700144929950525e-06,
|
| 361 |
+
"loss": 2.3432,
|
| 362 |
+
"step": 28000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.23,
|
| 366 |
+
"learning_rate": 9.69181562244915e-06,
|
| 367 |
+
"loss": 2.3475,
|
| 368 |
+
"step": 28500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.24,
|
| 372 |
+
"learning_rate": 9.683486314947776e-06,
|
| 373 |
+
"loss": 2.3455,
|
| 374 |
+
"step": 29000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.24,
|
| 378 |
+
"learning_rate": 9.675157007446401e-06,
|
| 379 |
+
"loss": 2.3468,
|
| 380 |
+
"step": 29500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.25,
|
| 384 |
+
"learning_rate": 9.666827699945028e-06,
|
| 385 |
+
"loss": 2.3424,
|
| 386 |
+
"step": 30000
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.25,
|
| 390 |
+
"eval_accuracy": 0.5764564327016705,
|
| 391 |
+
"eval_loss": 2.2639663219451904,
|
| 392 |
+
"eval_runtime": 1891.3875,
|
| 393 |
+
"eval_samples_per_second": 163.032,
|
| 394 |
+
"eval_steps_per_second": 3.397,
|
| 395 |
+
"step": 30000
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.25,
|
| 399 |
+
"learning_rate": 9.658498392443653e-06,
|
| 400 |
+
"loss": 2.3371,
|
| 401 |
+
"step": 30500
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.25,
|
| 405 |
+
"learning_rate": 9.65016908494228e-06,
|
| 406 |
+
"loss": 2.3429,
|
| 407 |
+
"step": 31000
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.26,
|
| 411 |
+
"learning_rate": 9.641839777440905e-06,
|
| 412 |
+
"loss": 2.343,
|
| 413 |
+
"step": 31500
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.26,
|
| 417 |
+
"learning_rate": 9.63351046993953e-06,
|
| 418 |
+
"loss": 2.344,
|
| 419 |
+
"step": 32000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.27,
|
| 423 |
+
"learning_rate": 9.625181162438155e-06,
|
| 424 |
+
"loss": 2.3401,
|
| 425 |
+
"step": 32500
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.27,
|
| 429 |
+
"learning_rate": 9.616851854936781e-06,
|
| 430 |
+
"loss": 2.3455,
|
| 431 |
+
"step": 33000
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.27,
|
| 435 |
+
"learning_rate": 9.608522547435406e-06,
|
| 436 |
+
"loss": 2.3416,
|
| 437 |
+
"step": 33500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.28,
|
| 441 |
+
"learning_rate": 9.600193239934033e-06,
|
| 442 |
+
"loss": 2.3373,
|
| 443 |
+
"step": 34000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.28,
|
| 447 |
+
"learning_rate": 9.591863932432658e-06,
|
| 448 |
+
"loss": 2.34,
|
| 449 |
+
"step": 34500
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.29,
|
| 453 |
+
"learning_rate": 9.583534624931285e-06,
|
| 454 |
+
"loss": 2.338,
|
| 455 |
+
"step": 35000
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.29,
|
| 459 |
+
"learning_rate": 9.57520531742991e-06,
|
| 460 |
+
"loss": 2.3441,
|
| 461 |
+
"step": 35500
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.29,
|
| 465 |
+
"learning_rate": 9.566876009928535e-06,
|
| 466 |
+
"loss": 2.338,
|
| 467 |
+
"step": 36000
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.3,
|
| 471 |
+
"learning_rate": 9.55854670242716e-06,
|
| 472 |
+
"loss": 2.3397,
|
| 473 |
+
"step": 36500
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.3,
|
| 477 |
+
"learning_rate": 9.550217394925787e-06,
|
| 478 |
+
"loss": 2.3369,
|
| 479 |
+
"step": 37000
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.31,
|
| 483 |
+
"learning_rate": 9.541888087424413e-06,
|
| 484 |
+
"loss": 2.3341,
|
| 485 |
+
"step": 37500
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.31,
|
| 489 |
+
"learning_rate": 9.533558779923038e-06,
|
| 490 |
+
"loss": 2.3323,
|
| 491 |
+
"step": 38000
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.32,
|
| 495 |
+
"learning_rate": 9.525229472421663e-06,
|
| 496 |
+
"loss": 2.3261,
|
| 497 |
+
"step": 38500
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.32,
|
| 501 |
+
"learning_rate": 9.51690016492029e-06,
|
| 502 |
+
"loss": 2.3324,
|
| 503 |
+
"step": 39000
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 0.32,
|
| 507 |
+
"learning_rate": 9.508570857418915e-06,
|
| 508 |
+
"loss": 2.3313,
|
| 509 |
+
"step": 39500
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.33,
|
| 513 |
+
"learning_rate": 9.50024154991754e-06,
|
| 514 |
+
"loss": 2.3442,
|
| 515 |
+
"step": 40000
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.33,
|
| 519 |
+
"eval_accuracy": 0.5778452791429078,
|
| 520 |
+
"eval_loss": 2.252516984939575,
|
| 521 |
+
"eval_runtime": 1892.6186,
|
| 522 |
+
"eval_samples_per_second": 162.926,
|
| 523 |
+
"eval_steps_per_second": 3.395,
|
| 524 |
+
"step": 40000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.33,
|
| 528 |
+
"learning_rate": 9.491912242416167e-06,
|
| 529 |
+
"loss": 2.3337,
|
| 530 |
+
"step": 40500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.34,
|
| 534 |
+
"learning_rate": 9.483582934914792e-06,
|
| 535 |
+
"loss": 2.3321,
|
| 536 |
+
"step": 41000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.34,
|
| 540 |
+
"learning_rate": 9.475253627413418e-06,
|
| 541 |
+
"loss": 2.33,
|
| 542 |
+
"step": 41500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.34,
|
| 546 |
+
"learning_rate": 9.466924319912043e-06,
|
| 547 |
+
"loss": 2.3366,
|
| 548 |
+
"step": 42000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.35,
|
| 552 |
+
"learning_rate": 9.458595012410668e-06,
|
| 553 |
+
"loss": 2.3265,
|
| 554 |
+
"step": 42500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.35,
|
| 558 |
+
"learning_rate": 9.450265704909295e-06,
|
| 559 |
+
"loss": 2.3293,
|
| 560 |
+
"step": 43000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.36,
|
| 564 |
+
"learning_rate": 9.44193639740792e-06,
|
| 565 |
+
"loss": 2.3218,
|
| 566 |
+
"step": 43500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.36,
|
| 570 |
+
"learning_rate": 9.433607089906545e-06,
|
| 571 |
+
"loss": 2.3291,
|
| 572 |
+
"step": 44000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.36,
|
| 576 |
+
"learning_rate": 9.425277782405172e-06,
|
| 577 |
+
"loss": 2.3256,
|
| 578 |
+
"step": 44500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.37,
|
| 582 |
+
"learning_rate": 9.416948474903798e-06,
|
| 583 |
+
"loss": 2.3276,
|
| 584 |
+
"step": 45000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.37,
|
| 588 |
+
"learning_rate": 9.408619167402423e-06,
|
| 589 |
+
"loss": 2.332,
|
| 590 |
+
"step": 45500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.38,
|
| 594 |
+
"learning_rate": 9.400289859901048e-06,
|
| 595 |
+
"loss": 2.3193,
|
| 596 |
+
"step": 46000
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.38,
|
| 600 |
+
"learning_rate": 9.391960552399673e-06,
|
| 601 |
+
"loss": 2.3206,
|
| 602 |
+
"step": 46500
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.39,
|
| 606 |
+
"learning_rate": 9.3836312448983e-06,
|
| 607 |
+
"loss": 2.3219,
|
| 608 |
+
"step": 47000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.39,
|
| 612 |
+
"learning_rate": 9.375301937396925e-06,
|
| 613 |
+
"loss": 2.3209,
|
| 614 |
+
"step": 47500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.39,
|
| 618 |
+
"learning_rate": 9.366972629895552e-06,
|
| 619 |
+
"loss": 2.3219,
|
| 620 |
+
"step": 48000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.4,
|
| 624 |
+
"learning_rate": 9.358643322394177e-06,
|
| 625 |
+
"loss": 2.3246,
|
| 626 |
+
"step": 48500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.4,
|
| 630 |
+
"learning_rate": 9.350314014892804e-06,
|
| 631 |
+
"loss": 2.32,
|
| 632 |
+
"step": 49000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.41,
|
| 636 |
+
"learning_rate": 9.341984707391429e-06,
|
| 637 |
+
"loss": 2.3204,
|
| 638 |
+
"step": 49500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.41,
|
| 642 |
+
"learning_rate": 9.333655399890054e-06,
|
| 643 |
+
"loss": 2.3228,
|
| 644 |
+
"step": 50000
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.41,
|
| 648 |
+
"eval_accuracy": 0.5792503659913376,
|
| 649 |
+
"eval_loss": 2.242736577987671,
|
| 650 |
+
"eval_runtime": 1890.8069,
|
| 651 |
+
"eval_samples_per_second": 163.082,
|
| 652 |
+
"eval_steps_per_second": 3.398,
|
| 653 |
+
"step": 50000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.41,
|
| 657 |
+
"learning_rate": 9.325326092388679e-06,
|
| 658 |
+
"loss": 2.3192,
|
| 659 |
+
"step": 50500
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.42,
|
| 663 |
+
"learning_rate": 9.316996784887305e-06,
|
| 664 |
+
"loss": 2.3221,
|
| 665 |
+
"step": 51000
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.42,
|
| 669 |
+
"learning_rate": 9.30866747738593e-06,
|
| 670 |
+
"loss": 2.3223,
|
| 671 |
+
"step": 51500
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 0.43,
|
| 675 |
+
"learning_rate": 9.300338169884557e-06,
|
| 676 |
+
"loss": 2.3238,
|
| 677 |
+
"step": 52000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.43,
|
| 681 |
+
"learning_rate": 9.292008862383182e-06,
|
| 682 |
+
"loss": 2.3186,
|
| 683 |
+
"step": 52500
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.43,
|
| 687 |
+
"learning_rate": 9.283679554881809e-06,
|
| 688 |
+
"loss": 2.3167,
|
| 689 |
+
"step": 53000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.44,
|
| 693 |
+
"learning_rate": 9.275350247380434e-06,
|
| 694 |
+
"loss": 2.3186,
|
| 695 |
+
"step": 53500
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.44,
|
| 699 |
+
"learning_rate": 9.267020939879059e-06,
|
| 700 |
+
"loss": 2.3126,
|
| 701 |
+
"step": 54000
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.45,
|
| 705 |
+
"learning_rate": 9.258691632377685e-06,
|
| 706 |
+
"loss": 2.3162,
|
| 707 |
+
"step": 54500
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 0.45,
|
| 711 |
+
"learning_rate": 9.25036232487631e-06,
|
| 712 |
+
"loss": 2.3158,
|
| 713 |
+
"step": 55000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.45,
|
| 717 |
+
"learning_rate": 9.242033017374937e-06,
|
| 718 |
+
"loss": 2.3174,
|
| 719 |
+
"step": 55500
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.46,
|
| 723 |
+
"learning_rate": 9.233703709873562e-06,
|
| 724 |
+
"loss": 2.3112,
|
| 725 |
+
"step": 56000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.46,
|
| 729 |
+
"learning_rate": 9.225374402372189e-06,
|
| 730 |
+
"loss": 2.3142,
|
| 731 |
+
"step": 56500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.47,
|
| 735 |
+
"learning_rate": 9.217045094870814e-06,
|
| 736 |
+
"loss": 2.3091,
|
| 737 |
+
"step": 57000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.47,
|
| 741 |
+
"learning_rate": 9.208715787369439e-06,
|
| 742 |
+
"loss": 2.3156,
|
| 743 |
+
"step": 57500
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 0.48,
|
| 747 |
+
"learning_rate": 9.200386479868064e-06,
|
| 748 |
+
"loss": 2.3101,
|
| 749 |
+
"step": 58000
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.48,
|
| 753 |
+
"learning_rate": 9.19205717236669e-06,
|
| 754 |
+
"loss": 2.3095,
|
| 755 |
+
"step": 58500
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 0.48,
|
| 759 |
+
"learning_rate": 9.183727864865315e-06,
|
| 760 |
+
"loss": 2.3163,
|
| 761 |
+
"step": 59000
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.49,
|
| 765 |
+
"learning_rate": 9.175398557363942e-06,
|
| 766 |
+
"loss": 2.3174,
|
| 767 |
+
"step": 59500
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.49,
|
| 771 |
+
"learning_rate": 9.167069249862567e-06,
|
| 772 |
+
"loss": 2.3179,
|
| 773 |
+
"step": 60000
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.49,
|
| 777 |
+
"eval_accuracy": 0.5809536977963716,
|
| 778 |
+
"eval_loss": 2.231283187866211,
|
| 779 |
+
"eval_runtime": 1893.8614,
|
| 780 |
+
"eval_samples_per_second": 162.819,
|
| 781 |
+
"eval_steps_per_second": 3.393,
|
| 782 |
+
"step": 60000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.5,
|
| 786 |
+
"learning_rate": 9.158739942361194e-06,
|
| 787 |
+
"loss": 2.3106,
|
| 788 |
+
"step": 60500
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.5,
|
| 792 |
+
"learning_rate": 9.150410634859819e-06,
|
| 793 |
+
"loss": 2.3052,
|
| 794 |
+
"step": 61000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.5,
|
| 798 |
+
"learning_rate": 9.142081327358444e-06,
|
| 799 |
+
"loss": 2.3124,
|
| 800 |
+
"step": 61500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.51,
|
| 804 |
+
"learning_rate": 9.133752019857069e-06,
|
| 805 |
+
"loss": 2.3052,
|
| 806 |
+
"step": 62000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.51,
|
| 810 |
+
"learning_rate": 9.125422712355696e-06,
|
| 811 |
+
"loss": 2.3096,
|
| 812 |
+
"step": 62500
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.52,
|
| 816 |
+
"learning_rate": 9.11709340485432e-06,
|
| 817 |
+
"loss": 2.3116,
|
| 818 |
+
"step": 63000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.52,
|
| 822 |
+
"learning_rate": 9.108764097352947e-06,
|
| 823 |
+
"loss": 2.3022,
|
| 824 |
+
"step": 63500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.52,
|
| 828 |
+
"learning_rate": 9.100434789851572e-06,
|
| 829 |
+
"loss": 2.3036,
|
| 830 |
+
"step": 64000
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.53,
|
| 834 |
+
"learning_rate": 9.092105482350199e-06,
|
| 835 |
+
"loss": 2.3055,
|
| 836 |
+
"step": 64500
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.53,
|
| 840 |
+
"learning_rate": 9.083776174848824e-06,
|
| 841 |
+
"loss": 2.3029,
|
| 842 |
+
"step": 65000
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.54,
|
| 846 |
+
"learning_rate": 9.075446867347449e-06,
|
| 847 |
+
"loss": 2.313,
|
| 848 |
+
"step": 65500
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.54,
|
| 852 |
+
"learning_rate": 9.067117559846076e-06,
|
| 853 |
+
"loss": 2.3085,
|
| 854 |
+
"step": 66000
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.54,
|
| 858 |
+
"learning_rate": 9.0587882523447e-06,
|
| 859 |
+
"loss": 2.3029,
|
| 860 |
+
"step": 66500
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.55,
|
| 864 |
+
"learning_rate": 9.050458944843327e-06,
|
| 865 |
+
"loss": 2.2954,
|
| 866 |
+
"step": 67000
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.55,
|
| 870 |
+
"learning_rate": 9.042129637341952e-06,
|
| 871 |
+
"loss": 2.303,
|
| 872 |
+
"step": 67500
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.56,
|
| 876 |
+
"learning_rate": 9.033800329840577e-06,
|
| 877 |
+
"loss": 2.3016,
|
| 878 |
+
"step": 68000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.56,
|
| 882 |
+
"learning_rate": 9.025471022339204e-06,
|
| 883 |
+
"loss": 2.3063,
|
| 884 |
+
"step": 68500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.57,
|
| 888 |
+
"learning_rate": 9.017141714837829e-06,
|
| 889 |
+
"loss": 2.3018,
|
| 890 |
+
"step": 69000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.57,
|
| 894 |
+
"learning_rate": 9.008812407336454e-06,
|
| 895 |
+
"loss": 2.3061,
|
| 896 |
+
"step": 69500
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.57,
|
| 900 |
+
"learning_rate": 9.00048309983508e-06,
|
| 901 |
+
"loss": 2.2993,
|
| 902 |
+
"step": 70000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.57,
|
| 906 |
+
"eval_accuracy": 0.5821933056736094,
|
| 907 |
+
"eval_loss": 2.223655939102173,
|
| 908 |
+
"eval_runtime": 1891.7765,
|
| 909 |
+
"eval_samples_per_second": 162.998,
|
| 910 |
+
"eval_steps_per_second": 3.396,
|
| 911 |
+
"step": 70000
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.58,
|
| 915 |
+
"learning_rate": 8.992153792333706e-06,
|
| 916 |
+
"loss": 2.2958,
|
| 917 |
+
"step": 70500
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.58,
|
| 921 |
+
"learning_rate": 8.983824484832332e-06,
|
| 922 |
+
"loss": 2.2968,
|
| 923 |
+
"step": 71000
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.59,
|
| 927 |
+
"learning_rate": 8.975495177330957e-06,
|
| 928 |
+
"loss": 2.3031,
|
| 929 |
+
"step": 71500
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.59,
|
| 933 |
+
"learning_rate": 8.967165869829582e-06,
|
| 934 |
+
"loss": 2.2966,
|
| 935 |
+
"step": 72000
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.59,
|
| 939 |
+
"learning_rate": 8.958836562328209e-06,
|
| 940 |
+
"loss": 2.2992,
|
| 941 |
+
"step": 72500
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.6,
|
| 945 |
+
"learning_rate": 8.950507254826834e-06,
|
| 946 |
+
"loss": 2.2936,
|
| 947 |
+
"step": 73000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.6,
|
| 951 |
+
"learning_rate": 8.942177947325459e-06,
|
| 952 |
+
"loss": 2.299,
|
| 953 |
+
"step": 73500
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.61,
|
| 957 |
+
"learning_rate": 8.933848639824086e-06,
|
| 958 |
+
"loss": 2.2978,
|
| 959 |
+
"step": 74000
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.61,
|
| 963 |
+
"learning_rate": 8.925519332322713e-06,
|
| 964 |
+
"loss": 2.2965,
|
| 965 |
+
"step": 74500
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 0.61,
|
| 969 |
+
"learning_rate": 8.917190024821338e-06,
|
| 970 |
+
"loss": 2.2998,
|
| 971 |
+
"step": 75000
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.62,
|
| 975 |
+
"learning_rate": 8.908860717319963e-06,
|
| 976 |
+
"loss": 2.2933,
|
| 977 |
+
"step": 75500
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.62,
|
| 981 |
+
"learning_rate": 8.900531409818588e-06,
|
| 982 |
+
"loss": 2.2915,
|
| 983 |
+
"step": 76000
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.63,
|
| 987 |
+
"learning_rate": 8.892202102317214e-06,
|
| 988 |
+
"loss": 2.29,
|
| 989 |
+
"step": 76500
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.63,
|
| 993 |
+
"learning_rate": 8.88387279481584e-06,
|
| 994 |
+
"loss": 2.2889,
|
| 995 |
+
"step": 77000
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 0.63,
|
| 999 |
+
"learning_rate": 8.875543487314466e-06,
|
| 1000 |
+
"loss": 2.2902,
|
| 1001 |
+
"step": 77500
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.64,
|
| 1005 |
+
"learning_rate": 8.867214179813091e-06,
|
| 1006 |
+
"loss": 2.2849,
|
| 1007 |
+
"step": 78000
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 0.64,
|
| 1011 |
+
"learning_rate": 8.858884872311718e-06,
|
| 1012 |
+
"loss": 2.2927,
|
| 1013 |
+
"step": 78500
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.65,
|
| 1017 |
+
"learning_rate": 8.850555564810343e-06,
|
| 1018 |
+
"loss": 2.2855,
|
| 1019 |
+
"step": 79000
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.65,
|
| 1023 |
+
"learning_rate": 8.842226257308968e-06,
|
| 1024 |
+
"loss": 2.2942,
|
| 1025 |
+
"step": 79500
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.66,
|
| 1029 |
+
"learning_rate": 8.833896949807593e-06,
|
| 1030 |
+
"loss": 2.2911,
|
| 1031 |
+
"step": 80000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.66,
|
| 1035 |
+
"eval_accuracy": 0.5830921567576977,
|
| 1036 |
+
"eval_loss": 2.2128355503082275,
|
| 1037 |
+
"eval_runtime": 1891.4198,
|
| 1038 |
+
"eval_samples_per_second": 163.029,
|
| 1039 |
+
"eval_steps_per_second": 3.397,
|
| 1040 |
+
"step": 80000
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.66,
|
| 1044 |
+
"learning_rate": 8.82556764230622e-06,
|
| 1045 |
+
"loss": 2.2885,
|
| 1046 |
+
"step": 80500
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.66,
|
| 1050 |
+
"learning_rate": 8.817238334804844e-06,
|
| 1051 |
+
"loss": 2.2926,
|
| 1052 |
+
"step": 81000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.67,
|
| 1056 |
+
"learning_rate": 8.808909027303471e-06,
|
| 1057 |
+
"loss": 2.2876,
|
| 1058 |
+
"step": 81500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.67,
|
| 1062 |
+
"learning_rate": 8.800579719802096e-06,
|
| 1063 |
+
"loss": 2.2852,
|
| 1064 |
+
"step": 82000
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.68,
|
| 1068 |
+
"learning_rate": 8.792250412300723e-06,
|
| 1069 |
+
"loss": 2.2769,
|
| 1070 |
+
"step": 82500
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.68,
|
| 1074 |
+
"learning_rate": 8.783921104799348e-06,
|
| 1075 |
+
"loss": 2.2883,
|
| 1076 |
+
"step": 83000
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.68,
|
| 1080 |
+
"learning_rate": 8.775591797297973e-06,
|
| 1081 |
+
"loss": 2.2879,
|
| 1082 |
+
"step": 83500
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.69,
|
| 1086 |
+
"learning_rate": 8.767262489796598e-06,
|
| 1087 |
+
"loss": 2.2853,
|
| 1088 |
+
"step": 84000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.69,
|
| 1092 |
+
"learning_rate": 8.758933182295224e-06,
|
| 1093 |
+
"loss": 2.2818,
|
| 1094 |
+
"step": 84500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.7,
|
| 1098 |
+
"learning_rate": 8.750603874793851e-06,
|
| 1099 |
+
"loss": 2.2824,
|
| 1100 |
+
"step": 85000
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.7,
|
| 1104 |
+
"learning_rate": 8.742274567292476e-06,
|
| 1105 |
+
"loss": 2.2873,
|
| 1106 |
+
"step": 85500
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.7,
|
| 1110 |
+
"learning_rate": 8.733945259791101e-06,
|
| 1111 |
+
"loss": 2.2792,
|
| 1112 |
+
"step": 86000
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.71,
|
| 1116 |
+
"learning_rate": 8.725615952289728e-06,
|
| 1117 |
+
"loss": 2.2823,
|
| 1118 |
+
"step": 86500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.71,
|
| 1122 |
+
"learning_rate": 8.717286644788353e-06,
|
| 1123 |
+
"loss": 2.2856,
|
| 1124 |
+
"step": 87000
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.72,
|
| 1128 |
+
"learning_rate": 8.708957337286978e-06,
|
| 1129 |
+
"loss": 2.2796,
|
| 1130 |
+
"step": 87500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.72,
|
| 1134 |
+
"learning_rate": 8.700628029785605e-06,
|
| 1135 |
+
"loss": 2.2776,
|
| 1136 |
+
"step": 88000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.73,
|
| 1140 |
+
"learning_rate": 8.69229872228423e-06,
|
| 1141 |
+
"loss": 2.2834,
|
| 1142 |
+
"step": 88500
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.73,
|
| 1146 |
+
"learning_rate": 8.683969414782856e-06,
|
| 1147 |
+
"loss": 2.2798,
|
| 1148 |
+
"step": 89000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.73,
|
| 1152 |
+
"learning_rate": 8.675640107281481e-06,
|
| 1153 |
+
"loss": 2.2759,
|
| 1154 |
+
"step": 89500
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.74,
|
| 1158 |
+
"learning_rate": 8.667310799780106e-06,
|
| 1159 |
+
"loss": 2.279,
|
| 1160 |
+
"step": 90000
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.74,
|
| 1164 |
+
"eval_accuracy": 0.5842267848402505,
|
| 1165 |
+
"eval_loss": 2.2007641792297363,
|
| 1166 |
+
"eval_runtime": 1891.3924,
|
| 1167 |
+
"eval_samples_per_second": 163.031,
|
| 1168 |
+
"eval_steps_per_second": 3.397,
|
| 1169 |
+
"step": 90000
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 0.74,
|
| 1173 |
+
"learning_rate": 8.658981492278733e-06,
|
| 1174 |
+
"loss": 2.2741,
|
| 1175 |
+
"step": 90500
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.75,
|
| 1179 |
+
"learning_rate": 8.650652184777358e-06,
|
| 1180 |
+
"loss": 2.277,
|
| 1181 |
+
"step": 91000
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 0.75,
|
| 1185 |
+
"learning_rate": 8.642322877275983e-06,
|
| 1186 |
+
"loss": 2.2797,
|
| 1187 |
+
"step": 91500
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 0.75,
|
| 1191 |
+
"learning_rate": 8.63399356977461e-06,
|
| 1192 |
+
"loss": 2.2834,
|
| 1193 |
+
"step": 92000
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.76,
|
| 1197 |
+
"learning_rate": 8.625664262273236e-06,
|
| 1198 |
+
"loss": 2.2777,
|
| 1199 |
+
"step": 92500
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.76,
|
| 1203 |
+
"learning_rate": 8.617334954771861e-06,
|
| 1204 |
+
"loss": 2.2804,
|
| 1205 |
+
"step": 93000
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 0.77,
|
| 1209 |
+
"learning_rate": 8.609005647270486e-06,
|
| 1210 |
+
"loss": 2.2734,
|
| 1211 |
+
"step": 93500
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 0.77,
|
| 1215 |
+
"learning_rate": 8.600676339769111e-06,
|
| 1216 |
+
"loss": 2.2727,
|
| 1217 |
+
"step": 94000
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.77,
|
| 1221 |
+
"learning_rate": 8.592347032267738e-06,
|
| 1222 |
+
"loss": 2.2745,
|
| 1223 |
+
"step": 94500
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 0.78,
|
| 1227 |
+
"learning_rate": 8.584017724766363e-06,
|
| 1228 |
+
"loss": 2.2736,
|
| 1229 |
+
"step": 95000
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.78,
|
| 1233 |
+
"learning_rate": 8.57568841726499e-06,
|
| 1234 |
+
"loss": 2.2764,
|
| 1235 |
+
"step": 95500
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.79,
|
| 1239 |
+
"learning_rate": 8.567359109763615e-06,
|
| 1240 |
+
"loss": 2.2744,
|
| 1241 |
+
"step": 96000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.79,
|
| 1245 |
+
"learning_rate": 8.559029802262241e-06,
|
| 1246 |
+
"loss": 2.2704,
|
| 1247 |
+
"step": 96500
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 0.79,
|
| 1251 |
+
"learning_rate": 8.550700494760866e-06,
|
| 1252 |
+
"loss": 2.2738,
|
| 1253 |
+
"step": 97000
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 0.8,
|
| 1257 |
+
"learning_rate": 8.542371187259491e-06,
|
| 1258 |
+
"loss": 2.2739,
|
| 1259 |
+
"step": 97500
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.8,
|
| 1263 |
+
"learning_rate": 8.534041879758118e-06,
|
| 1264 |
+
"loss": 2.274,
|
| 1265 |
+
"step": 98000
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.81,
|
| 1269 |
+
"learning_rate": 8.525712572256743e-06,
|
| 1270 |
+
"loss": 2.2783,
|
| 1271 |
+
"step": 98500
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.81,
|
| 1275 |
+
"learning_rate": 8.517383264755368e-06,
|
| 1276 |
+
"loss": 2.275,
|
| 1277 |
+
"step": 99000
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.82,
|
| 1281 |
+
"learning_rate": 8.509053957253995e-06,
|
| 1282 |
+
"loss": 2.268,
|
| 1283 |
+
"step": 99500
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.82,
|
| 1287 |
+
"learning_rate": 8.500724649752622e-06,
|
| 1288 |
+
"loss": 2.2715,
|
| 1289 |
+
"step": 100000
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 0.82,
|
| 1293 |
+
"eval_accuracy": 0.5856086866591657,
|
| 1294 |
+
"eval_loss": 2.196564197540283,
|
| 1295 |
+
"eval_runtime": 1891.0578,
|
| 1296 |
+
"eval_samples_per_second": 163.06,
|
| 1297 |
+
"eval_steps_per_second": 3.398,
|
| 1298 |
+
"step": 100000
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.82,
|
| 1302 |
+
"learning_rate": 8.492395342251247e-06,
|
| 1303 |
+
"loss": 2.2917,
|
| 1304 |
+
"step": 100500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.83,
|
| 1308 |
+
"learning_rate": 8.484066034749872e-06,
|
| 1309 |
+
"loss": 2.281,
|
| 1310 |
+
"step": 101000
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 0.83,
|
| 1314 |
+
"learning_rate": 8.475736727248497e-06,
|
| 1315 |
+
"loss": 2.2912,
|
| 1316 |
+
"step": 101500
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 0.83,
|
| 1320 |
+
"step": 101882,
|
| 1321 |
+
"total_flos": 1.4489445094328893e+18,
|
| 1322 |
+
"train_loss": 2.3200887040276985,
|
| 1323 |
+
"train_runtime": 107983.587,
|
| 1324 |
+
"train_samples_per_second": 271.28,
|
| 1325 |
+
"train_steps_per_second": 5.652
|
| 1326 |
+
}
|
| 1327 |
+
],
|
| 1328 |
+
"max_steps": 610290,
|
| 1329 |
+
"num_train_epochs": 5,
|
| 1330 |
+
"total_flos": 1.4489445094328893e+18,
|
| 1331 |
+
"trial_name": null,
|
| 1332 |
+
"trial_params": null
|
| 1333 |
+
}
|