End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +1480 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.99,
|
| 3 |
+
"eval_accuracy": 0.6431874521705574,
|
| 4 |
+
"eval_loss": 1.926223635673523,
|
| 5 |
+
"eval_runtime": 10.14,
|
| 6 |
+
"eval_samples": 2017,
|
| 7 |
+
"eval_samples_per_second": 198.916,
|
| 8 |
+
"eval_steps_per_second": 99.507,
|
| 9 |
+
"perplexity": 6.863542004501074,
|
| 10 |
+
"train_loss": 2.1596999946215476,
|
| 11 |
+
"train_runtime": 1037.2411,
|
| 12 |
+
"train_samples": 18139,
|
| 13 |
+
"train_samples_per_second": 87.439,
|
| 14 |
+
"train_steps_per_second": 0.68
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.99,
|
| 3 |
+
"eval_accuracy": 0.6431874521705574,
|
| 4 |
+
"eval_loss": 1.926223635673523,
|
| 5 |
+
"eval_runtime": 10.14,
|
| 6 |
+
"eval_samples": 2017,
|
| 7 |
+
"eval_samples_per_second": 198.916,
|
| 8 |
+
"eval_steps_per_second": 99.507,
|
| 9 |
+
"perplexity": 6.863542004501074
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.99,
|
| 3 |
+
"train_loss": 2.1596999946215476,
|
| 4 |
+
"train_runtime": 1037.2411,
|
| 5 |
+
"train_samples": 18139,
|
| 6 |
+
"train_samples_per_second": 87.439,
|
| 7 |
+
"train_steps_per_second": 0.68
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1480 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.6451979106514778,
|
| 3 |
+
"best_model_checkpoint": "./runtime-masked/bert_uncased_L-2_H-768_A-12-mlm-multi-emails-hq/checkpoint-705",
|
| 4 |
+
"epoch": 4.994708994708994,
|
| 5 |
+
"global_step": 705,
|
| 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": 2.4999999999999998e-05,
|
| 13 |
+
"loss": 3.2694,
|
| 14 |
+
"step": 3
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.04,
|
| 18 |
+
"learning_rate": 4.9999999999999996e-05,
|
| 19 |
+
"loss": 3.2366,
|
| 20 |
+
"step": 6
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.06,
|
| 24 |
+
"learning_rate": 7.5e-05,
|
| 25 |
+
"loss": 3.2013,
|
| 26 |
+
"step": 9
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.08,
|
| 30 |
+
"learning_rate": 9.999999999999999e-05,
|
| 31 |
+
"loss": 3.0257,
|
| 32 |
+
"step": 12
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.11,
|
| 36 |
+
"learning_rate": 0.000125,
|
| 37 |
+
"loss": 3.0287,
|
| 38 |
+
"step": 15
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.13,
|
| 42 |
+
"learning_rate": 0.00015,
|
| 43 |
+
"loss": 2.9585,
|
| 44 |
+
"step": 18
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.15,
|
| 48 |
+
"learning_rate": 0.000175,
|
| 49 |
+
"loss": 2.8967,
|
| 50 |
+
"step": 21
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.17,
|
| 54 |
+
"learning_rate": 0.00019999999999999998,
|
| 55 |
+
"loss": 2.7588,
|
| 56 |
+
"step": 24
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.19,
|
| 60 |
+
"learning_rate": 0.000225,
|
| 61 |
+
"loss": 2.7953,
|
| 62 |
+
"step": 27
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.21,
|
| 66 |
+
"learning_rate": 0.00025,
|
| 67 |
+
"loss": 2.752,
|
| 68 |
+
"step": 30
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.23,
|
| 72 |
+
"learning_rate": 0.00027499999999999996,
|
| 73 |
+
"loss": 2.6644,
|
| 74 |
+
"step": 33
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.25,
|
| 78 |
+
"learning_rate": 0.0003,
|
| 79 |
+
"loss": 2.736,
|
| 80 |
+
"step": 36
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.28,
|
| 84 |
+
"learning_rate": 0.00029998511516242686,
|
| 85 |
+
"loss": 2.5937,
|
| 86 |
+
"step": 39
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3,
|
| 90 |
+
"learning_rate": 0.00029994046360381946,
|
| 91 |
+
"loss": 2.6411,
|
| 92 |
+
"step": 42
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.32,
|
| 96 |
+
"learning_rate": 0.0002998660541859271,
|
| 97 |
+
"loss": 2.587,
|
| 98 |
+
"step": 45
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.34,
|
| 102 |
+
"learning_rate": 0.0002997619016763776,
|
| 103 |
+
"loss": 2.5726,
|
| 104 |
+
"step": 48
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.36,
|
| 108 |
+
"learning_rate": 0.000299628026745747,
|
| 109 |
+
"loss": 2.5472,
|
| 110 |
+
"step": 51
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.38,
|
| 114 |
+
"learning_rate": 0.0002994644559634565,
|
| 115 |
+
"loss": 2.5179,
|
| 116 |
+
"step": 54
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.4,
|
| 120 |
+
"learning_rate": 0.00029927122179249977,
|
| 121 |
+
"loss": 2.4876,
|
| 122 |
+
"step": 57
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.42,
|
| 126 |
+
"learning_rate": 0.00029904836258300014,
|
| 127 |
+
"loss": 2.5026,
|
| 128 |
+
"step": 60
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.44,
|
| 132 |
+
"learning_rate": 0.0002987959225645994,
|
| 133 |
+
"loss": 2.4842,
|
| 134 |
+
"step": 63
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.47,
|
| 138 |
+
"learning_rate": 0.00029851395183767983,
|
| 139 |
+
"loss": 2.442,
|
| 140 |
+
"step": 66
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.49,
|
| 144 |
+
"learning_rate": 0.00029820250636342104,
|
| 145 |
+
"loss": 2.4273,
|
| 146 |
+
"step": 69
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.51,
|
| 150 |
+
"learning_rate": 0.00029786164795269374,
|
| 151 |
+
"loss": 2.4646,
|
| 152 |
+
"step": 72
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.53,
|
| 156 |
+
"learning_rate": 0.00029749144425379216,
|
| 157 |
+
"loss": 2.3918,
|
| 158 |
+
"step": 75
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.55,
|
| 162 |
+
"learning_rate": 0.0002970919687390088,
|
| 163 |
+
"loss": 2.4216,
|
| 164 |
+
"step": 78
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.57,
|
| 168 |
+
"learning_rate": 0.00029666330069005225,
|
| 169 |
+
"loss": 2.3679,
|
| 170 |
+
"step": 81
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.59,
|
| 174 |
+
"learning_rate": 0.000296205525182313,
|
| 175 |
+
"loss": 2.4738,
|
| 176 |
+
"step": 84
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.61,
|
| 180 |
+
"learning_rate": 0.00029571873306797876,
|
| 181 |
+
"loss": 2.4424,
|
| 182 |
+
"step": 87
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.63,
|
| 186 |
+
"learning_rate": 0.0002952030209580035,
|
| 187 |
+
"loss": 2.4183,
|
| 188 |
+
"step": 90
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.66,
|
| 192 |
+
"learning_rate": 0.00029465849120293385,
|
| 193 |
+
"loss": 2.368,
|
| 194 |
+
"step": 93
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.68,
|
| 198 |
+
"learning_rate": 0.0002940852518725959,
|
| 199 |
+
"loss": 2.3876,
|
| 200 |
+
"step": 96
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.7,
|
| 204 |
+
"learning_rate": 0.00029348341673464726,
|
| 205 |
+
"loss": 2.3105,
|
| 206 |
+
"step": 99
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.72,
|
| 210 |
+
"learning_rate": 0.00029285310523199823,
|
| 211 |
+
"loss": 2.3397,
|
| 212 |
+
"step": 102
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.74,
|
| 216 |
+
"learning_rate": 0.0002921944424591067,
|
| 217 |
+
"loss": 2.3759,
|
| 218 |
+
"step": 105
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.76,
|
| 222 |
+
"learning_rate": 0.0002915075591371512,
|
| 223 |
+
"loss": 2.4231,
|
| 224 |
+
"step": 108
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.78,
|
| 228 |
+
"learning_rate": 0.0002907925915880874,
|
| 229 |
+
"loss": 2.3697,
|
| 230 |
+
"step": 111
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.8,
|
| 234 |
+
"learning_rate": 0.00029004968170759317,
|
| 235 |
+
"loss": 2.309,
|
| 236 |
+
"step": 114
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.83,
|
| 240 |
+
"learning_rate": 0.0002892789769369072,
|
| 241 |
+
"loss": 2.2677,
|
| 242 |
+
"step": 117
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.85,
|
| 246 |
+
"learning_rate": 0.00028848063023356724,
|
| 247 |
+
"loss": 2.3125,
|
| 248 |
+
"step": 120
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.87,
|
| 252 |
+
"learning_rate": 0.0002876548000410533,
|
| 253 |
+
"loss": 2.3249,
|
| 254 |
+
"step": 123
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.89,
|
| 258 |
+
"learning_rate": 0.0002868016502573425,
|
| 259 |
+
"loss": 2.276,
|
| 260 |
+
"step": 126
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.91,
|
| 264 |
+
"learning_rate": 0.0002859213502023809,
|
| 265 |
+
"loss": 2.3264,
|
| 266 |
+
"step": 129
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.93,
|
| 270 |
+
"learning_rate": 0.0002850140745844796,
|
| 271 |
+
"loss": 2.2838,
|
| 272 |
+
"step": 132
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.95,
|
| 276 |
+
"learning_rate": 0.0002840800034656413,
|
| 277 |
+
"loss": 2.268,
|
| 278 |
+
"step": 135
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.97,
|
| 282 |
+
"learning_rate": 0.0002831193222258246,
|
| 283 |
+
"loss": 2.2771,
|
| 284 |
+
"step": 138
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.99,
|
| 288 |
+
"learning_rate": 0.00028213222152615234,
|
| 289 |
+
"loss": 2.3053,
|
| 290 |
+
"step": 141
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.99,
|
| 294 |
+
"eval_accuracy": 0.6064227132633339,
|
| 295 |
+
"eval_loss": 2.1757757663726807,
|
| 296 |
+
"eval_runtime": 10.2199,
|
| 297 |
+
"eval_samples_per_second": 197.361,
|
| 298 |
+
"eval_steps_per_second": 98.729,
|
| 299 |
+
"step": 141
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.02,
|
| 303 |
+
"learning_rate": 0.0002811188972710721,
|
| 304 |
+
"loss": 2.7816,
|
| 305 |
+
"step": 144
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.04,
|
| 309 |
+
"learning_rate": 0.00028007955056947675,
|
| 310 |
+
"loss": 2.2354,
|
| 311 |
+
"step": 147
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.06,
|
| 315 |
+
"learning_rate": 0.00027901438769479056,
|
| 316 |
+
"loss": 2.2302,
|
| 317 |
+
"step": 150
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.08,
|
| 321 |
+
"learning_rate": 0.0002779236200440321,
|
| 322 |
+
"loss": 2.2056,
|
| 323 |
+
"step": 153
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 1.11,
|
| 327 |
+
"learning_rate": 0.00027680746409585865,
|
| 328 |
+
"loss": 2.2774,
|
| 329 |
+
"step": 156
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.13,
|
| 333 |
+
"learning_rate": 0.00027566614136760364,
|
| 334 |
+
"loss": 2.2402,
|
| 335 |
+
"step": 159
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 1.15,
|
| 339 |
+
"learning_rate": 0.0002744998783713127,
|
| 340 |
+
"loss": 2.2697,
|
| 341 |
+
"step": 162
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 1.17,
|
| 345 |
+
"learning_rate": 0.00027330890656878943,
|
| 346 |
+
"loss": 2.2966,
|
| 347 |
+
"step": 165
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.19,
|
| 351 |
+
"learning_rate": 0.0002720934623256582,
|
| 352 |
+
"loss": 2.2673,
|
| 353 |
+
"step": 168
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.21,
|
| 357 |
+
"learning_rate": 0.00027085378686445433,
|
| 358 |
+
"loss": 2.2083,
|
| 359 |
+
"step": 171
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.23,
|
| 363 |
+
"learning_rate": 0.0002695901262167495,
|
| 364 |
+
"loss": 2.1908,
|
| 365 |
+
"step": 174
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 1.25,
|
| 369 |
+
"learning_rate": 0.0002683027311743236,
|
| 370 |
+
"loss": 2.1983,
|
| 371 |
+
"step": 177
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.28,
|
| 375 |
+
"learning_rate": 0.00026699185723939124,
|
| 376 |
+
"loss": 2.1856,
|
| 377 |
+
"step": 180
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.3,
|
| 381 |
+
"learning_rate": 0.0002656577645738938,
|
| 382 |
+
"loss": 2.232,
|
| 383 |
+
"step": 183
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.32,
|
| 387 |
+
"learning_rate": 0.00026430071794786644,
|
| 388 |
+
"loss": 2.2705,
|
| 389 |
+
"step": 186
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.34,
|
| 393 |
+
"learning_rate": 0.00026292098668689043,
|
| 394 |
+
"loss": 2.1916,
|
| 395 |
+
"step": 189
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.36,
|
| 399 |
+
"learning_rate": 0.00026151884461864214,
|
| 400 |
+
"loss": 2.2189,
|
| 401 |
+
"step": 192
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.38,
|
| 405 |
+
"learning_rate": 0.0002600945700185474,
|
| 406 |
+
"loss": 2.2082,
|
| 407 |
+
"step": 195
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 1.4,
|
| 411 |
+
"learning_rate": 0.0002586484455545538,
|
| 412 |
+
"loss": 2.1623,
|
| 413 |
+
"step": 198
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.42,
|
| 417 |
+
"learning_rate": 0.00025718075823103165,
|
| 418 |
+
"loss": 2.1774,
|
| 419 |
+
"step": 201
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 1.44,
|
| 423 |
+
"learning_rate": 0.0002556917993318131,
|
| 424 |
+
"loss": 2.1522,
|
| 425 |
+
"step": 204
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.47,
|
| 429 |
+
"learning_rate": 0.000254181864362383,
|
| 430 |
+
"loss": 2.1744,
|
| 431 |
+
"step": 207
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.49,
|
| 435 |
+
"learning_rate": 0.0002526512529912317,
|
| 436 |
+
"loss": 2.2299,
|
| 437 |
+
"step": 210
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.51,
|
| 441 |
+
"learning_rate": 0.00025110026899038105,
|
| 442 |
+
"loss": 2.2391,
|
| 443 |
+
"step": 213
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.53,
|
| 447 |
+
"learning_rate": 0.00024952922017509687,
|
| 448 |
+
"loss": 2.1867,
|
| 449 |
+
"step": 216
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 1.55,
|
| 453 |
+
"learning_rate": 0.00024793841834279824,
|
| 454 |
+
"loss": 2.2197,
|
| 455 |
+
"step": 219
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.57,
|
| 459 |
+
"learning_rate": 0.00024632817921117696,
|
| 460 |
+
"loss": 2.1504,
|
| 461 |
+
"step": 222
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.59,
|
| 465 |
+
"learning_rate": 0.0002446988223555388,
|
| 466 |
+
"loss": 2.1684,
|
| 467 |
+
"step": 225
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 1.61,
|
| 471 |
+
"learning_rate": 0.00024305067114537896,
|
| 472 |
+
"loss": 2.1997,
|
| 473 |
+
"step": 228
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.63,
|
| 477 |
+
"learning_rate": 0.0002413840526802049,
|
| 478 |
+
"loss": 2.1836,
|
| 479 |
+
"step": 231
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.66,
|
| 483 |
+
"learning_rate": 0.00023969929772461865,
|
| 484 |
+
"loss": 2.1628,
|
| 485 |
+
"step": 234
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.68,
|
| 489 |
+
"learning_rate": 0.0002379967406426717,
|
| 490 |
+
"loss": 2.1886,
|
| 491 |
+
"step": 237
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 1.7,
|
| 495 |
+
"learning_rate": 0.00023627671933150574,
|
| 496 |
+
"loss": 2.1225,
|
| 497 |
+
"step": 240
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.72,
|
| 501 |
+
"learning_rate": 0.00023453957515429192,
|
| 502 |
+
"loss": 2.2524,
|
| 503 |
+
"step": 243
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 1.74,
|
| 507 |
+
"learning_rate": 0.0002327856528724825,
|
| 508 |
+
"loss": 2.1975,
|
| 509 |
+
"step": 246
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.76,
|
| 513 |
+
"learning_rate": 0.00023101530057738793,
|
| 514 |
+
"loss": 2.1573,
|
| 515 |
+
"step": 249
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.78,
|
| 519 |
+
"learning_rate": 0.00022922886962109296,
|
| 520 |
+
"loss": 2.0817,
|
| 521 |
+
"step": 252
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.8,
|
| 525 |
+
"learning_rate": 0.00022742671454672588,
|
| 526 |
+
"loss": 2.1674,
|
| 527 |
+
"step": 255
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.83,
|
| 531 |
+
"learning_rate": 0.00022560919301809414,
|
| 532 |
+
"loss": 2.131,
|
| 533 |
+
"step": 258
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 1.85,
|
| 537 |
+
"learning_rate": 0.00022377666574870103,
|
| 538 |
+
"loss": 2.1338,
|
| 539 |
+
"step": 261
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.87,
|
| 543 |
+
"learning_rate": 0.00022192949643015655,
|
| 544 |
+
"loss": 2.1217,
|
| 545 |
+
"step": 264
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 1.89,
|
| 549 |
+
"learning_rate": 0.00022006805165999764,
|
| 550 |
+
"loss": 2.1887,
|
| 551 |
+
"step": 267
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 1.91,
|
| 555 |
+
"learning_rate": 0.00021819270086893185,
|
| 556 |
+
"loss": 2.1461,
|
| 557 |
+
"step": 270
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.93,
|
| 561 |
+
"learning_rate": 0.00021630381624751795,
|
| 562 |
+
"loss": 2.0757,
|
| 563 |
+
"step": 273
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.95,
|
| 567 |
+
"learning_rate": 0.00021440177267229984,
|
| 568 |
+
"loss": 2.1336,
|
| 569 |
+
"step": 276
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.97,
|
| 573 |
+
"learning_rate": 0.0002124869476314064,
|
| 574 |
+
"loss": 2.133,
|
| 575 |
+
"step": 279
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 1.99,
|
| 579 |
+
"learning_rate": 0.0002105597211496339,
|
| 580 |
+
"loss": 2.1556,
|
| 581 |
+
"step": 282
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 1.99,
|
| 585 |
+
"eval_accuracy": 0.6236541534592513,
|
| 586 |
+
"eval_loss": 2.058671236038208,
|
| 587 |
+
"eval_runtime": 10.2746,
|
| 588 |
+
"eval_samples_per_second": 196.309,
|
| 589 |
+
"eval_steps_per_second": 98.203,
|
| 590 |
+
"step": 282
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.02,
|
| 594 |
+
"learning_rate": 0.0002086204757130243,
|
| 595 |
+
"loss": 2.6096,
|
| 596 |
+
"step": 285
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 2.04,
|
| 600 |
+
"learning_rate": 0.0002066695961929554,
|
| 601 |
+
"loss": 2.1364,
|
| 602 |
+
"step": 288
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 2.06,
|
| 606 |
+
"learning_rate": 0.00020470746976975762,
|
| 607 |
+
"loss": 2.0328,
|
| 608 |
+
"step": 291
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.08,
|
| 612 |
+
"learning_rate": 0.00020273448585587248,
|
| 613 |
+
"loss": 2.1069,
|
| 614 |
+
"step": 294
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.11,
|
| 618 |
+
"learning_rate": 0.00020075103601856773,
|
| 619 |
+
"loss": 2.1156,
|
| 620 |
+
"step": 297
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.13,
|
| 624 |
+
"learning_rate": 0.00019875751390222566,
|
| 625 |
+
"loss": 2.0617,
|
| 626 |
+
"step": 300
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 2.15,
|
| 630 |
+
"learning_rate": 0.0001967543151502182,
|
| 631 |
+
"loss": 2.1158,
|
| 632 |
+
"step": 303
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.17,
|
| 636 |
+
"learning_rate": 0.00019474183732638608,
|
| 637 |
+
"loss": 2.0996,
|
| 638 |
+
"step": 306
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.19,
|
| 642 |
+
"learning_rate": 0.00019272047983613628,
|
| 643 |
+
"loss": 2.1118,
|
| 644 |
+
"step": 309
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 2.21,
|
| 648 |
+
"learning_rate": 0.00019069064384717442,
|
| 649 |
+
"loss": 2.0661,
|
| 650 |
+
"step": 312
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.23,
|
| 654 |
+
"learning_rate": 0.0001886527322098871,
|
| 655 |
+
"loss": 2.0963,
|
| 656 |
+
"step": 315
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.25,
|
| 660 |
+
"learning_rate": 0.00018660714937739043,
|
| 661 |
+
"loss": 2.039,
|
| 662 |
+
"step": 318
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.28,
|
| 666 |
+
"learning_rate": 0.00018455430132526047,
|
| 667 |
+
"loss": 2.1014,
|
| 668 |
+
"step": 321
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 2.3,
|
| 672 |
+
"learning_rate": 0.00018249459547096147,
|
| 673 |
+
"loss": 2.1209,
|
| 674 |
+
"step": 324
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.32,
|
| 678 |
+
"learning_rate": 0.00018042844059298798,
|
| 679 |
+
"loss": 2.0982,
|
| 680 |
+
"step": 327
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 2.34,
|
| 684 |
+
"learning_rate": 0.00017835624674973673,
|
| 685 |
+
"loss": 2.0784,
|
| 686 |
+
"step": 330
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 2.36,
|
| 690 |
+
"learning_rate": 0.00017627842519812485,
|
| 691 |
+
"loss": 2.0276,
|
| 692 |
+
"step": 333
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.38,
|
| 696 |
+
"learning_rate": 0.0001741953883119696,
|
| 697 |
+
"loss": 2.0388,
|
| 698 |
+
"step": 336
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 2.4,
|
| 702 |
+
"learning_rate": 0.00017210754950014714,
|
| 703 |
+
"loss": 2.0352,
|
| 704 |
+
"step": 339
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 2.42,
|
| 708 |
+
"learning_rate": 0.00017001532312454533,
|
| 709 |
+
"loss": 2.0596,
|
| 710 |
+
"step": 342
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 2.44,
|
| 714 |
+
"learning_rate": 0.00016791912441782778,
|
| 715 |
+
"loss": 2.0419,
|
| 716 |
+
"step": 345
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.47,
|
| 720 |
+
"learning_rate": 0.00016581936940102474,
|
| 721 |
+
"loss": 2.1015,
|
| 722 |
+
"step": 348
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 2.49,
|
| 726 |
+
"learning_rate": 0.00016371647480096783,
|
| 727 |
+
"loss": 2.0491,
|
| 728 |
+
"step": 351
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 2.51,
|
| 732 |
+
"learning_rate": 0.00016161085796758442,
|
| 733 |
+
"loss": 2.0816,
|
| 734 |
+
"step": 354
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 2.53,
|
| 738 |
+
"learning_rate": 0.00015950293679106852,
|
| 739 |
+
"loss": 2.0958,
|
| 740 |
+
"step": 357
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 2.55,
|
| 744 |
+
"learning_rate": 0.00015739312961894465,
|
| 745 |
+
"loss": 2.084,
|
| 746 |
+
"step": 360
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 2.57,
|
| 750 |
+
"learning_rate": 0.00015528185517304027,
|
| 751 |
+
"loss": 2.0794,
|
| 752 |
+
"step": 363
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 2.59,
|
| 756 |
+
"learning_rate": 0.00015316953246638482,
|
| 757 |
+
"loss": 2.0948,
|
| 758 |
+
"step": 366
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.61,
|
| 762 |
+
"learning_rate": 0.00015105658072005004,
|
| 763 |
+
"loss": 2.0494,
|
| 764 |
+
"step": 369
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 2.63,
|
| 768 |
+
"learning_rate": 0.00014894341927994994,
|
| 769 |
+
"loss": 2.0161,
|
| 770 |
+
"step": 372
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 2.66,
|
| 774 |
+
"learning_rate": 0.0001468304675336152,
|
| 775 |
+
"loss": 2.0325,
|
| 776 |
+
"step": 375
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.68,
|
| 780 |
+
"learning_rate": 0.00014471814482695968,
|
| 781 |
+
"loss": 2.0537,
|
| 782 |
+
"step": 378
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.7,
|
| 786 |
+
"learning_rate": 0.00014260687038105532,
|
| 787 |
+
"loss": 2.0511,
|
| 788 |
+
"step": 381
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 2.72,
|
| 792 |
+
"learning_rate": 0.00014049706320893148,
|
| 793 |
+
"loss": 2.069,
|
| 794 |
+
"step": 384
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 2.74,
|
| 798 |
+
"learning_rate": 0.0001383891420324156,
|
| 799 |
+
"loss": 2.0702,
|
| 800 |
+
"step": 387
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.76,
|
| 804 |
+
"learning_rate": 0.00013628352519903214,
|
| 805 |
+
"loss": 2.0198,
|
| 806 |
+
"step": 390
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 2.78,
|
| 810 |
+
"learning_rate": 0.00013418063059897524,
|
| 811 |
+
"loss": 2.0515,
|
| 812 |
+
"step": 393
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 2.8,
|
| 816 |
+
"learning_rate": 0.0001320808755821722,
|
| 817 |
+
"loss": 2.0652,
|
| 818 |
+
"step": 396
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.83,
|
| 822 |
+
"learning_rate": 0.00012998467687545465,
|
| 823 |
+
"loss": 2.0401,
|
| 824 |
+
"step": 399
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.85,
|
| 828 |
+
"learning_rate": 0.00012789245049985284,
|
| 829 |
+
"loss": 1.9955,
|
| 830 |
+
"step": 402
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.87,
|
| 834 |
+
"learning_rate": 0.00012580461168803036,
|
| 835 |
+
"loss": 1.9932,
|
| 836 |
+
"step": 405
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.89,
|
| 840 |
+
"learning_rate": 0.00012372157480187513,
|
| 841 |
+
"loss": 1.9957,
|
| 842 |
+
"step": 408
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.91,
|
| 846 |
+
"learning_rate": 0.00012164375325026323,
|
| 847 |
+
"loss": 2.0278,
|
| 848 |
+
"step": 411
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 2.93,
|
| 852 |
+
"learning_rate": 0.00011957155940701203,
|
| 853 |
+
"loss": 2.0055,
|
| 854 |
+
"step": 414
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 2.95,
|
| 858 |
+
"learning_rate": 0.00011750540452903852,
|
| 859 |
+
"loss": 2.1282,
|
| 860 |
+
"step": 417
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.97,
|
| 864 |
+
"learning_rate": 0.00011544569867473953,
|
| 865 |
+
"loss": 2.0752,
|
| 866 |
+
"step": 420
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.99,
|
| 870 |
+
"learning_rate": 0.00011339285062260957,
|
| 871 |
+
"loss": 2.0616,
|
| 872 |
+
"step": 423
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.99,
|
| 876 |
+
"eval_accuracy": 0.6354889456059845,
|
| 877 |
+
"eval_loss": 1.9779837131500244,
|
| 878 |
+
"eval_runtime": 10.2076,
|
| 879 |
+
"eval_samples_per_second": 197.597,
|
| 880 |
+
"eval_steps_per_second": 98.848,
|
| 881 |
+
"step": 423
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 3.02,
|
| 885 |
+
"learning_rate": 0.00011134726779011288,
|
| 886 |
+
"loss": 2.5405,
|
| 887 |
+
"step": 426
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 3.04,
|
| 891 |
+
"learning_rate": 0.00010930935615282558,
|
| 892 |
+
"loss": 2.0141,
|
| 893 |
+
"step": 429
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 3.06,
|
| 897 |
+
"learning_rate": 0.00010727952016386371,
|
| 898 |
+
"loss": 2.039,
|
| 899 |
+
"step": 432
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 3.08,
|
| 903 |
+
"learning_rate": 0.00010525816267361396,
|
| 904 |
+
"loss": 2.0632,
|
| 905 |
+
"step": 435
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.11,
|
| 909 |
+
"learning_rate": 0.00010324568484978182,
|
| 910 |
+
"loss": 1.9622,
|
| 911 |
+
"step": 438
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 3.13,
|
| 915 |
+
"learning_rate": 0.00010124248609777434,
|
| 916 |
+
"loss": 2.0226,
|
| 917 |
+
"step": 441
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 3.15,
|
| 921 |
+
"learning_rate": 9.924896398143224e-05,
|
| 922 |
+
"loss": 2.0174,
|
| 923 |
+
"step": 444
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 3.17,
|
| 927 |
+
"learning_rate": 9.726551414412753e-05,
|
| 928 |
+
"loss": 1.9791,
|
| 929 |
+
"step": 447
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 3.19,
|
| 933 |
+
"learning_rate": 9.529253023024234e-05,
|
| 934 |
+
"loss": 2.0418,
|
| 935 |
+
"step": 450
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 3.21,
|
| 939 |
+
"learning_rate": 9.33304038070446e-05,
|
| 940 |
+
"loss": 1.9825,
|
| 941 |
+
"step": 453
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 3.23,
|
| 945 |
+
"learning_rate": 9.137952428697568e-05,
|
| 946 |
+
"loss": 2.0014,
|
| 947 |
+
"step": 456
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.25,
|
| 951 |
+
"learning_rate": 8.944027885036605e-05,
|
| 952 |
+
"loss": 1.9395,
|
| 953 |
+
"step": 459
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 3.28,
|
| 957 |
+
"learning_rate": 8.751305236859359e-05,
|
| 958 |
+
"loss": 2.0189,
|
| 959 |
+
"step": 462
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 3.3,
|
| 963 |
+
"learning_rate": 8.559822732770019e-05,
|
| 964 |
+
"loss": 1.9629,
|
| 965 |
+
"step": 465
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 3.32,
|
| 969 |
+
"learning_rate": 8.3696183752482e-05,
|
| 970 |
+
"loss": 2.0085,
|
| 971 |
+
"step": 468
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 3.34,
|
| 975 |
+
"learning_rate": 8.180729913106815e-05,
|
| 976 |
+
"loss": 1.9725,
|
| 977 |
+
"step": 471
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 3.36,
|
| 981 |
+
"learning_rate": 7.993194834000229e-05,
|
| 982 |
+
"loss": 1.9946,
|
| 983 |
+
"step": 474
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 3.38,
|
| 987 |
+
"learning_rate": 7.807050356984345e-05,
|
| 988 |
+
"loss": 2.0058,
|
| 989 |
+
"step": 477
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 3.4,
|
| 993 |
+
"learning_rate": 7.622333425129895e-05,
|
| 994 |
+
"loss": 2.0525,
|
| 995 |
+
"step": 480
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 3.42,
|
| 999 |
+
"learning_rate": 7.439080698190579e-05,
|
| 1000 |
+
"loss": 1.9817,
|
| 1001 |
+
"step": 483
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 3.44,
|
| 1005 |
+
"learning_rate": 7.25732854532741e-05,
|
| 1006 |
+
"loss": 2.0297,
|
| 1007 |
+
"step": 486
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 3.47,
|
| 1011 |
+
"learning_rate": 7.077113037890702e-05,
|
| 1012 |
+
"loss": 2.0163,
|
| 1013 |
+
"step": 489
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 3.49,
|
| 1017 |
+
"learning_rate": 6.898469942261206e-05,
|
| 1018 |
+
"loss": 2.0449,
|
| 1019 |
+
"step": 492
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 3.51,
|
| 1023 |
+
"learning_rate": 6.721434712751744e-05,
|
| 1024 |
+
"loss": 1.9419,
|
| 1025 |
+
"step": 495
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 3.53,
|
| 1029 |
+
"learning_rate": 6.546042484570807e-05,
|
| 1030 |
+
"loss": 2.0181,
|
| 1031 |
+
"step": 498
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.55,
|
| 1035 |
+
"learning_rate": 6.372328066849426e-05,
|
| 1036 |
+
"loss": 1.9761,
|
| 1037 |
+
"step": 501
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 3.57,
|
| 1041 |
+
"learning_rate": 6.200325935732827e-05,
|
| 1042 |
+
"loss": 1.9969,
|
| 1043 |
+
"step": 504
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 3.59,
|
| 1047 |
+
"learning_rate": 6.030070227538133e-05,
|
| 1048 |
+
"loss": 1.9515,
|
| 1049 |
+
"step": 507
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 3.61,
|
| 1053 |
+
"learning_rate": 5.8615947319795044e-05,
|
| 1054 |
+
"loss": 2.0222,
|
| 1055 |
+
"step": 510
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 3.63,
|
| 1059 |
+
"learning_rate": 5.6949328854621045e-05,
|
| 1060 |
+
"loss": 2.0148,
|
| 1061 |
+
"step": 513
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 3.66,
|
| 1065 |
+
"learning_rate": 5.5301177644461164e-05,
|
| 1066 |
+
"loss": 2.0086,
|
| 1067 |
+
"step": 516
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 3.68,
|
| 1071 |
+
"learning_rate": 5.367182078882299e-05,
|
| 1072 |
+
"loss": 2.0048,
|
| 1073 |
+
"step": 519
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.7,
|
| 1077 |
+
"learning_rate": 5.206158165720175e-05,
|
| 1078 |
+
"loss": 1.967,
|
| 1079 |
+
"step": 522
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 3.72,
|
| 1083 |
+
"learning_rate": 5.04707798249031e-05,
|
| 1084 |
+
"loss": 1.9731,
|
| 1085 |
+
"step": 525
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 3.74,
|
| 1089 |
+
"learning_rate": 4.889973100961892e-05,
|
| 1090 |
+
"loss": 1.961,
|
| 1091 |
+
"step": 528
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 3.76,
|
| 1095 |
+
"learning_rate": 4.734874700876826e-05,
|
| 1096 |
+
"loss": 1.9674,
|
| 1097 |
+
"step": 531
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 3.78,
|
| 1101 |
+
"learning_rate": 4.5818135637616936e-05,
|
| 1102 |
+
"loss": 1.9797,
|
| 1103 |
+
"step": 534
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 3.8,
|
| 1107 |
+
"learning_rate": 4.430820066818692e-05,
|
| 1108 |
+
"loss": 2.0148,
|
| 1109 |
+
"step": 537
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 3.83,
|
| 1113 |
+
"learning_rate": 4.281924176896837e-05,
|
| 1114 |
+
"loss": 1.9882,
|
| 1115 |
+
"step": 540
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 3.85,
|
| 1119 |
+
"learning_rate": 4.13515544454462e-05,
|
| 1120 |
+
"loss": 1.9908,
|
| 1121 |
+
"step": 543
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 3.87,
|
| 1125 |
+
"learning_rate": 3.990542998145262e-05,
|
| 1126 |
+
"loss": 1.9337,
|
| 1127 |
+
"step": 546
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 3.89,
|
| 1131 |
+
"learning_rate": 3.8481155381357825e-05,
|
| 1132 |
+
"loss": 1.9709,
|
| 1133 |
+
"step": 549
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 3.91,
|
| 1137 |
+
"learning_rate": 3.7079013313109535e-05,
|
| 1138 |
+
"loss": 1.9485,
|
| 1139 |
+
"step": 552
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 3.93,
|
| 1143 |
+
"learning_rate": 3.569928205213354e-05,
|
| 1144 |
+
"loss": 1.955,
|
| 1145 |
+
"step": 555
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 3.95,
|
| 1149 |
+
"learning_rate": 3.434223542610615e-05,
|
| 1150 |
+
"loss": 1.9797,
|
| 1151 |
+
"step": 558
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 3.97,
|
| 1155 |
+
"learning_rate": 3.30081427606087e-05,
|
| 1156 |
+
"loss": 1.9471,
|
| 1157 |
+
"step": 561
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 3.99,
|
| 1161 |
+
"learning_rate": 3.169726882567642e-05,
|
| 1162 |
+
"loss": 2.0084,
|
| 1163 |
+
"step": 564
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 3.99,
|
| 1167 |
+
"eval_accuracy": 0.6422005133308889,
|
| 1168 |
+
"eval_loss": 1.9317268133163452,
|
| 1169 |
+
"eval_runtime": 10.2419,
|
| 1170 |
+
"eval_samples_per_second": 196.936,
|
| 1171 |
+
"eval_steps_per_second": 98.517,
|
| 1172 |
+
"step": 564
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 4.02,
|
| 1176 |
+
"learning_rate": 3.0409873783250476e-05,
|
| 1177 |
+
"loss": 2.5298,
|
| 1178 |
+
"step": 567
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 4.04,
|
| 1182 |
+
"learning_rate": 2.914621313554567e-05,
|
| 1183 |
+
"loss": 1.9335,
|
| 1184 |
+
"step": 570
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 4.06,
|
| 1188 |
+
"learning_rate": 2.790653767434181e-05,
|
| 1189 |
+
"loss": 1.9663,
|
| 1190 |
+
"step": 573
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 4.08,
|
| 1194 |
+
"learning_rate": 2.6691093431210596e-05,
|
| 1195 |
+
"loss": 1.9546,
|
| 1196 |
+
"step": 576
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 4.11,
|
| 1200 |
+
"learning_rate": 2.550012162868731e-05,
|
| 1201 |
+
"loss": 1.9588,
|
| 1202 |
+
"step": 579
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 4.13,
|
| 1206 |
+
"learning_rate": 2.4333858632396336e-05,
|
| 1207 |
+
"loss": 1.9079,
|
| 1208 |
+
"step": 582
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 4.15,
|
| 1212 |
+
"learning_rate": 2.319253590414132e-05,
|
| 1213 |
+
"loss": 1.9706,
|
| 1214 |
+
"step": 585
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 4.17,
|
| 1218 |
+
"learning_rate": 2.207637995596791e-05,
|
| 1219 |
+
"loss": 1.9244,
|
| 1220 |
+
"step": 588
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 4.19,
|
| 1224 |
+
"learning_rate": 2.0985612305209404e-05,
|
| 1225 |
+
"loss": 1.9571,
|
| 1226 |
+
"step": 591
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 4.21,
|
| 1230 |
+
"learning_rate": 1.9920449430523266e-05,
|
| 1231 |
+
"loss": 1.9935,
|
| 1232 |
+
"step": 594
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 4.23,
|
| 1236 |
+
"learning_rate": 1.8881102728927854e-05,
|
| 1237 |
+
"loss": 1.965,
|
| 1238 |
+
"step": 597
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 4.25,
|
| 1242 |
+
"learning_rate": 1.7867778473847695e-05,
|
| 1243 |
+
"loss": 1.9879,
|
| 1244 |
+
"step": 600
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 4.28,
|
| 1248 |
+
"learning_rate": 1.6880677774175354e-05,
|
| 1249 |
+
"loss": 1.9765,
|
| 1250 |
+
"step": 603
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 4.3,
|
| 1254 |
+
"learning_rate": 1.5919996534358635e-05,
|
| 1255 |
+
"loss": 1.9608,
|
| 1256 |
+
"step": 606
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 4.32,
|
| 1260 |
+
"learning_rate": 1.4985925415520395e-05,
|
| 1261 |
+
"loss": 1.8867,
|
| 1262 |
+
"step": 609
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 4.34,
|
| 1266 |
+
"learning_rate": 1.407864979761908e-05,
|
| 1267 |
+
"loss": 1.9532,
|
| 1268 |
+
"step": 612
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 4.36,
|
| 1272 |
+
"learning_rate": 1.31983497426575e-05,
|
| 1273 |
+
"loss": 1.9672,
|
| 1274 |
+
"step": 615
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 4.38,
|
| 1278 |
+
"learning_rate": 1.2345199958946673e-05,
|
| 1279 |
+
"loss": 1.9372,
|
| 1280 |
+
"step": 618
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 4.4,
|
| 1284 |
+
"learning_rate": 1.1519369766432762e-05,
|
| 1285 |
+
"loss": 1.9676,
|
| 1286 |
+
"step": 621
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 4.42,
|
| 1290 |
+
"learning_rate": 1.072102306309276e-05,
|
| 1291 |
+
"loss": 1.9726,
|
| 1292 |
+
"step": 624
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 4.44,
|
| 1296 |
+
"learning_rate": 9.950318292406816e-06,
|
| 1297 |
+
"loss": 1.9565,
|
| 1298 |
+
"step": 627
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 4.47,
|
| 1302 |
+
"learning_rate": 9.207408411912603e-06,
|
| 1303 |
+
"loss": 2.0322,
|
| 1304 |
+
"step": 630
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 4.49,
|
| 1308 |
+
"learning_rate": 8.492440862848787e-06,
|
| 1309 |
+
"loss": 1.9868,
|
| 1310 |
+
"step": 633
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 4.51,
|
| 1314 |
+
"learning_rate": 7.805557540893276e-06,
|
| 1315 |
+
"loss": 1.9788,
|
| 1316 |
+
"step": 636
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 4.53,
|
| 1320 |
+
"learning_rate": 7.146894768001743e-06,
|
| 1321 |
+
"loss": 1.9226,
|
| 1322 |
+
"step": 639
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 4.55,
|
| 1326 |
+
"learning_rate": 6.516583265352754e-06,
|
| 1327 |
+
"loss": 1.9214,
|
| 1328 |
+
"step": 642
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 4.57,
|
| 1332 |
+
"learning_rate": 5.914748127404101e-06,
|
| 1333 |
+
"loss": 1.9815,
|
| 1334 |
+
"step": 645
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 4.59,
|
| 1338 |
+
"learning_rate": 5.34150879706613e-06,
|
| 1339 |
+
"loss": 1.916,
|
| 1340 |
+
"step": 648
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 4.61,
|
| 1344 |
+
"learning_rate": 4.796979041996485e-06,
|
| 1345 |
+
"loss": 1.9162,
|
| 1346 |
+
"step": 651
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 4.63,
|
| 1350 |
+
"learning_rate": 4.2812669320212375e-06,
|
| 1351 |
+
"loss": 1.9955,
|
| 1352 |
+
"step": 654
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 4.66,
|
| 1356 |
+
"learning_rate": 3.7944748176869766e-06,
|
| 1357 |
+
"loss": 1.926,
|
| 1358 |
+
"step": 657
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 4.68,
|
| 1362 |
+
"learning_rate": 3.336699309947721e-06,
|
| 1363 |
+
"loss": 1.9637,
|
| 1364 |
+
"step": 660
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 4.7,
|
| 1368 |
+
"learning_rate": 2.908031260991195e-06,
|
| 1369 |
+
"loss": 1.9602,
|
| 1370 |
+
"step": 663
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 4.72,
|
| 1374 |
+
"learning_rate": 2.5085557462078134e-06,
|
| 1375 |
+
"loss": 2.0124,
|
| 1376 |
+
"step": 666
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 4.74,
|
| 1380 |
+
"learning_rate": 2.1383520473062743e-06,
|
| 1381 |
+
"loss": 1.9393,
|
| 1382 |
+
"step": 669
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 4.76,
|
| 1386 |
+
"learning_rate": 1.7974936365789116e-06,
|
| 1387 |
+
"loss": 1.8888,
|
| 1388 |
+
"step": 672
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 4.78,
|
| 1392 |
+
"learning_rate": 1.4860481623201414e-06,
|
| 1393 |
+
"loss": 1.9409,
|
| 1394 |
+
"step": 675
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 4.8,
|
| 1398 |
+
"learning_rate": 1.2040774354005856e-06,
|
| 1399 |
+
"loss": 1.9617,
|
| 1400 |
+
"step": 678
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 4.83,
|
| 1404 |
+
"learning_rate": 9.516374169998508e-07,
|
| 1405 |
+
"loss": 1.9433,
|
| 1406 |
+
"step": 681
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 4.85,
|
| 1410 |
+
"learning_rate": 7.287782075002302e-07,
|
| 1411 |
+
"loss": 2.0248,
|
| 1412 |
+
"step": 684
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 4.87,
|
| 1416 |
+
"learning_rate": 5.355440365434893e-07,
|
| 1417 |
+
"loss": 1.9528,
|
| 1418 |
+
"step": 687
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 4.89,
|
| 1422 |
+
"learning_rate": 3.7197325425297186e-07,
|
| 1423 |
+
"loss": 1.9823,
|
| 1424 |
+
"step": 690
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 4.91,
|
| 1428 |
+
"learning_rate": 2.380983236223377e-07,
|
| 1429 |
+
"loss": 1.9716,
|
| 1430 |
+
"step": 693
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 4.93,
|
| 1434 |
+
"learning_rate": 1.3394581407289996e-07,
|
| 1435 |
+
"loss": 1.9205,
|
| 1436 |
+
"step": 696
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 4.95,
|
| 1440 |
+
"learning_rate": 5.9536396180492886e-08,
|
| 1441 |
+
"loss": 1.953,
|
| 1442 |
+
"step": 699
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 4.97,
|
| 1446 |
+
"learning_rate": 1.4884837573087006e-08,
|
| 1447 |
+
"loss": 2.0213,
|
| 1448 |
+
"step": 702
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 4.99,
|
| 1452 |
+
"learning_rate": 0.0,
|
| 1453 |
+
"loss": 1.9621,
|
| 1454 |
+
"step": 705
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 4.99,
|
| 1458 |
+
"eval_accuracy": 0.6451979106514778,
|
| 1459 |
+
"eval_loss": 1.9133403301239014,
|
| 1460 |
+
"eval_runtime": 10.2297,
|
| 1461 |
+
"eval_samples_per_second": 197.171,
|
| 1462 |
+
"eval_steps_per_second": 98.634,
|
| 1463 |
+
"step": 705
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 4.99,
|
| 1467 |
+
"step": 705,
|
| 1468 |
+
"total_flos": 4119345677467648.0,
|
| 1469 |
+
"train_loss": 2.1596999946215476,
|
| 1470 |
+
"train_runtime": 1037.2411,
|
| 1471 |
+
"train_samples_per_second": 87.439,
|
| 1472 |
+
"train_steps_per_second": 0.68
|
| 1473 |
+
}
|
| 1474 |
+
],
|
| 1475 |
+
"max_steps": 705,
|
| 1476 |
+
"num_train_epochs": 5,
|
| 1477 |
+
"total_flos": 4119345677467648.0,
|
| 1478 |
+
"trial_name": null,
|
| 1479 |
+
"trial_params": null
|
| 1480 |
+
}
|