End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +2353 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.99,
|
| 3 |
+
"eval_accuracy": 0.4732233053015494,
|
| 4 |
+
"eval_loss": 3.0945498943328857,
|
| 5 |
+
"eval_runtime": 7.191,
|
| 6 |
+
"eval_samples": 2017,
|
| 7 |
+
"eval_samples_per_second": 280.489,
|
| 8 |
+
"eval_steps_per_second": 140.314,
|
| 9 |
+
"perplexity": 22.077299183356594,
|
| 10 |
+
"train_loss": 3.6050218030916037,
|
| 11 |
+
"train_runtime": 557.9113,
|
| 12 |
+
"train_samples": 18139,
|
| 13 |
+
"train_samples_per_second": 260.099,
|
| 14 |
+
"train_steps_per_second": 2.022
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.99,
|
| 3 |
+
"eval_accuracy": 0.4732233053015494,
|
| 4 |
+
"eval_loss": 3.0945498943328857,
|
| 5 |
+
"eval_runtime": 7.191,
|
| 6 |
+
"eval_samples": 2017,
|
| 7 |
+
"eval_samples_per_second": 280.489,
|
| 8 |
+
"eval_steps_per_second": 140.314,
|
| 9 |
+
"perplexity": 22.077299183356594
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 7.99,
|
| 3 |
+
"train_loss": 3.6050218030916037,
|
| 4 |
+
"train_runtime": 557.9113,
|
| 5 |
+
"train_samples": 18139,
|
| 6 |
+
"train_samples_per_second": 260.099,
|
| 7 |
+
"train_steps_per_second": 2.022
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2353 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.47660093674971465,
|
| 3 |
+
"best_model_checkpoint": "./runtime-masked/bert_uncased_L-2_H-128_A-2-mlm-multi-emails-hq/checkpoint-987",
|
| 4 |
+
"epoch": 7.994708994708994,
|
| 5 |
+
"global_step": 1128,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.02,
|
| 12 |
+
"learning_rate": 1.5789473684210522e-05,
|
| 13 |
+
"loss": 5.0353,
|
| 14 |
+
"step": 3
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 3.1578947368421045e-05,
|
| 19 |
+
"loss": 5.0947,
|
| 20 |
+
"step": 6
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.06,
|
| 24 |
+
"learning_rate": 4.7368421052631574e-05,
|
| 25 |
+
"loss": 5.0512,
|
| 26 |
+
"step": 9
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.08,
|
| 30 |
+
"learning_rate": 6.315789473684209e-05,
|
| 31 |
+
"loss": 4.9084,
|
| 32 |
+
"step": 12
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.11,
|
| 36 |
+
"learning_rate": 7.894736842105262e-05,
|
| 37 |
+
"loss": 4.9194,
|
| 38 |
+
"step": 15
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.13,
|
| 42 |
+
"learning_rate": 9.473684210526315e-05,
|
| 43 |
+
"loss": 4.8347,
|
| 44 |
+
"step": 18
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.15,
|
| 48 |
+
"learning_rate": 0.00011052631578947366,
|
| 49 |
+
"loss": 4.7759,
|
| 50 |
+
"step": 21
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.17,
|
| 54 |
+
"learning_rate": 0.00012631578947368418,
|
| 55 |
+
"loss": 4.7372,
|
| 56 |
+
"step": 24
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.19,
|
| 60 |
+
"learning_rate": 0.0001421052631578947,
|
| 61 |
+
"loss": 4.7002,
|
| 62 |
+
"step": 27
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.21,
|
| 66 |
+
"learning_rate": 0.00015789473684210524,
|
| 67 |
+
"loss": 4.6299,
|
| 68 |
+
"step": 30
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.23,
|
| 72 |
+
"learning_rate": 0.0001736842105263158,
|
| 73 |
+
"loss": 4.5503,
|
| 74 |
+
"step": 33
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.25,
|
| 78 |
+
"learning_rate": 0.0001894736842105263,
|
| 79 |
+
"loss": 4.5794,
|
| 80 |
+
"step": 36
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.28,
|
| 84 |
+
"learning_rate": 0.00020526315789473683,
|
| 85 |
+
"loss": 4.4404,
|
| 86 |
+
"step": 39
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3,
|
| 90 |
+
"learning_rate": 0.00022105263157894733,
|
| 91 |
+
"loss": 4.5015,
|
| 92 |
+
"step": 42
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.32,
|
| 96 |
+
"learning_rate": 0.00023684210526315788,
|
| 97 |
+
"loss": 4.3954,
|
| 98 |
+
"step": 45
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.34,
|
| 102 |
+
"learning_rate": 0.00025263157894736836,
|
| 103 |
+
"loss": 4.3344,
|
| 104 |
+
"step": 48
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.36,
|
| 108 |
+
"learning_rate": 0.0002684210526315789,
|
| 109 |
+
"loss": 4.3077,
|
| 110 |
+
"step": 51
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.38,
|
| 114 |
+
"learning_rate": 0.0002842105263157894,
|
| 115 |
+
"loss": 4.275,
|
| 116 |
+
"step": 54
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.4,
|
| 120 |
+
"learning_rate": 0.0003,
|
| 121 |
+
"loss": 4.2386,
|
| 122 |
+
"step": 57
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.42,
|
| 126 |
+
"learning_rate": 0.00029999419206464364,
|
| 127 |
+
"loss": 4.2538,
|
| 128 |
+
"step": 60
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.44,
|
| 132 |
+
"learning_rate": 0.0002999767687083362,
|
| 133 |
+
"loss": 4.1491,
|
| 134 |
+
"step": 63
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.47,
|
| 138 |
+
"learning_rate": 0.0002999477312803273,
|
| 139 |
+
"loss": 4.1288,
|
| 140 |
+
"step": 66
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.49,
|
| 144 |
+
"learning_rate": 0.00029990708202925034,
|
| 145 |
+
"loss": 4.1081,
|
| 146 |
+
"step": 69
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.51,
|
| 150 |
+
"learning_rate": 0.0002998548241029483,
|
| 151 |
+
"loss": 4.1276,
|
| 152 |
+
"step": 72
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.53,
|
| 156 |
+
"learning_rate": 0.00029979096154823,
|
| 157 |
+
"loss": 4.0664,
|
| 158 |
+
"step": 75
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.55,
|
| 162 |
+
"learning_rate": 0.0002997154993105566,
|
| 163 |
+
"loss": 4.0946,
|
| 164 |
+
"step": 78
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.57,
|
| 168 |
+
"learning_rate": 0.0002996284432336587,
|
| 169 |
+
"loss": 4.0326,
|
| 170 |
+
"step": 81
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.59,
|
| 174 |
+
"learning_rate": 0.00029952980005908385,
|
| 175 |
+
"loss": 4.1449,
|
| 176 |
+
"step": 84
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.61,
|
| 180 |
+
"learning_rate": 0.0002994195774256746,
|
| 181 |
+
"loss": 4.0859,
|
| 182 |
+
"step": 87
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.63,
|
| 186 |
+
"learning_rate": 0.0002992977838689765,
|
| 187 |
+
"loss": 4.052,
|
| 188 |
+
"step": 90
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.66,
|
| 192 |
+
"learning_rate": 0.0002991644288205777,
|
| 193 |
+
"loss": 3.9777,
|
| 194 |
+
"step": 93
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.68,
|
| 198 |
+
"learning_rate": 0.00029901952260737825,
|
| 199 |
+
"loss": 4.0233,
|
| 200 |
+
"step": 96
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.7,
|
| 204 |
+
"learning_rate": 0.0002988630764507904,
|
| 205 |
+
"loss": 3.9186,
|
| 206 |
+
"step": 99
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.72,
|
| 210 |
+
"learning_rate": 0.0002986951024658695,
|
| 211 |
+
"loss": 3.9352,
|
| 212 |
+
"step": 102
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.74,
|
| 216 |
+
"learning_rate": 0.0002985156136603764,
|
| 217 |
+
"loss": 3.9825,
|
| 218 |
+
"step": 105
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.76,
|
| 222 |
+
"learning_rate": 0.0002983246239337692,
|
| 223 |
+
"loss": 3.9961,
|
| 224 |
+
"step": 108
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.78,
|
| 228 |
+
"learning_rate": 0.0002981221480761281,
|
| 229 |
+
"loss": 3.9801,
|
| 230 |
+
"step": 111
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.8,
|
| 234 |
+
"learning_rate": 0.0002979082017670087,
|
| 235 |
+
"loss": 3.9136,
|
| 236 |
+
"step": 114
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.83,
|
| 240 |
+
"learning_rate": 0.00029768280157422883,
|
| 241 |
+
"loss": 3.8618,
|
| 242 |
+
"step": 117
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.85,
|
| 246 |
+
"learning_rate": 0.00029744596495258525,
|
| 247 |
+
"loss": 3.8953,
|
| 248 |
+
"step": 120
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.87,
|
| 252 |
+
"learning_rate": 0.00029719771024250166,
|
| 253 |
+
"loss": 3.95,
|
| 254 |
+
"step": 123
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.89,
|
| 258 |
+
"learning_rate": 0.000296938056668609,
|
| 259 |
+
"loss": 3.8829,
|
| 260 |
+
"step": 126
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.91,
|
| 264 |
+
"learning_rate": 0.0002966670243382561,
|
| 265 |
+
"loss": 3.8937,
|
| 266 |
+
"step": 129
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.93,
|
| 270 |
+
"learning_rate": 0.00029638463423995305,
|
| 271 |
+
"loss": 3.8596,
|
| 272 |
+
"step": 132
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.95,
|
| 276 |
+
"learning_rate": 0.0002960909082417457,
|
| 277 |
+
"loss": 3.8494,
|
| 278 |
+
"step": 135
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.97,
|
| 282 |
+
"learning_rate": 0.0002957858690895221,
|
| 283 |
+
"loss": 3.8545,
|
| 284 |
+
"step": 138
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.99,
|
| 288 |
+
"learning_rate": 0.0002954695404052514,
|
| 289 |
+
"loss": 3.8974,
|
| 290 |
+
"step": 141
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.99,
|
| 294 |
+
"eval_accuracy": 0.42182521025962494,
|
| 295 |
+
"eval_loss": 3.5129101276397705,
|
| 296 |
+
"eval_runtime": 7.1205,
|
| 297 |
+
"eval_samples_per_second": 283.265,
|
| 298 |
+
"eval_steps_per_second": 141.703,
|
| 299 |
+
"step": 141
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.02,
|
| 303 |
+
"learning_rate": 0.00029514194668515416,
|
| 304 |
+
"loss": 4.7932,
|
| 305 |
+
"step": 144
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.04,
|
| 309 |
+
"learning_rate": 0.00029480311329780576,
|
| 310 |
+
"loss": 3.8152,
|
| 311 |
+
"step": 147
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.06,
|
| 315 |
+
"learning_rate": 0.0002944530664821717,
|
| 316 |
+
"loss": 3.7853,
|
| 317 |
+
"step": 150
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.08,
|
| 321 |
+
"learning_rate": 0.00029409183334557556,
|
| 322 |
+
"loss": 3.7879,
|
| 323 |
+
"step": 153
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 1.11,
|
| 327 |
+
"learning_rate": 0.00029371944186160016,
|
| 328 |
+
"loss": 3.8584,
|
| 329 |
+
"step": 156
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.13,
|
| 333 |
+
"learning_rate": 0.00029333592086792107,
|
| 334 |
+
"loss": 3.8126,
|
| 335 |
+
"step": 159
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 1.15,
|
| 339 |
+
"learning_rate": 0.0002929413000640735,
|
| 340 |
+
"loss": 3.8505,
|
| 341 |
+
"step": 162
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 1.17,
|
| 345 |
+
"learning_rate": 0.0002925356100091522,
|
| 346 |
+
"loss": 3.8638,
|
| 347 |
+
"step": 165
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.19,
|
| 351 |
+
"learning_rate": 0.00029211888211944553,
|
| 352 |
+
"loss": 3.8271,
|
| 353 |
+
"step": 168
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.21,
|
| 357 |
+
"learning_rate": 0.0002916911486660021,
|
| 358 |
+
"loss": 3.76,
|
| 359 |
+
"step": 171
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.23,
|
| 363 |
+
"learning_rate": 0.0002912524427721317,
|
| 364 |
+
"loss": 3.7647,
|
| 365 |
+
"step": 174
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 1.25,
|
| 369 |
+
"learning_rate": 0.00029080279841084075,
|
| 370 |
+
"loss": 3.7656,
|
| 371 |
+
"step": 177
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.28,
|
| 375 |
+
"learning_rate": 0.000290342250402201,
|
| 376 |
+
"loss": 3.7425,
|
| 377 |
+
"step": 180
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.3,
|
| 381 |
+
"learning_rate": 0.0002898708344106533,
|
| 382 |
+
"loss": 3.7809,
|
| 383 |
+
"step": 183
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.32,
|
| 387 |
+
"learning_rate": 0.00028938858694224574,
|
| 388 |
+
"loss": 3.8349,
|
| 389 |
+
"step": 186
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.34,
|
| 393 |
+
"learning_rate": 0.0002888955453418066,
|
| 394 |
+
"loss": 3.7542,
|
| 395 |
+
"step": 189
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.36,
|
| 399 |
+
"learning_rate": 0.0002883917477900524,
|
| 400 |
+
"loss": 3.7652,
|
| 401 |
+
"step": 192
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.38,
|
| 405 |
+
"learning_rate": 0.0002878772333006314,
|
| 406 |
+
"loss": 3.7324,
|
| 407 |
+
"step": 195
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 1.4,
|
| 411 |
+
"learning_rate": 0.0002873520417171021,
|
| 412 |
+
"loss": 3.7123,
|
| 413 |
+
"step": 198
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.42,
|
| 417 |
+
"learning_rate": 0.00028681621370984817,
|
| 418 |
+
"loss": 3.7046,
|
| 419 |
+
"step": 201
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 1.44,
|
| 423 |
+
"learning_rate": 0.0002862697907729285,
|
| 424 |
+
"loss": 3.6747,
|
| 425 |
+
"step": 204
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.47,
|
| 429 |
+
"learning_rate": 0.00028571281522086453,
|
| 430 |
+
"loss": 3.6714,
|
| 431 |
+
"step": 207
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.49,
|
| 435 |
+
"learning_rate": 0.0002851453301853628,
|
| 436 |
+
"loss": 3.7639,
|
| 437 |
+
"step": 210
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.51,
|
| 441 |
+
"learning_rate": 0.0002845673796119755,
|
| 442 |
+
"loss": 3.811,
|
| 443 |
+
"step": 213
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.53,
|
| 447 |
+
"learning_rate": 0.0002839790082566967,
|
| 448 |
+
"loss": 3.7407,
|
| 449 |
+
"step": 216
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 1.55,
|
| 453 |
+
"learning_rate": 0.00028338026168249714,
|
| 454 |
+
"loss": 3.7508,
|
| 455 |
+
"step": 219
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.57,
|
| 459 |
+
"learning_rate": 0.0002827711862557954,
|
| 460 |
+
"loss": 3.6624,
|
| 461 |
+
"step": 222
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.59,
|
| 465 |
+
"learning_rate": 0.00028215182914286766,
|
| 466 |
+
"loss": 3.7081,
|
| 467 |
+
"step": 225
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 1.61,
|
| 471 |
+
"learning_rate": 0.0002815222383061948,
|
| 472 |
+
"loss": 3.7027,
|
| 473 |
+
"step": 228
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.63,
|
| 477 |
+
"learning_rate": 0.00028088246250074857,
|
| 478 |
+
"loss": 3.7365,
|
| 479 |
+
"step": 231
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.66,
|
| 483 |
+
"learning_rate": 0.0002802325512702159,
|
| 484 |
+
"loss": 3.6906,
|
| 485 |
+
"step": 234
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.68,
|
| 489 |
+
"learning_rate": 0.0002795725549431624,
|
| 490 |
+
"loss": 3.7296,
|
| 491 |
+
"step": 237
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 1.7,
|
| 495 |
+
"learning_rate": 0.0002789025246291347,
|
| 496 |
+
"loss": 3.6572,
|
| 497 |
+
"step": 240
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.72,
|
| 501 |
+
"learning_rate": 0.0002782225122147029,
|
| 502 |
+
"loss": 3.7676,
|
| 503 |
+
"step": 243
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 1.74,
|
| 507 |
+
"learning_rate": 0.0002775325703594421,
|
| 508 |
+
"loss": 3.7506,
|
| 509 |
+
"step": 246
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.76,
|
| 513 |
+
"learning_rate": 0.00027683275249185504,
|
| 514 |
+
"loss": 3.6966,
|
| 515 |
+
"step": 249
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.78,
|
| 519 |
+
"learning_rate": 0.0002761231128052341,
|
| 520 |
+
"loss": 3.6229,
|
| 521 |
+
"step": 252
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.8,
|
| 525 |
+
"learning_rate": 0.00027540370625346507,
|
| 526 |
+
"loss": 3.6964,
|
| 527 |
+
"step": 255
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.83,
|
| 531 |
+
"learning_rate": 0.0002746745885467712,
|
| 532 |
+
"loss": 3.6367,
|
| 533 |
+
"step": 258
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 1.85,
|
| 537 |
+
"learning_rate": 0.00027393581614739923,
|
| 538 |
+
"loss": 3.6899,
|
| 539 |
+
"step": 261
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.87,
|
| 543 |
+
"learning_rate": 0.000273187446265247,
|
| 544 |
+
"loss": 3.661,
|
| 545 |
+
"step": 264
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 1.89,
|
| 549 |
+
"learning_rate": 0.00027242953685343327,
|
| 550 |
+
"loss": 3.7046,
|
| 551 |
+
"step": 267
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 1.91,
|
| 555 |
+
"learning_rate": 0.0002716621466038095,
|
| 556 |
+
"loss": 3.6271,
|
| 557 |
+
"step": 270
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.93,
|
| 561 |
+
"learning_rate": 0.0002708853349424152,
|
| 562 |
+
"loss": 3.5838,
|
| 563 |
+
"step": 273
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.95,
|
| 567 |
+
"learning_rate": 0.000270099162024876,
|
| 568 |
+
"loss": 3.6223,
|
| 569 |
+
"step": 276
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.97,
|
| 573 |
+
"learning_rate": 0.0002693036887317449,
|
| 574 |
+
"loss": 3.6492,
|
| 575 |
+
"step": 279
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 1.99,
|
| 579 |
+
"learning_rate": 0.0002684989766637882,
|
| 580 |
+
"loss": 3.7009,
|
| 581 |
+
"step": 282
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 1.99,
|
| 585 |
+
"eval_accuracy": 0.44521651429020515,
|
| 586 |
+
"eval_loss": 3.329479217529297,
|
| 587 |
+
"eval_runtime": 7.098,
|
| 588 |
+
"eval_samples_per_second": 284.164,
|
| 589 |
+
"eval_steps_per_second": 142.153,
|
| 590 |
+
"step": 282
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.02,
|
| 594 |
+
"learning_rate": 0.0002676850881372147,
|
| 595 |
+
"loss": 4.5343,
|
| 596 |
+
"step": 285
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 2.04,
|
| 600 |
+
"learning_rate": 0.00026686208617885055,
|
| 601 |
+
"loss": 3.6713,
|
| 602 |
+
"step": 288
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 2.06,
|
| 606 |
+
"learning_rate": 0.0002660300345212579,
|
| 607 |
+
"loss": 3.5313,
|
| 608 |
+
"step": 291
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.08,
|
| 612 |
+
"learning_rate": 0.0002651889975978001,
|
| 613 |
+
"loss": 3.6284,
|
| 614 |
+
"step": 294
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.11,
|
| 618 |
+
"learning_rate": 0.0002643390405376515,
|
| 619 |
+
"loss": 3.6498,
|
| 620 |
+
"step": 297
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.13,
|
| 624 |
+
"learning_rate": 0.00026348022916075436,
|
| 625 |
+
"loss": 3.5984,
|
| 626 |
+
"step": 300
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 2.15,
|
| 630 |
+
"learning_rate": 0.00026261262997272135,
|
| 631 |
+
"loss": 3.6395,
|
| 632 |
+
"step": 303
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.17,
|
| 636 |
+
"learning_rate": 0.00026173631015968585,
|
| 637 |
+
"loss": 3.6397,
|
| 638 |
+
"step": 306
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.19,
|
| 642 |
+
"learning_rate": 0.00026085133758309883,
|
| 643 |
+
"loss": 3.6392,
|
| 644 |
+
"step": 309
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 2.21,
|
| 648 |
+
"learning_rate": 0.0002599577807744739,
|
| 649 |
+
"loss": 3.6054,
|
| 650 |
+
"step": 312
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.23,
|
| 654 |
+
"learning_rate": 0.0002590557089300802,
|
| 655 |
+
"loss": 3.649,
|
| 656 |
+
"step": 315
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.25,
|
| 660 |
+
"learning_rate": 0.0002581451919055837,
|
| 661 |
+
"loss": 3.5249,
|
| 662 |
+
"step": 318
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.28,
|
| 666 |
+
"learning_rate": 0.0002572263002106381,
|
| 667 |
+
"loss": 3.599,
|
| 668 |
+
"step": 321
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 2.3,
|
| 672 |
+
"learning_rate": 0.00025629910500342423,
|
| 673 |
+
"loss": 3.6279,
|
| 674 |
+
"step": 324
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.32,
|
| 678 |
+
"learning_rate": 0.00025536367808513973,
|
| 679 |
+
"loss": 3.5986,
|
| 680 |
+
"step": 327
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 2.34,
|
| 684 |
+
"learning_rate": 0.000254420091894439,
|
| 685 |
+
"loss": 3.613,
|
| 686 |
+
"step": 330
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 2.36,
|
| 690 |
+
"learning_rate": 0.0002534684195018232,
|
| 691 |
+
"loss": 3.5667,
|
| 692 |
+
"step": 333
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.38,
|
| 696 |
+
"learning_rate": 0.0002525087346039822,
|
| 697 |
+
"loss": 3.5672,
|
| 698 |
+
"step": 336
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 2.4,
|
| 702 |
+
"learning_rate": 0.0002515411115180875,
|
| 703 |
+
"loss": 3.5449,
|
| 704 |
+
"step": 339
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 2.42,
|
| 708 |
+
"learning_rate": 0.00025056562517603666,
|
| 709 |
+
"loss": 3.5743,
|
| 710 |
+
"step": 342
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 2.44,
|
| 714 |
+
"learning_rate": 0.0002495823511186512,
|
| 715 |
+
"loss": 3.5829,
|
| 716 |
+
"step": 345
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.47,
|
| 720 |
+
"learning_rate": 0.0002485913654898268,
|
| 721 |
+
"loss": 3.6092,
|
| 722 |
+
"step": 348
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 2.49,
|
| 726 |
+
"learning_rate": 0.0002475927450306363,
|
| 727 |
+
"loss": 3.5628,
|
| 728 |
+
"step": 351
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 2.51,
|
| 732 |
+
"learning_rate": 0.0002465865670733873,
|
| 733 |
+
"loss": 3.613,
|
| 734 |
+
"step": 354
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 2.53,
|
| 738 |
+
"learning_rate": 0.00024557290953563366,
|
| 739 |
+
"loss": 3.5926,
|
| 740 |
+
"step": 357
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 2.55,
|
| 744 |
+
"learning_rate": 0.00024455185091414136,
|
| 745 |
+
"loss": 3.6004,
|
| 746 |
+
"step": 360
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 2.57,
|
| 750 |
+
"learning_rate": 0.00024352347027881003,
|
| 751 |
+
"loss": 3.6115,
|
| 752 |
+
"step": 363
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 2.59,
|
| 756 |
+
"learning_rate": 0.00024248784726654961,
|
| 757 |
+
"loss": 3.6296,
|
| 758 |
+
"step": 366
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.61,
|
| 762 |
+
"learning_rate": 0.0002414450620751136,
|
| 763 |
+
"loss": 3.567,
|
| 764 |
+
"step": 369
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 2.63,
|
| 768 |
+
"learning_rate": 0.00024039519545688846,
|
| 769 |
+
"loss": 3.5447,
|
| 770 |
+
"step": 372
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 2.66,
|
| 774 |
+
"learning_rate": 0.00023933832871264016,
|
| 775 |
+
"loss": 3.5445,
|
| 776 |
+
"step": 375
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.68,
|
| 780 |
+
"learning_rate": 0.00023827454368521844,
|
| 781 |
+
"loss": 3.602,
|
| 782 |
+
"step": 378
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.7,
|
| 786 |
+
"learning_rate": 0.00023720392275321895,
|
| 787 |
+
"loss": 3.5547,
|
| 788 |
+
"step": 381
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 2.72,
|
| 792 |
+
"learning_rate": 0.0002361265488246039,
|
| 793 |
+
"loss": 3.5888,
|
| 794 |
+
"step": 384
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 2.74,
|
| 798 |
+
"learning_rate": 0.0002350425053302817,
|
| 799 |
+
"loss": 3.555,
|
| 800 |
+
"step": 387
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.76,
|
| 804 |
+
"learning_rate": 0.0002339518762176462,
|
| 805 |
+
"loss": 3.5593,
|
| 806 |
+
"step": 390
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 2.78,
|
| 810 |
+
"learning_rate": 0.00023285474594407585,
|
| 811 |
+
"loss": 3.5636,
|
| 812 |
+
"step": 393
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 2.8,
|
| 816 |
+
"learning_rate": 0.00023175119947039342,
|
| 817 |
+
"loss": 3.5646,
|
| 818 |
+
"step": 396
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.83,
|
| 822 |
+
"learning_rate": 0.00023064132225428659,
|
| 823 |
+
"loss": 3.5013,
|
| 824 |
+
"step": 399
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.85,
|
| 828 |
+
"learning_rate": 0.0002295252002436904,
|
| 829 |
+
"loss": 3.4809,
|
| 830 |
+
"step": 402
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.87,
|
| 834 |
+
"learning_rate": 0.00022840291987013143,
|
| 835 |
+
"loss": 3.4732,
|
| 836 |
+
"step": 405
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.89,
|
| 840 |
+
"learning_rate": 0.00022727456804203436,
|
| 841 |
+
"loss": 3.4971,
|
| 842 |
+
"step": 408
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.91,
|
| 846 |
+
"learning_rate": 0.00022614023213799234,
|
| 847 |
+
"loss": 3.502,
|
| 848 |
+
"step": 411
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 2.93,
|
| 852 |
+
"learning_rate": 0.000225,
|
| 853 |
+
"loss": 3.5042,
|
| 854 |
+
"step": 414
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 2.95,
|
| 858 |
+
"learning_rate": 0.00022385395992665128,
|
| 859 |
+
"loss": 3.6605,
|
| 860 |
+
"step": 417
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.97,
|
| 864 |
+
"learning_rate": 0.0002227022006663018,
|
| 865 |
+
"loss": 3.5878,
|
| 866 |
+
"step": 420
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.99,
|
| 870 |
+
"learning_rate": 0.00022154481141019585,
|
| 871 |
+
"loss": 3.5845,
|
| 872 |
+
"step": 423
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.99,
|
| 876 |
+
"eval_accuracy": 0.45894501850558433,
|
| 877 |
+
"eval_loss": 3.2218551635742188,
|
| 878 |
+
"eval_runtime": 7.1194,
|
| 879 |
+
"eval_samples_per_second": 283.31,
|
| 880 |
+
"eval_steps_per_second": 141.725,
|
| 881 |
+
"step": 423
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 3.02,
|
| 885 |
+
"learning_rate": 0.00022038188178555992,
|
| 886 |
+
"loss": 4.4562,
|
| 887 |
+
"step": 426
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 3.04,
|
| 891 |
+
"learning_rate": 0.00021921350184866178,
|
| 892 |
+
"loss": 3.5152,
|
| 893 |
+
"step": 429
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 3.06,
|
| 897 |
+
"learning_rate": 0.0002180397620778366,
|
| 898 |
+
"loss": 3.5466,
|
| 899 |
+
"step": 432
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 3.08,
|
| 903 |
+
"learning_rate": 0.00021686075336648075,
|
| 904 |
+
"loss": 3.5805,
|
| 905 |
+
"step": 435
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.11,
|
| 909 |
+
"learning_rate": 0.00021567656701601247,
|
| 910 |
+
"loss": 3.4612,
|
| 911 |
+
"step": 438
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 3.13,
|
| 915 |
+
"learning_rate": 0.00021448729472880215,
|
| 916 |
+
"loss": 3.5251,
|
| 917 |
+
"step": 441
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 3.15,
|
| 921 |
+
"learning_rate": 0.00021329302860107063,
|
| 922 |
+
"loss": 3.5363,
|
| 923 |
+
"step": 444
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 3.17,
|
| 927 |
+
"learning_rate": 0.00021209386111575753,
|
| 928 |
+
"loss": 3.5013,
|
| 929 |
+
"step": 447
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 3.19,
|
| 933 |
+
"learning_rate": 0.00021088988513535931,
|
| 934 |
+
"loss": 3.542,
|
| 935 |
+
"step": 450
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 3.21,
|
| 939 |
+
"learning_rate": 0.0002096811938947381,
|
| 940 |
+
"loss": 3.5095,
|
| 941 |
+
"step": 453
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 3.23,
|
| 945 |
+
"learning_rate": 0.00020846788099390188,
|
| 946 |
+
"loss": 3.5267,
|
| 947 |
+
"step": 456
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.25,
|
| 951 |
+
"learning_rate": 0.00020725004039075587,
|
| 952 |
+
"loss": 3.4362,
|
| 953 |
+
"step": 459
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 3.28,
|
| 957 |
+
"learning_rate": 0.00020602776639382672,
|
| 958 |
+
"loss": 3.5144,
|
| 959 |
+
"step": 462
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 3.3,
|
| 963 |
+
"learning_rate": 0.00020480115365495926,
|
| 964 |
+
"loss": 3.4616,
|
| 965 |
+
"step": 465
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 3.32,
|
| 969 |
+
"learning_rate": 0.0002035702971619867,
|
| 970 |
+
"loss": 3.5137,
|
| 971 |
+
"step": 468
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 3.34,
|
| 975 |
+
"learning_rate": 0.00020233529223137502,
|
| 976 |
+
"loss": 3.4768,
|
| 977 |
+
"step": 471
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 3.36,
|
| 981 |
+
"learning_rate": 0.00020109623450084154,
|
| 982 |
+
"loss": 3.5287,
|
| 983 |
+
"step": 474
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 3.38,
|
| 987 |
+
"learning_rate": 0.00019985321992194892,
|
| 988 |
+
"loss": 3.5211,
|
| 989 |
+
"step": 477
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.4,
|
| 993 |
+
"learning_rate": 0.00019860634475267488,
|
| 994 |
+
"loss": 3.5589,
|
| 995 |
+
"step": 480
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 3.42,
|
| 999 |
+
"learning_rate": 0.00019735570554995776,
|
| 1000 |
+
"loss": 3.4963,
|
| 1001 |
+
"step": 483
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 3.44,
|
| 1005 |
+
"learning_rate": 0.00019610139916221943,
|
| 1006 |
+
"loss": 3.5406,
|
| 1007 |
+
"step": 486
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 3.47,
|
| 1011 |
+
"learning_rate": 0.0001948435227218655,
|
| 1012 |
+
"loss": 3.5183,
|
| 1013 |
+
"step": 489
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 3.49,
|
| 1017 |
+
"learning_rate": 0.0001935821736377634,
|
| 1018 |
+
"loss": 3.5413,
|
| 1019 |
+
"step": 492
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 3.51,
|
| 1023 |
+
"learning_rate": 0.0001923174495876989,
|
| 1024 |
+
"loss": 3.4245,
|
| 1025 |
+
"step": 495
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 3.53,
|
| 1029 |
+
"learning_rate": 0.00019104944851081244,
|
| 1030 |
+
"loss": 3.5263,
|
| 1031 |
+
"step": 498
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.55,
|
| 1035 |
+
"learning_rate": 0.00018977826860001442,
|
| 1036 |
+
"loss": 3.4864,
|
| 1037 |
+
"step": 501
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 3.57,
|
| 1041 |
+
"learning_rate": 0.00018850400829438155,
|
| 1042 |
+
"loss": 3.5034,
|
| 1043 |
+
"step": 504
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 3.59,
|
| 1047 |
+
"learning_rate": 0.00018722676627153358,
|
| 1048 |
+
"loss": 3.4382,
|
| 1049 |
+
"step": 507
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 3.61,
|
| 1053 |
+
"learning_rate": 0.00018594664143999187,
|
| 1054 |
+
"loss": 3.5283,
|
| 1055 |
+
"step": 510
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.63,
|
| 1059 |
+
"learning_rate": 0.00018466373293152,
|
| 1060 |
+
"loss": 3.5367,
|
| 1061 |
+
"step": 513
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 3.66,
|
| 1065 |
+
"learning_rate": 0.00018337814009344714,
|
| 1066 |
+
"loss": 3.5376,
|
| 1067 |
+
"step": 516
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 3.68,
|
| 1071 |
+
"learning_rate": 0.00018208996248097458,
|
| 1072 |
+
"loss": 3.5159,
|
| 1073 |
+
"step": 519
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.7,
|
| 1077 |
+
"learning_rate": 0.00018079929984946637,
|
| 1078 |
+
"loss": 3.489,
|
| 1079 |
+
"step": 522
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 3.72,
|
| 1083 |
+
"learning_rate": 0.0001795062521467242,
|
| 1084 |
+
"loss": 3.4791,
|
| 1085 |
+
"step": 525
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 3.74,
|
| 1089 |
+
"learning_rate": 0.00017821091950524768,
|
| 1090 |
+
"loss": 3.4478,
|
| 1091 |
+
"step": 528
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 3.76,
|
| 1095 |
+
"learning_rate": 0.00017691340223448016,
|
| 1096 |
+
"loss": 3.4824,
|
| 1097 |
+
"step": 531
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 3.78,
|
| 1101 |
+
"learning_rate": 0.00017561380081304058,
|
| 1102 |
+
"loss": 3.4876,
|
| 1103 |
+
"step": 534
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 3.8,
|
| 1107 |
+
"learning_rate": 0.00017431221588094307,
|
| 1108 |
+
"loss": 3.5219,
|
| 1109 |
+
"step": 537
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 3.83,
|
| 1113 |
+
"learning_rate": 0.00017300874823180282,
|
| 1114 |
+
"loss": 3.4727,
|
| 1115 |
+
"step": 540
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 3.85,
|
| 1119 |
+
"learning_rate": 0.00017170349880503107,
|
| 1120 |
+
"loss": 3.4474,
|
| 1121 |
+
"step": 543
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 3.87,
|
| 1125 |
+
"learning_rate": 0.00017039656867801845,
|
| 1126 |
+
"loss": 3.4553,
|
| 1127 |
+
"step": 546
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 3.89,
|
| 1131 |
+
"learning_rate": 0.00016908805905830752,
|
| 1132 |
+
"loss": 3.4837,
|
| 1133 |
+
"step": 549
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 3.91,
|
| 1137 |
+
"learning_rate": 0.00016777807127575542,
|
| 1138 |
+
"loss": 3.4804,
|
| 1139 |
+
"step": 552
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 3.93,
|
| 1143 |
+
"learning_rate": 0.00016646670677468686,
|
| 1144 |
+
"loss": 3.4665,
|
| 1145 |
+
"step": 555
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 3.95,
|
| 1149 |
+
"learning_rate": 0.00016515406710603867,
|
| 1150 |
+
"loss": 3.4832,
|
| 1151 |
+
"step": 558
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 3.97,
|
| 1155 |
+
"learning_rate": 0.0001638402539194953,
|
| 1156 |
+
"loss": 3.428,
|
| 1157 |
+
"step": 561
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 3.99,
|
| 1161 |
+
"learning_rate": 0.00016252536895561753,
|
| 1162 |
+
"loss": 3.4976,
|
| 1163 |
+
"step": 564
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 3.99,
|
| 1167 |
+
"eval_accuracy": 0.46655492116704206,
|
| 1168 |
+
"eval_loss": 3.1618497371673584,
|
| 1169 |
+
"eval_runtime": 7.1255,
|
| 1170 |
+
"eval_samples_per_second": 283.069,
|
| 1171 |
+
"eval_steps_per_second": 141.605,
|
| 1172 |
+
"step": 564
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 4.02,
|
| 1176 |
+
"learning_rate": 0.00016120951403796364,
|
| 1177 |
+
"loss": 4.4347,
|
| 1178 |
+
"step": 567
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 4.04,
|
| 1182 |
+
"learning_rate": 0.00015989279106520427,
|
| 1183 |
+
"loss": 3.4532,
|
| 1184 |
+
"step": 570
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 4.06,
|
| 1188 |
+
"learning_rate": 0.00015857530200323163,
|
| 1189 |
+
"loss": 3.4226,
|
| 1190 |
+
"step": 573
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 4.08,
|
| 1194 |
+
"learning_rate": 0.000157257148877263,
|
| 1195 |
+
"loss": 3.4455,
|
| 1196 |
+
"step": 576
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 4.11,
|
| 1200 |
+
"learning_rate": 0.00015593843376394043,
|
| 1201 |
+
"loss": 3.4435,
|
| 1202 |
+
"step": 579
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 4.13,
|
| 1206 |
+
"learning_rate": 0.00015461925878342556,
|
| 1207 |
+
"loss": 3.4103,
|
| 1208 |
+
"step": 582
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 4.15,
|
| 1212 |
+
"learning_rate": 0.00015329972609149191,
|
| 1213 |
+
"loss": 3.4701,
|
| 1214 |
+
"step": 585
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 4.17,
|
| 1218 |
+
"learning_rate": 0.00015197993787161384,
|
| 1219 |
+
"loss": 3.4195,
|
| 1220 |
+
"step": 588
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 4.19,
|
| 1224 |
+
"learning_rate": 0.00015065999632705353,
|
| 1225 |
+
"loss": 3.4592,
|
| 1226 |
+
"step": 591
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 4.21,
|
| 1230 |
+
"learning_rate": 0.00014934000367294647,
|
| 1231 |
+
"loss": 3.4675,
|
| 1232 |
+
"step": 594
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 4.23,
|
| 1236 |
+
"learning_rate": 0.00014802006212838616,
|
| 1237 |
+
"loss": 3.5138,
|
| 1238 |
+
"step": 597
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 4.25,
|
| 1242 |
+
"learning_rate": 0.00014670027390850809,
|
| 1243 |
+
"loss": 3.485,
|
| 1244 |
+
"step": 600
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 4.28,
|
| 1248 |
+
"learning_rate": 0.00014538074121657447,
|
| 1249 |
+
"loss": 3.4549,
|
| 1250 |
+
"step": 603
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 4.3,
|
| 1254 |
+
"learning_rate": 0.0001440615662360596,
|
| 1255 |
+
"loss": 3.4515,
|
| 1256 |
+
"step": 606
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 4.32,
|
| 1260 |
+
"learning_rate": 0.000142742851122737,
|
| 1261 |
+
"loss": 3.4109,
|
| 1262 |
+
"step": 609
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 4.34,
|
| 1266 |
+
"learning_rate": 0.00014142469799676831,
|
| 1267 |
+
"loss": 3.4207,
|
| 1268 |
+
"step": 612
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 4.36,
|
| 1272 |
+
"learning_rate": 0.00014010720893479567,
|
| 1273 |
+
"loss": 3.4871,
|
| 1274 |
+
"step": 615
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 4.38,
|
| 1278 |
+
"learning_rate": 0.00013879048596203636,
|
| 1279 |
+
"loss": 3.4328,
|
| 1280 |
+
"step": 618
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 4.4,
|
| 1284 |
+
"learning_rate": 0.00013747463104438247,
|
| 1285 |
+
"loss": 3.4716,
|
| 1286 |
+
"step": 621
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 4.42,
|
| 1290 |
+
"learning_rate": 0.0001361597460805047,
|
| 1291 |
+
"loss": 3.432,
|
| 1292 |
+
"step": 624
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 4.44,
|
| 1296 |
+
"learning_rate": 0.00013484593289396133,
|
| 1297 |
+
"loss": 3.4307,
|
| 1298 |
+
"step": 627
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 4.47,
|
| 1302 |
+
"learning_rate": 0.00013353329322531311,
|
| 1303 |
+
"loss": 3.5387,
|
| 1304 |
+
"step": 630
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 4.49,
|
| 1308 |
+
"learning_rate": 0.0001322219287242446,
|
| 1309 |
+
"loss": 3.466,
|
| 1310 |
+
"step": 633
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 4.51,
|
| 1314 |
+
"learning_rate": 0.0001309119409416925,
|
| 1315 |
+
"loss": 3.4677,
|
| 1316 |
+
"step": 636
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 4.53,
|
| 1320 |
+
"learning_rate": 0.00012960343132198158,
|
| 1321 |
+
"loss": 3.3951,
|
| 1322 |
+
"step": 639
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 4.55,
|
| 1326 |
+
"learning_rate": 0.00012829650119496896,
|
| 1327 |
+
"loss": 3.433,
|
| 1328 |
+
"step": 642
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 4.57,
|
| 1332 |
+
"learning_rate": 0.00012699125176819716,
|
| 1333 |
+
"loss": 3.4642,
|
| 1334 |
+
"step": 645
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 4.59,
|
| 1338 |
+
"learning_rate": 0.00012568778411905688,
|
| 1339 |
+
"loss": 3.429,
|
| 1340 |
+
"step": 648
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 4.61,
|
| 1344 |
+
"learning_rate": 0.00012438619918695934,
|
| 1345 |
+
"loss": 3.4187,
|
| 1346 |
+
"step": 651
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 4.63,
|
| 1350 |
+
"learning_rate": 0.00012308659776551984,
|
| 1351 |
+
"loss": 3.466,
|
| 1352 |
+
"step": 654
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 4.66,
|
| 1356 |
+
"learning_rate": 0.00012178908049475228,
|
| 1357 |
+
"loss": 3.4039,
|
| 1358 |
+
"step": 657
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 4.68,
|
| 1362 |
+
"learning_rate": 0.00012049374785327578,
|
| 1363 |
+
"loss": 3.4337,
|
| 1364 |
+
"step": 660
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 4.7,
|
| 1368 |
+
"learning_rate": 0.00011920070015053363,
|
| 1369 |
+
"loss": 3.4728,
|
| 1370 |
+
"step": 663
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 4.72,
|
| 1374 |
+
"learning_rate": 0.00011791003751902542,
|
| 1375 |
+
"loss": 3.5019,
|
| 1376 |
+
"step": 666
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 4.74,
|
| 1380 |
+
"learning_rate": 0.00011662185990655284,
|
| 1381 |
+
"loss": 3.3873,
|
| 1382 |
+
"step": 669
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 4.76,
|
| 1386 |
+
"learning_rate": 0.00011533626706848,
|
| 1387 |
+
"loss": 3.3441,
|
| 1388 |
+
"step": 672
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 4.78,
|
| 1392 |
+
"learning_rate": 0.00011405335856000816,
|
| 1393 |
+
"loss": 3.4262,
|
| 1394 |
+
"step": 675
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 4.8,
|
| 1398 |
+
"learning_rate": 0.00011277323372846643,
|
| 1399 |
+
"loss": 3.4619,
|
| 1400 |
+
"step": 678
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 4.83,
|
| 1404 |
+
"learning_rate": 0.00011149599170561842,
|
| 1405 |
+
"loss": 3.4034,
|
| 1406 |
+
"step": 681
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 4.85,
|
| 1410 |
+
"learning_rate": 0.00011022173139998555,
|
| 1411 |
+
"loss": 3.5172,
|
| 1412 |
+
"step": 684
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 4.87,
|
| 1416 |
+
"learning_rate": 0.00010895055148918756,
|
| 1417 |
+
"loss": 3.4081,
|
| 1418 |
+
"step": 687
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 4.89,
|
| 1422 |
+
"learning_rate": 0.00010768255041230107,
|
| 1423 |
+
"loss": 3.4555,
|
| 1424 |
+
"step": 690
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 4.91,
|
| 1428 |
+
"learning_rate": 0.00010641782636223658,
|
| 1429 |
+
"loss": 3.4756,
|
| 1430 |
+
"step": 693
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 4.93,
|
| 1434 |
+
"learning_rate": 0.00010515647727813445,
|
| 1435 |
+
"loss": 3.3765,
|
| 1436 |
+
"step": 696
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 4.95,
|
| 1440 |
+
"learning_rate": 0.00010389860083778056,
|
| 1441 |
+
"loss": 3.4485,
|
| 1442 |
+
"step": 699
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 4.97,
|
| 1446 |
+
"learning_rate": 0.00010264429445004229,
|
| 1447 |
+
"loss": 3.4879,
|
| 1448 |
+
"step": 702
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 4.99,
|
| 1452 |
+
"learning_rate": 0.00010139365524732513,
|
| 1453 |
+
"loss": 3.4356,
|
| 1454 |
+
"step": 705
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 4.99,
|
| 1458 |
+
"eval_accuracy": 0.4739487637146307,
|
| 1459 |
+
"eval_loss": 3.100182056427002,
|
| 1460 |
+
"eval_runtime": 7.093,
|
| 1461 |
+
"eval_samples_per_second": 284.366,
|
| 1462 |
+
"eval_steps_per_second": 142.254,
|
| 1463 |
+
"step": 705
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 5.02,
|
| 1467 |
+
"learning_rate": 0.00010014678007805106,
|
| 1468 |
+
"loss": 4.224,
|
| 1469 |
+
"step": 708
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 5.04,
|
| 1473 |
+
"learning_rate": 9.890376549915847e-05,
|
| 1474 |
+
"loss": 3.4427,
|
| 1475 |
+
"step": 711
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 5.06,
|
| 1479 |
+
"learning_rate": 9.766470776862499e-05,
|
| 1480 |
+
"loss": 3.4274,
|
| 1481 |
+
"step": 714
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 5.08,
|
| 1485 |
+
"learning_rate": 9.642970283801327e-05,
|
| 1486 |
+
"loss": 3.4132,
|
| 1487 |
+
"step": 717
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 5.11,
|
| 1491 |
+
"learning_rate": 9.519884634504074e-05,
|
| 1492 |
+
"loss": 3.3957,
|
| 1493 |
+
"step": 720
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 5.13,
|
| 1497 |
+
"learning_rate": 9.397223360617325e-05,
|
| 1498 |
+
"loss": 3.4081,
|
| 1499 |
+
"step": 723
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 5.15,
|
| 1503 |
+
"learning_rate": 9.274995960924409e-05,
|
| 1504 |
+
"loss": 3.4419,
|
| 1505 |
+
"step": 726
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 5.17,
|
| 1509 |
+
"learning_rate": 9.15321190060981e-05,
|
| 1510 |
+
"loss": 3.4608,
|
| 1511 |
+
"step": 729
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 5.19,
|
| 1515 |
+
"learning_rate": 9.031880610526189e-05,
|
| 1516 |
+
"loss": 3.401,
|
| 1517 |
+
"step": 732
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 5.21,
|
| 1521 |
+
"learning_rate": 8.91101148646407e-05,
|
| 1522 |
+
"loss": 3.3991,
|
| 1523 |
+
"step": 735
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 5.23,
|
| 1527 |
+
"learning_rate": 8.790613888424248e-05,
|
| 1528 |
+
"loss": 3.4329,
|
| 1529 |
+
"step": 738
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 5.25,
|
| 1533 |
+
"learning_rate": 8.670697139892933e-05,
|
| 1534 |
+
"loss": 3.4085,
|
| 1535 |
+
"step": 741
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 5.28,
|
| 1539 |
+
"learning_rate": 8.551270527119782e-05,
|
| 1540 |
+
"loss": 3.4269,
|
| 1541 |
+
"step": 744
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 5.3,
|
| 1545 |
+
"learning_rate": 8.432343298398748e-05,
|
| 1546 |
+
"loss": 3.3984,
|
| 1547 |
+
"step": 747
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 5.32,
|
| 1551 |
+
"learning_rate": 8.313924663351926e-05,
|
| 1552 |
+
"loss": 3.3986,
|
| 1553 |
+
"step": 750
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 5.34,
|
| 1557 |
+
"learning_rate": 8.196023792216334e-05,
|
| 1558 |
+
"loss": 3.4602,
|
| 1559 |
+
"step": 753
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 5.36,
|
| 1563 |
+
"learning_rate": 8.078649815133826e-05,
|
| 1564 |
+
"loss": 3.4649,
|
| 1565 |
+
"step": 756
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 5.38,
|
| 1569 |
+
"learning_rate": 7.961811821444008e-05,
|
| 1570 |
+
"loss": 3.3909,
|
| 1571 |
+
"step": 759
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 5.4,
|
| 1575 |
+
"learning_rate": 7.845518858980415e-05,
|
| 1576 |
+
"loss": 3.3896,
|
| 1577 |
+
"step": 762
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 5.42,
|
| 1581 |
+
"learning_rate": 7.729779933369818e-05,
|
| 1582 |
+
"loss": 3.3746,
|
| 1583 |
+
"step": 765
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 5.44,
|
| 1587 |
+
"learning_rate": 7.614604007334866e-05,
|
| 1588 |
+
"loss": 3.4258,
|
| 1589 |
+
"step": 768
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 5.47,
|
| 1593 |
+
"learning_rate": 7.500000000000002e-05,
|
| 1594 |
+
"loss": 3.3834,
|
| 1595 |
+
"step": 771
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 5.49,
|
| 1599 |
+
"learning_rate": 7.385976786200765e-05,
|
| 1600 |
+
"loss": 3.3626,
|
| 1601 |
+
"step": 774
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 5.51,
|
| 1605 |
+
"learning_rate": 7.272543195796558e-05,
|
| 1606 |
+
"loss": 3.4169,
|
| 1607 |
+
"step": 777
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 5.53,
|
| 1611 |
+
"learning_rate": 7.15970801298685e-05,
|
| 1612 |
+
"loss": 3.4288,
|
| 1613 |
+
"step": 780
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 5.55,
|
| 1617 |
+
"learning_rate": 7.047479975630955e-05,
|
| 1618 |
+
"loss": 3.3825,
|
| 1619 |
+
"step": 783
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 5.57,
|
| 1623 |
+
"learning_rate": 6.935867774571337e-05,
|
| 1624 |
+
"loss": 3.3973,
|
| 1625 |
+
"step": 786
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 5.59,
|
| 1629 |
+
"learning_rate": 6.82488005296066e-05,
|
| 1630 |
+
"loss": 3.4212,
|
| 1631 |
+
"step": 789
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 5.61,
|
| 1635 |
+
"learning_rate": 6.714525405592412e-05,
|
| 1636 |
+
"loss": 3.4038,
|
| 1637 |
+
"step": 792
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 5.63,
|
| 1641 |
+
"learning_rate": 6.604812378235381e-05,
|
| 1642 |
+
"loss": 3.404,
|
| 1643 |
+
"step": 795
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 5.66,
|
| 1647 |
+
"learning_rate": 6.495749466971827e-05,
|
| 1648 |
+
"loss": 3.4048,
|
| 1649 |
+
"step": 798
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 5.68,
|
| 1653 |
+
"learning_rate": 6.38734511753961e-05,
|
| 1654 |
+
"loss": 3.4066,
|
| 1655 |
+
"step": 801
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 5.7,
|
| 1659 |
+
"learning_rate": 6.279607724678103e-05,
|
| 1660 |
+
"loss": 3.3839,
|
| 1661 |
+
"step": 804
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 5.72,
|
| 1665 |
+
"learning_rate": 6.172545631478156e-05,
|
| 1666 |
+
"loss": 3.4525,
|
| 1667 |
+
"step": 807
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 5.74,
|
| 1671 |
+
"learning_rate": 6.066167128735988e-05,
|
| 1672 |
+
"loss": 3.3788,
|
| 1673 |
+
"step": 810
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 5.76,
|
| 1677 |
+
"learning_rate": 5.960480454311155e-05,
|
| 1678 |
+
"loss": 3.3995,
|
| 1679 |
+
"step": 813
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 5.78,
|
| 1683 |
+
"learning_rate": 5.855493792488637e-05,
|
| 1684 |
+
"loss": 3.4489,
|
| 1685 |
+
"step": 816
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 5.8,
|
| 1689 |
+
"learning_rate": 5.751215273345036e-05,
|
| 1690 |
+
"loss": 3.3495,
|
| 1691 |
+
"step": 819
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 5.83,
|
| 1695 |
+
"learning_rate": 5.6476529721189974e-05,
|
| 1696 |
+
"loss": 3.4403,
|
| 1697 |
+
"step": 822
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 5.85,
|
| 1701 |
+
"learning_rate": 5.5448149085858596e-05,
|
| 1702 |
+
"loss": 3.3715,
|
| 1703 |
+
"step": 825
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 5.87,
|
| 1707 |
+
"learning_rate": 5.4427090464366323e-05,
|
| 1708 |
+
"loss": 3.3936,
|
| 1709 |
+
"step": 828
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 5.89,
|
| 1713 |
+
"learning_rate": 5.3413432926612655e-05,
|
| 1714 |
+
"loss": 3.3675,
|
| 1715 |
+
"step": 831
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 5.91,
|
| 1719 |
+
"learning_rate": 5.240725496936372e-05,
|
| 1720 |
+
"loss": 3.405,
|
| 1721 |
+
"step": 834
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 5.93,
|
| 1725 |
+
"learning_rate": 5.140863451017318e-05,
|
| 1726 |
+
"loss": 3.4457,
|
| 1727 |
+
"step": 837
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 5.95,
|
| 1731 |
+
"learning_rate": 5.0417648881348774e-05,
|
| 1732 |
+
"loss": 3.4159,
|
| 1733 |
+
"step": 840
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 5.97,
|
| 1737 |
+
"learning_rate": 4.943437482396336e-05,
|
| 1738 |
+
"loss": 3.3632,
|
| 1739 |
+
"step": 843
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 5.99,
|
| 1743 |
+
"learning_rate": 4.845888848191248e-05,
|
| 1744 |
+
"loss": 3.4493,
|
| 1745 |
+
"step": 846
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 5.99,
|
| 1749 |
+
"eval_accuracy": 0.4745788157886148,
|
| 1750 |
+
"eval_loss": 3.1028223037719727,
|
| 1751 |
+
"eval_runtime": 7.1484,
|
| 1752 |
+
"eval_samples_per_second": 282.16,
|
| 1753 |
+
"eval_steps_per_second": 141.15,
|
| 1754 |
+
"step": 846
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 6.02,
|
| 1758 |
+
"learning_rate": 4.74912653960177e-05,
|
| 1759 |
+
"loss": 4.272,
|
| 1760 |
+
"step": 849
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 6.04,
|
| 1764 |
+
"learning_rate": 4.65315804981768e-05,
|
| 1765 |
+
"loss": 3.4332,
|
| 1766 |
+
"step": 852
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 6.06,
|
| 1770 |
+
"learning_rate": 4.5579908105561016e-05,
|
| 1771 |
+
"loss": 3.3962,
|
| 1772 |
+
"step": 855
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 6.08,
|
| 1776 |
+
"learning_rate": 4.463632191486022e-05,
|
| 1777 |
+
"loss": 3.3835,
|
| 1778 |
+
"step": 858
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 6.11,
|
| 1782 |
+
"learning_rate": 4.3700894996575764e-05,
|
| 1783 |
+
"loss": 3.3852,
|
| 1784 |
+
"step": 861
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 6.13,
|
| 1788 |
+
"learning_rate": 4.277369978936187e-05,
|
| 1789 |
+
"loss": 3.4379,
|
| 1790 |
+
"step": 864
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 6.15,
|
| 1794 |
+
"learning_rate": 4.185480809441631e-05,
|
| 1795 |
+
"loss": 3.4083,
|
| 1796 |
+
"step": 867
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 6.17,
|
| 1800 |
+
"learning_rate": 4.094429106991981e-05,
|
| 1801 |
+
"loss": 3.447,
|
| 1802 |
+
"step": 870
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 6.19,
|
| 1806 |
+
"learning_rate": 4.004221922552608e-05,
|
| 1807 |
+
"loss": 3.3643,
|
| 1808 |
+
"step": 873
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 6.21,
|
| 1812 |
+
"learning_rate": 3.914866241690115e-05,
|
| 1813 |
+
"loss": 3.4416,
|
| 1814 |
+
"step": 876
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 6.23,
|
| 1818 |
+
"learning_rate": 3.826368984031414e-05,
|
| 1819 |
+
"loss": 3.4144,
|
| 1820 |
+
"step": 879
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 6.25,
|
| 1824 |
+
"learning_rate": 3.738737002727863e-05,
|
| 1825 |
+
"loss": 3.3951,
|
| 1826 |
+
"step": 882
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 6.28,
|
| 1830 |
+
"learning_rate": 3.651977083924563e-05,
|
| 1831 |
+
"loss": 3.39,
|
| 1832 |
+
"step": 885
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 6.3,
|
| 1836 |
+
"learning_rate": 3.566095946234842e-05,
|
| 1837 |
+
"loss": 3.4053,
|
| 1838 |
+
"step": 888
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 6.32,
|
| 1842 |
+
"learning_rate": 3.4811002402199874e-05,
|
| 1843 |
+
"loss": 3.4033,
|
| 1844 |
+
"step": 891
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 6.34,
|
| 1848 |
+
"learning_rate": 3.396996547874203e-05,
|
| 1849 |
+
"loss": 3.2884,
|
| 1850 |
+
"step": 894
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 6.36,
|
| 1854 |
+
"learning_rate": 3.3137913821149425e-05,
|
| 1855 |
+
"loss": 3.415,
|
| 1856 |
+
"step": 897
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 6.38,
|
| 1860 |
+
"learning_rate": 3.2314911862785275e-05,
|
| 1861 |
+
"loss": 3.3832,
|
| 1862 |
+
"step": 900
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 6.4,
|
| 1866 |
+
"learning_rate": 3.150102333621181e-05,
|
| 1867 |
+
"loss": 3.3488,
|
| 1868 |
+
"step": 903
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 6.42,
|
| 1872 |
+
"learning_rate": 3.0696311268255093e-05,
|
| 1873 |
+
"loss": 3.3449,
|
| 1874 |
+
"step": 906
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 6.44,
|
| 1878 |
+
"learning_rate": 2.990083797512401e-05,
|
| 1879 |
+
"loss": 3.3563,
|
| 1880 |
+
"step": 909
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 6.47,
|
| 1884 |
+
"learning_rate": 2.9114665057584768e-05,
|
| 1885 |
+
"loss": 3.4071,
|
| 1886 |
+
"step": 912
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 6.49,
|
| 1890 |
+
"learning_rate": 2.8337853396190503e-05,
|
| 1891 |
+
"loss": 3.456,
|
| 1892 |
+
"step": 915
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 6.51,
|
| 1896 |
+
"learning_rate": 2.7570463146566758e-05,
|
| 1897 |
+
"loss": 3.4042,
|
| 1898 |
+
"step": 918
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 6.53,
|
| 1902 |
+
"learning_rate": 2.6812553734752956e-05,
|
| 1903 |
+
"loss": 3.375,
|
| 1904 |
+
"step": 921
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 6.55,
|
| 1908 |
+
"learning_rate": 2.6064183852600797e-05,
|
| 1909 |
+
"loss": 3.3807,
|
| 1910 |
+
"step": 924
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 6.57,
|
| 1914 |
+
"learning_rate": 2.532541145322881e-05,
|
| 1915 |
+
"loss": 3.3436,
|
| 1916 |
+
"step": 927
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 6.59,
|
| 1920 |
+
"learning_rate": 2.4596293746534905e-05,
|
| 1921 |
+
"loss": 3.2923,
|
| 1922 |
+
"step": 930
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 6.61,
|
| 1926 |
+
"learning_rate": 2.3876887194765874e-05,
|
| 1927 |
+
"loss": 3.4021,
|
| 1928 |
+
"step": 933
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 6.63,
|
| 1932 |
+
"learning_rate": 2.3167247508144954e-05,
|
| 1933 |
+
"loss": 3.3814,
|
| 1934 |
+
"step": 936
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 6.66,
|
| 1938 |
+
"learning_rate": 2.24674296405579e-05,
|
| 1939 |
+
"loss": 3.4269,
|
| 1940 |
+
"step": 939
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 6.68,
|
| 1944 |
+
"learning_rate": 2.1777487785297114e-05,
|
| 1945 |
+
"loss": 3.3474,
|
| 1946 |
+
"step": 942
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 6.7,
|
| 1950 |
+
"learning_rate": 2.1097475370865242e-05,
|
| 1951 |
+
"loss": 3.4056,
|
| 1952 |
+
"step": 945
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 6.72,
|
| 1956 |
+
"learning_rate": 2.0427445056837544e-05,
|
| 1957 |
+
"loss": 3.4254,
|
| 1958 |
+
"step": 948
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 6.74,
|
| 1962 |
+
"learning_rate": 1.9767448729784046e-05,
|
| 1963 |
+
"loss": 3.3823,
|
| 1964 |
+
"step": 951
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 6.76,
|
| 1968 |
+
"learning_rate": 1.9117537499251413e-05,
|
| 1969 |
+
"loss": 3.3903,
|
| 1970 |
+
"step": 954
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 6.78,
|
| 1974 |
+
"learning_rate": 1.8477761693805203e-05,
|
| 1975 |
+
"loss": 3.3717,
|
| 1976 |
+
"step": 957
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 6.8,
|
| 1980 |
+
"learning_rate": 1.7848170857132325e-05,
|
| 1981 |
+
"loss": 3.3791,
|
| 1982 |
+
"step": 960
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 6.83,
|
| 1986 |
+
"learning_rate": 1.7228813744204556e-05,
|
| 1987 |
+
"loss": 3.3805,
|
| 1988 |
+
"step": 963
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 6.85,
|
| 1992 |
+
"learning_rate": 1.661973831750283e-05,
|
| 1993 |
+
"loss": 3.3905,
|
| 1994 |
+
"step": 966
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 6.87,
|
| 1998 |
+
"learning_rate": 1.6020991743303263e-05,
|
| 1999 |
+
"loss": 3.4504,
|
| 2000 |
+
"step": 969
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 6.89,
|
| 2004 |
+
"learning_rate": 1.5432620388024512e-05,
|
| 2005 |
+
"loss": 3.3748,
|
| 2006 |
+
"step": 972
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 6.91,
|
| 2010 |
+
"learning_rate": 1.4854669814637143e-05,
|
| 2011 |
+
"loss": 3.4381,
|
| 2012 |
+
"step": 975
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 6.93,
|
| 2016 |
+
"learning_rate": 1.428718477913543e-05,
|
| 2017 |
+
"loss": 3.3435,
|
| 2018 |
+
"step": 978
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 6.95,
|
| 2022 |
+
"learning_rate": 1.3730209227071436e-05,
|
| 2023 |
+
"loss": 3.3974,
|
| 2024 |
+
"step": 981
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 6.97,
|
| 2028 |
+
"learning_rate": 1.3183786290151838e-05,
|
| 2029 |
+
"loss": 3.3399,
|
| 2030 |
+
"step": 984
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 6.99,
|
| 2034 |
+
"learning_rate": 1.264795828289787e-05,
|
| 2035 |
+
"loss": 3.4199,
|
| 2036 |
+
"step": 987
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 6.99,
|
| 2040 |
+
"eval_accuracy": 0.47660093674971465,
|
| 2041 |
+
"eval_loss": 3.0857138633728027,
|
| 2042 |
+
"eval_runtime": 7.115,
|
| 2043 |
+
"eval_samples_per_second": 283.486,
|
| 2044 |
+
"eval_steps_per_second": 141.813,
|
| 2045 |
+
"step": 987
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 7.02,
|
| 2049 |
+
"learning_rate": 1.2122766699368603e-05,
|
| 2050 |
+
"loss": 4.2135,
|
| 2051 |
+
"step": 990
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 7.04,
|
| 2055 |
+
"learning_rate": 1.1608252209947566e-05,
|
| 2056 |
+
"loss": 3.4164,
|
| 2057 |
+
"step": 993
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 7.06,
|
| 2061 |
+
"learning_rate": 1.1104454658193395e-05,
|
| 2062 |
+
"loss": 3.4037,
|
| 2063 |
+
"step": 996
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 7.08,
|
| 2067 |
+
"learning_rate": 1.061141305775422e-05,
|
| 2068 |
+
"loss": 3.3729,
|
| 2069 |
+
"step": 999
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 7.11,
|
| 2073 |
+
"learning_rate": 1.0129165589346643e-05,
|
| 2074 |
+
"loss": 3.4295,
|
| 2075 |
+
"step": 1002
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 7.13,
|
| 2079 |
+
"learning_rate": 9.657749597798947e-06,
|
| 2080 |
+
"loss": 3.4044,
|
| 2081 |
+
"step": 1005
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 7.15,
|
| 2085 |
+
"learning_rate": 9.197201589159215e-06,
|
| 2086 |
+
"loss": 3.4028,
|
| 2087 |
+
"step": 1008
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 7.17,
|
| 2091 |
+
"learning_rate": 8.747557227868263e-06,
|
| 2092 |
+
"loss": 3.3725,
|
| 2093 |
+
"step": 1011
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 7.19,
|
| 2097 |
+
"learning_rate": 8.308851333997917e-06,
|
| 2098 |
+
"loss": 3.33,
|
| 2099 |
+
"step": 1014
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 7.21,
|
| 2103 |
+
"learning_rate": 7.881117880554421e-06,
|
| 2104 |
+
"loss": 3.4139,
|
| 2105 |
+
"step": 1017
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 7.23,
|
| 2109 |
+
"learning_rate": 7.4643899908477525e-06,
|
| 2110 |
+
"loss": 3.4329,
|
| 2111 |
+
"step": 1020
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 7.25,
|
| 2115 |
+
"learning_rate": 7.058699935926526e-06,
|
| 2116 |
+
"loss": 3.3783,
|
| 2117 |
+
"step": 1023
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 7.28,
|
| 2121 |
+
"learning_rate": 6.664079132078881e-06,
|
| 2122 |
+
"loss": 3.3888,
|
| 2123 |
+
"step": 1026
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 7.3,
|
| 2127 |
+
"learning_rate": 6.280558138399805e-06,
|
| 2128 |
+
"loss": 3.3738,
|
| 2129 |
+
"step": 1029
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 7.32,
|
| 2133 |
+
"learning_rate": 5.908166654424412e-06,
|
| 2134 |
+
"loss": 3.3097,
|
| 2135 |
+
"step": 1032
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 7.34,
|
| 2139 |
+
"learning_rate": 5.546933517828317e-06,
|
| 2140 |
+
"loss": 3.4202,
|
| 2141 |
+
"step": 1035
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 7.36,
|
| 2145 |
+
"learning_rate": 5.196886702194203e-06,
|
| 2146 |
+
"loss": 3.3502,
|
| 2147 |
+
"step": 1038
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 7.38,
|
| 2151 |
+
"learning_rate": 4.8580533148458225e-06,
|
| 2152 |
+
"loss": 3.3874,
|
| 2153 |
+
"step": 1041
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 7.4,
|
| 2157 |
+
"learning_rate": 4.530459594748592e-06,
|
| 2158 |
+
"loss": 3.3369,
|
| 2159 |
+
"step": 1044
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 7.42,
|
| 2163 |
+
"learning_rate": 4.214130910477853e-06,
|
| 2164 |
+
"loss": 3.3929,
|
| 2165 |
+
"step": 1047
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 7.44,
|
| 2169 |
+
"learning_rate": 3.90909175825429e-06,
|
| 2170 |
+
"loss": 3.362,
|
| 2171 |
+
"step": 1050
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 7.47,
|
| 2175 |
+
"learning_rate": 3.6153657600469176e-06,
|
| 2176 |
+
"loss": 3.3281,
|
| 2177 |
+
"step": 1053
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 7.49,
|
| 2181 |
+
"learning_rate": 3.3329756617438917e-06,
|
| 2182 |
+
"loss": 3.352,
|
| 2183 |
+
"step": 1056
|
| 2184 |
+
},
|
| 2185 |
+
{
|
| 2186 |
+
"epoch": 7.51,
|
| 2187 |
+
"learning_rate": 3.06194333139097e-06,
|
| 2188 |
+
"loss": 3.3818,
|
| 2189 |
+
"step": 1059
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 7.53,
|
| 2193 |
+
"learning_rate": 2.8022897574982826e-06,
|
| 2194 |
+
"loss": 3.4417,
|
| 2195 |
+
"step": 1062
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 7.55,
|
| 2199 |
+
"learning_rate": 2.554035047414732e-06,
|
| 2200 |
+
"loss": 3.3567,
|
| 2201 |
+
"step": 1065
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 7.57,
|
| 2205 |
+
"learning_rate": 2.3171984257711385e-06,
|
| 2206 |
+
"loss": 3.4005,
|
| 2207 |
+
"step": 1068
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 7.59,
|
| 2211 |
+
"learning_rate": 2.0917982329912987e-06,
|
| 2212 |
+
"loss": 3.3674,
|
| 2213 |
+
"step": 1071
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 7.61,
|
| 2217 |
+
"learning_rate": 1.8778519238719203e-06,
|
| 2218 |
+
"loss": 3.3507,
|
| 2219 |
+
"step": 1074
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 7.63,
|
| 2223 |
+
"learning_rate": 1.6753760662307215e-06,
|
| 2224 |
+
"loss": 3.3754,
|
| 2225 |
+
"step": 1077
|
| 2226 |
+
},
|
| 2227 |
+
{
|
| 2228 |
+
"epoch": 7.66,
|
| 2229 |
+
"learning_rate": 1.4843863396236388e-06,
|
| 2230 |
+
"loss": 3.4043,
|
| 2231 |
+
"step": 1080
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 7.68,
|
| 2235 |
+
"learning_rate": 1.3048975341304835e-06,
|
| 2236 |
+
"loss": 3.4084,
|
| 2237 |
+
"step": 1083
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 7.7,
|
| 2241 |
+
"learning_rate": 1.1369235492096397e-06,
|
| 2242 |
+
"loss": 3.4293,
|
| 2243 |
+
"step": 1086
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 7.72,
|
| 2247 |
+
"learning_rate": 9.804773926217092e-07,
|
| 2248 |
+
"loss": 3.4001,
|
| 2249 |
+
"step": 1089
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 7.74,
|
| 2253 |
+
"learning_rate": 8.355711794222342e-07,
|
| 2254 |
+
"loss": 3.3856,
|
| 2255 |
+
"step": 1092
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 7.76,
|
| 2259 |
+
"learning_rate": 7.022161310234642e-07,
|
| 2260 |
+
"loss": 3.3659,
|
| 2261 |
+
"step": 1095
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 7.78,
|
| 2265 |
+
"learning_rate": 5.80422574325401e-07,
|
| 2266 |
+
"loss": 3.3653,
|
| 2267 |
+
"step": 1098
|
| 2268 |
+
},
|
| 2269 |
+
{
|
| 2270 |
+
"epoch": 7.8,
|
| 2271 |
+
"learning_rate": 4.70199940916105e-07,
|
| 2272 |
+
"loss": 3.437,
|
| 2273 |
+
"step": 1101
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 7.83,
|
| 2277 |
+
"learning_rate": 3.715567663412966e-07,
|
| 2278 |
+
"loss": 3.382,
|
| 2279 |
+
"step": 1104
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 7.85,
|
| 2283 |
+
"learning_rate": 2.845006894433843e-07,
|
| 2284 |
+
"loss": 3.3835,
|
| 2285 |
+
"step": 1107
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 7.87,
|
| 2289 |
+
"learning_rate": 2.0903845176995503e-07,
|
| 2290 |
+
"loss": 3.3901,
|
| 2291 |
+
"step": 1110
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 7.89,
|
| 2295 |
+
"learning_rate": 1.4517589705164145e-07,
|
| 2296 |
+
"loss": 3.3445,
|
| 2297 |
+
"step": 1113
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 7.91,
|
| 2301 |
+
"learning_rate": 9.291797074961748e-08,
|
| 2302 |
+
"loss": 3.4404,
|
| 2303 |
+
"step": 1116
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 7.93,
|
| 2307 |
+
"learning_rate": 5.226871967267121e-08,
|
| 2308 |
+
"loss": 3.404,
|
| 2309 |
+
"step": 1119
|
| 2310 |
+
},
|
| 2311 |
+
{
|
| 2312 |
+
"epoch": 7.95,
|
| 2313 |
+
"learning_rate": 2.3231291663772245e-08,
|
| 2314 |
+
"loss": 3.3689,
|
| 2315 |
+
"step": 1122
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 7.97,
|
| 2319 |
+
"learning_rate": 5.80793535631674e-09,
|
| 2320 |
+
"loss": 3.3911,
|
| 2321 |
+
"step": 1125
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 7.99,
|
| 2325 |
+
"learning_rate": 0.0,
|
| 2326 |
+
"loss": 3.4086,
|
| 2327 |
+
"step": 1128
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 7.99,
|
| 2331 |
+
"eval_accuracy": 0.47283051057151976,
|
| 2332 |
+
"eval_loss": 3.0980591773986816,
|
| 2333 |
+
"eval_runtime": 7.1137,
|
| 2334 |
+
"eval_samples_per_second": 283.538,
|
| 2335 |
+
"eval_steps_per_second": 141.839,
|
| 2336 |
+
"step": 1128
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 7.99,
|
| 2340 |
+
"step": 1128,
|
| 2341 |
+
"total_flos": 197844098719744.0,
|
| 2342 |
+
"train_loss": 3.6050218030916037,
|
| 2343 |
+
"train_runtime": 557.9113,
|
| 2344 |
+
"train_samples_per_second": 260.099,
|
| 2345 |
+
"train_steps_per_second": 2.022
|
| 2346 |
+
}
|
| 2347 |
+
],
|
| 2348 |
+
"max_steps": 1128,
|
| 2349 |
+
"num_train_epochs": 8,
|
| 2350 |
+
"total_flos": 197844098719744.0,
|
| 2351 |
+
"trial_name": null,
|
| 2352 |
+
"trial_params": null
|
| 2353 |
+
}
|