End of training
Browse files- README.md +3 -1
- all_results.json +12 -0
- eval_results.json +8 -0
- train_results.json +7 -0
- trainer_state.json +1494 -0
README.md
CHANGED
|
@@ -2,6 +2,8 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: microsoft/beit-large-patch16-512
|
| 4 |
tags:
|
|
|
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
|
@@ -15,7 +17,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# beit-large-patch16-512-types-of-film-shots-vN
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [microsoft/beit-large-patch16-512](https://huggingface.co/microsoft/beit-large-patch16-512) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 1.2335
|
| 21 |
- Accuracy: 0.6763
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: microsoft/beit-large-patch16-512
|
| 4 |
tags:
|
| 5 |
+
- image-classification
|
| 6 |
+
- vision
|
| 7 |
- generated_from_trainer
|
| 8 |
metrics:
|
| 9 |
- accuracy
|
|
|
|
| 17 |
|
| 18 |
# beit-large-patch16-512-types-of-film-shots-vN
|
| 19 |
|
| 20 |
+
This model is a fine-tuned version of [microsoft/beit-large-patch16-512](https://huggingface.co/microsoft/beit-large-patch16-512) on the szymonrucinski/types-of-film-shots dataset.
|
| 21 |
It achieves the following results on the evaluation set:
|
| 22 |
- Loss: 1.2335
|
| 23 |
- Accuracy: 0.6763
|
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.0,
|
| 3 |
+
"eval_accuracy": 0.6762589928057554,
|
| 4 |
+
"eval_loss": 1.233504295349121,
|
| 5 |
+
"eval_runtime": 14.4279,
|
| 6 |
+
"eval_samples_per_second": 9.634,
|
| 7 |
+
"eval_steps_per_second": 2.426,
|
| 8 |
+
"train_loss": 0.9992087238718637,
|
| 9 |
+
"train_runtime": 3103.0559,
|
| 10 |
+
"train_samples_per_second": 1.52,
|
| 11 |
+
"train_steps_per_second": 0.76
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.0,
|
| 3 |
+
"eval_accuracy": 0.6762589928057554,
|
| 4 |
+
"eval_loss": 1.233504295349121,
|
| 5 |
+
"eval_runtime": 14.4279,
|
| 6 |
+
"eval_samples_per_second": 9.634,
|
| 7 |
+
"eval_steps_per_second": 2.426
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.0,
|
| 3 |
+
"train_loss": 0.9992087238718637,
|
| 4 |
+
"train_runtime": 3103.0559,
|
| 5 |
+
"train_samples_per_second": 1.52,
|
| 6 |
+
"train_steps_per_second": 0.76
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1494 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2358,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03,
|
| 13 |
+
"learning_rate": 1.6949152542372882e-06,
|
| 14 |
+
"loss": 1.9885,
|
| 15 |
+
"step": 10
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.05,
|
| 19 |
+
"learning_rate": 3.3898305084745763e-06,
|
| 20 |
+
"loss": 2.7383,
|
| 21 |
+
"step": 20
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.08,
|
| 25 |
+
"learning_rate": 5.084745762711865e-06,
|
| 26 |
+
"loss": 2.5169,
|
| 27 |
+
"step": 30
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.1,
|
| 31 |
+
"learning_rate": 6.779661016949153e-06,
|
| 32 |
+
"loss": 2.3984,
|
| 33 |
+
"step": 40
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.13,
|
| 37 |
+
"learning_rate": 8.47457627118644e-06,
|
| 38 |
+
"loss": 1.9978,
|
| 39 |
+
"step": 50
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.15,
|
| 43 |
+
"learning_rate": 1.016949152542373e-05,
|
| 44 |
+
"loss": 2.2204,
|
| 45 |
+
"step": 60
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.18,
|
| 49 |
+
"learning_rate": 1.1864406779661018e-05,
|
| 50 |
+
"loss": 1.9677,
|
| 51 |
+
"step": 70
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.2,
|
| 55 |
+
"learning_rate": 1.3559322033898305e-05,
|
| 56 |
+
"loss": 2.2108,
|
| 57 |
+
"step": 80
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.23,
|
| 61 |
+
"learning_rate": 1.5254237288135594e-05,
|
| 62 |
+
"loss": 1.5879,
|
| 63 |
+
"step": 90
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.25,
|
| 67 |
+
"learning_rate": 1.694915254237288e-05,
|
| 68 |
+
"loss": 2.2145,
|
| 69 |
+
"step": 100
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.28,
|
| 73 |
+
"learning_rate": 1.864406779661017e-05,
|
| 74 |
+
"loss": 2.0709,
|
| 75 |
+
"step": 110
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.31,
|
| 79 |
+
"learning_rate": 1.9999960660084782e-05,
|
| 80 |
+
"loss": 2.4549,
|
| 81 |
+
"step": 120
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.33,
|
| 85 |
+
"learning_rate": 1.9998583795552086e-05,
|
| 86 |
+
"loss": 2.0769,
|
| 87 |
+
"step": 130
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.36,
|
| 91 |
+
"learning_rate": 1.9995240244773315e-05,
|
| 92 |
+
"loss": 1.801,
|
| 93 |
+
"step": 140
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.38,
|
| 97 |
+
"learning_rate": 1.9989930665413148e-05,
|
| 98 |
+
"loss": 1.9293,
|
| 99 |
+
"step": 150
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.41,
|
| 103 |
+
"learning_rate": 1.998265610184716e-05,
|
| 104 |
+
"loss": 1.9712,
|
| 105 |
+
"step": 160
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.43,
|
| 109 |
+
"learning_rate": 1.9973417984956404e-05,
|
| 110 |
+
"loss": 2.4024,
|
| 111 |
+
"step": 170
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.46,
|
| 115 |
+
"learning_rate": 1.996221813184596e-05,
|
| 116 |
+
"loss": 2.1186,
|
| 117 |
+
"step": 180
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.48,
|
| 121 |
+
"learning_rate": 1.9949058745487524e-05,
|
| 122 |
+
"loss": 2.1329,
|
| 123 |
+
"step": 190
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.51,
|
| 127 |
+
"learning_rate": 1.9933942414286082e-05,
|
| 128 |
+
"loss": 1.8044,
|
| 129 |
+
"step": 200
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.53,
|
| 133 |
+
"learning_rate": 1.9916872111570782e-05,
|
| 134 |
+
"loss": 2.5074,
|
| 135 |
+
"step": 210
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.56,
|
| 139 |
+
"learning_rate": 1.98978511950101e-05,
|
| 140 |
+
"loss": 1.6585,
|
| 141 |
+
"step": 220
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.59,
|
| 145 |
+
"learning_rate": 1.9876883405951378e-05,
|
| 146 |
+
"loss": 1.9255,
|
| 147 |
+
"step": 230
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.61,
|
| 151 |
+
"learning_rate": 1.9853972868684942e-05,
|
| 152 |
+
"loss": 1.5736,
|
| 153 |
+
"step": 240
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.64,
|
| 157 |
+
"learning_rate": 1.982912408963285e-05,
|
| 158 |
+
"loss": 1.6555,
|
| 159 |
+
"step": 250
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.66,
|
| 163 |
+
"learning_rate": 1.980234195646248e-05,
|
| 164 |
+
"loss": 2.1425,
|
| 165 |
+
"step": 260
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.69,
|
| 169 |
+
"learning_rate": 1.9773631737125192e-05,
|
| 170 |
+
"loss": 1.1852,
|
| 171 |
+
"step": 270
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.71,
|
| 175 |
+
"learning_rate": 1.974299907882009e-05,
|
| 176 |
+
"loss": 2.1502,
|
| 177 |
+
"step": 280
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.74,
|
| 181 |
+
"learning_rate": 1.9710450006883258e-05,
|
| 182 |
+
"loss": 1.4109,
|
| 183 |
+
"step": 290
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.76,
|
| 187 |
+
"learning_rate": 1.96759909236026e-05,
|
| 188 |
+
"loss": 1.7527,
|
| 189 |
+
"step": 300
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.79,
|
| 193 |
+
"learning_rate": 1.9639628606958535e-05,
|
| 194 |
+
"loss": 1.2232,
|
| 195 |
+
"step": 310
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.81,
|
| 199 |
+
"learning_rate": 1.960137020929078e-05,
|
| 200 |
+
"loss": 2.0229,
|
| 201 |
+
"step": 320
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.84,
|
| 205 |
+
"learning_rate": 1.956122325589152e-05,
|
| 206 |
+
"loss": 1.5771,
|
| 207 |
+
"step": 330
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.87,
|
| 211 |
+
"learning_rate": 1.9519195643525205e-05,
|
| 212 |
+
"loss": 1.5255,
|
| 213 |
+
"step": 340
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.89,
|
| 217 |
+
"learning_rate": 1.947529563887529e-05,
|
| 218 |
+
"loss": 1.6161,
|
| 219 |
+
"step": 350
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.92,
|
| 223 |
+
"learning_rate": 1.9429531876918193e-05,
|
| 224 |
+
"loss": 1.9704,
|
| 225 |
+
"step": 360
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.94,
|
| 229 |
+
"learning_rate": 1.9381913359224844e-05,
|
| 230 |
+
"loss": 1.2751,
|
| 231 |
+
"step": 370
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.97,
|
| 235 |
+
"learning_rate": 1.9332449452190096e-05,
|
| 236 |
+
"loss": 1.6944,
|
| 237 |
+
"step": 380
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.99,
|
| 241 |
+
"learning_rate": 1.928114988519039e-05,
|
| 242 |
+
"loss": 1.0435,
|
| 243 |
+
"step": 390
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.0,
|
| 247 |
+
"eval_accuracy": 0.4892086330935252,
|
| 248 |
+
"eval_loss": 1.4798991680145264,
|
| 249 |
+
"eval_runtime": 14.722,
|
| 250 |
+
"eval_samples_per_second": 9.442,
|
| 251 |
+
"eval_steps_per_second": 2.377,
|
| 252 |
+
"step": 393
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.02,
|
| 256 |
+
"learning_rate": 1.922802474867003e-05,
|
| 257 |
+
"loss": 1.9562,
|
| 258 |
+
"step": 400
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 1.04,
|
| 262 |
+
"learning_rate": 1.9173084492156406e-05,
|
| 263 |
+
"loss": 1.6383,
|
| 264 |
+
"step": 410
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.07,
|
| 268 |
+
"learning_rate": 1.911633992220464e-05,
|
| 269 |
+
"loss": 1.1736,
|
| 270 |
+
"step": 420
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.09,
|
| 274 |
+
"learning_rate": 1.9057802200271943e-05,
|
| 275 |
+
"loss": 1.0087,
|
| 276 |
+
"step": 430
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.12,
|
| 280 |
+
"learning_rate": 1.8997482840522218e-05,
|
| 281 |
+
"loss": 1.5162,
|
| 282 |
+
"step": 440
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.15,
|
| 286 |
+
"learning_rate": 1.893539370756125e-05,
|
| 287 |
+
"loss": 1.2644,
|
| 288 |
+
"step": 450
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 1.17,
|
| 292 |
+
"learning_rate": 1.8871547014103e-05,
|
| 293 |
+
"loss": 1.2083,
|
| 294 |
+
"step": 460
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.2,
|
| 298 |
+
"learning_rate": 1.880595531856738e-05,
|
| 299 |
+
"loss": 1.0534,
|
| 300 |
+
"step": 470
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 1.22,
|
| 304 |
+
"learning_rate": 1.8738631522610077e-05,
|
| 305 |
+
"loss": 0.939,
|
| 306 |
+
"step": 480
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 1.25,
|
| 310 |
+
"learning_rate": 1.8669588868584832e-05,
|
| 311 |
+
"loss": 1.4496,
|
| 312 |
+
"step": 490
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.27,
|
| 316 |
+
"learning_rate": 1.859884093693872e-05,
|
| 317 |
+
"loss": 1.2077,
|
| 318 |
+
"step": 500
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.3,
|
| 322 |
+
"learning_rate": 1.8526401643540924e-05,
|
| 323 |
+
"loss": 1.1323,
|
| 324 |
+
"step": 510
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.32,
|
| 328 |
+
"learning_rate": 1.845228523694553e-05,
|
| 329 |
+
"loss": 1.2509,
|
| 330 |
+
"step": 520
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 1.35,
|
| 334 |
+
"learning_rate": 1.8376506295588887e-05,
|
| 335 |
+
"loss": 1.2086,
|
| 336 |
+
"step": 530
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.37,
|
| 340 |
+
"learning_rate": 1.8299079724922087e-05,
|
| 341 |
+
"loss": 1.4408,
|
| 342 |
+
"step": 540
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 1.4,
|
| 346 |
+
"learning_rate": 1.8220020754479104e-05,
|
| 347 |
+
"loss": 0.8179,
|
| 348 |
+
"step": 550
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 1.42,
|
| 352 |
+
"learning_rate": 1.8139344934881195e-05,
|
| 353 |
+
"loss": 0.831,
|
| 354 |
+
"step": 560
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.45,
|
| 358 |
+
"learning_rate": 1.8057068134778174e-05,
|
| 359 |
+
"loss": 1.8693,
|
| 360 |
+
"step": 570
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.48,
|
| 364 |
+
"learning_rate": 1.797320653772707e-05,
|
| 365 |
+
"loss": 0.7718,
|
| 366 |
+
"step": 580
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.5,
|
| 370 |
+
"learning_rate": 1.7887776639008912e-05,
|
| 371 |
+
"loss": 1.3469,
|
| 372 |
+
"step": 590
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 1.53,
|
| 376 |
+
"learning_rate": 1.7800795242384126e-05,
|
| 377 |
+
"loss": 1.297,
|
| 378 |
+
"step": 600
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.55,
|
| 382 |
+
"learning_rate": 1.7712279456787334e-05,
|
| 383 |
+
"loss": 1.0095,
|
| 384 |
+
"step": 610
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.58,
|
| 388 |
+
"learning_rate": 1.762224669296206e-05,
|
| 389 |
+
"loss": 1.8096,
|
| 390 |
+
"step": 620
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 1.6,
|
| 394 |
+
"learning_rate": 1.7530714660036112e-05,
|
| 395 |
+
"loss": 1.1618,
|
| 396 |
+
"step": 630
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.63,
|
| 400 |
+
"learning_rate": 1.743770136203825e-05,
|
| 401 |
+
"loss": 1.0257,
|
| 402 |
+
"step": 640
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.65,
|
| 406 |
+
"learning_rate": 1.7343225094356857e-05,
|
| 407 |
+
"loss": 1.644,
|
| 408 |
+
"step": 650
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.68,
|
| 412 |
+
"learning_rate": 1.724730444014132e-05,
|
| 413 |
+
"loss": 1.4658,
|
| 414 |
+
"step": 660
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 1.7,
|
| 418 |
+
"learning_rate": 1.7149958266646756e-05,
|
| 419 |
+
"loss": 1.1911,
|
| 420 |
+
"step": 670
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.73,
|
| 424 |
+
"learning_rate": 1.705120572152292e-05,
|
| 425 |
+
"loss": 1.1788,
|
| 426 |
+
"step": 680
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 1.76,
|
| 430 |
+
"learning_rate": 1.695106622904791e-05,
|
| 431 |
+
"loss": 1.56,
|
| 432 |
+
"step": 690
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 1.78,
|
| 436 |
+
"learning_rate": 1.6849559486307493e-05,
|
| 437 |
+
"loss": 1.5506,
|
| 438 |
+
"step": 700
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.81,
|
| 442 |
+
"learning_rate": 1.6746705459320746e-05,
|
| 443 |
+
"loss": 1.3003,
|
| 444 |
+
"step": 710
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.83,
|
| 448 |
+
"learning_rate": 1.664252437911282e-05,
|
| 449 |
+
"loss": 1.2845,
|
| 450 |
+
"step": 720
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.86,
|
| 454 |
+
"learning_rate": 1.6537036737735563e-05,
|
| 455 |
+
"loss": 1.9069,
|
| 456 |
+
"step": 730
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 1.88,
|
| 460 |
+
"learning_rate": 1.6430263284236803e-05,
|
| 461 |
+
"loss": 1.178,
|
| 462 |
+
"step": 740
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 1.91,
|
| 466 |
+
"learning_rate": 1.63222250205791e-05,
|
| 467 |
+
"loss": 1.3253,
|
| 468 |
+
"step": 750
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 1.93,
|
| 472 |
+
"learning_rate": 1.6212943197508706e-05,
|
| 473 |
+
"loss": 1.0718,
|
| 474 |
+
"step": 760
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.96,
|
| 478 |
+
"learning_rate": 1.6102439310375637e-05,
|
| 479 |
+
"loss": 0.9187,
|
| 480 |
+
"step": 770
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.98,
|
| 484 |
+
"learning_rate": 1.5990735094905622e-05,
|
| 485 |
+
"loss": 1.1554,
|
| 486 |
+
"step": 780
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 2.0,
|
| 490 |
+
"eval_accuracy": 0.4892086330935252,
|
| 491 |
+
"eval_loss": 1.4938154220581055,
|
| 492 |
+
"eval_runtime": 14.7051,
|
| 493 |
+
"eval_samples_per_second": 9.453,
|
| 494 |
+
"eval_steps_per_second": 2.38,
|
| 495 |
+
"step": 786
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 2.01,
|
| 499 |
+
"learning_rate": 1.5877852522924733e-05,
|
| 500 |
+
"loss": 1.1346,
|
| 501 |
+
"step": 790
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.04,
|
| 505 |
+
"learning_rate": 1.576381379803763e-05,
|
| 506 |
+
"loss": 1.2287,
|
| 507 |
+
"step": 800
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 2.06,
|
| 511 |
+
"learning_rate": 1.564864135126018e-05,
|
| 512 |
+
"loss": 1.02,
|
| 513 |
+
"step": 810
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.09,
|
| 517 |
+
"learning_rate": 1.553235783660735e-05,
|
| 518 |
+
"loss": 1.2653,
|
| 519 |
+
"step": 820
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 2.11,
|
| 523 |
+
"learning_rate": 1.541498612663726e-05,
|
| 524 |
+
"loss": 0.7407,
|
| 525 |
+
"step": 830
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 2.14,
|
| 529 |
+
"learning_rate": 1.5296549307952203e-05,
|
| 530 |
+
"loss": 1.2226,
|
| 531 |
+
"step": 840
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.16,
|
| 535 |
+
"learning_rate": 1.5177070676657632e-05,
|
| 536 |
+
"loss": 0.9274,
|
| 537 |
+
"step": 850
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 2.19,
|
| 541 |
+
"learning_rate": 1.5056573733779848e-05,
|
| 542 |
+
"loss": 0.9529,
|
| 543 |
+
"step": 860
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.21,
|
| 547 |
+
"learning_rate": 1.493508218064347e-05,
|
| 548 |
+
"loss": 0.8886,
|
| 549 |
+
"step": 870
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 2.24,
|
| 553 |
+
"learning_rate": 1.4812619914209455e-05,
|
| 554 |
+
"loss": 1.0895,
|
| 555 |
+
"step": 880
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.26,
|
| 559 |
+
"learning_rate": 1.4689211022374645e-05,
|
| 560 |
+
"loss": 1.669,
|
| 561 |
+
"step": 890
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 2.29,
|
| 565 |
+
"learning_rate": 1.4564879779233773e-05,
|
| 566 |
+
"loss": 1.1092,
|
| 567 |
+
"step": 900
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 2.32,
|
| 571 |
+
"learning_rate": 1.4439650640304822e-05,
|
| 572 |
+
"loss": 1.1258,
|
| 573 |
+
"step": 910
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.34,
|
| 577 |
+
"learning_rate": 1.431354823771872e-05,
|
| 578 |
+
"loss": 0.8034,
|
| 579 |
+
"step": 920
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 2.37,
|
| 583 |
+
"learning_rate": 1.4186597375374283e-05,
|
| 584 |
+
"loss": 1.1004,
|
| 585 |
+
"step": 930
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.39,
|
| 589 |
+
"learning_rate": 1.4058823024059374e-05,
|
| 590 |
+
"loss": 0.8297,
|
| 591 |
+
"step": 940
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 2.42,
|
| 595 |
+
"learning_rate": 1.3930250316539237e-05,
|
| 596 |
+
"loss": 0.8974,
|
| 597 |
+
"step": 950
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.44,
|
| 601 |
+
"learning_rate": 1.380090454261298e-05,
|
| 602 |
+
"loss": 1.3636,
|
| 603 |
+
"step": 960
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 2.47,
|
| 607 |
+
"learning_rate": 1.3670811144139144e-05,
|
| 608 |
+
"loss": 0.9185,
|
| 609 |
+
"step": 970
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 2.49,
|
| 613 |
+
"learning_rate": 1.3539995710031391e-05,
|
| 614 |
+
"loss": 0.8012,
|
| 615 |
+
"step": 980
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.52,
|
| 619 |
+
"learning_rate": 1.340848397122525e-05,
|
| 620 |
+
"loss": 0.9984,
|
| 621 |
+
"step": 990
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.54,
|
| 625 |
+
"learning_rate": 1.3276301795616937e-05,
|
| 626 |
+
"loss": 0.8015,
|
| 627 |
+
"step": 1000
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.57,
|
| 631 |
+
"learning_rate": 1.3143475182975222e-05,
|
| 632 |
+
"loss": 0.9148,
|
| 633 |
+
"step": 1010
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 2.6,
|
| 637 |
+
"learning_rate": 1.3010030259827378e-05,
|
| 638 |
+
"loss": 1.1759,
|
| 639 |
+
"step": 1020
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.62,
|
| 643 |
+
"learning_rate": 1.2875993274320173e-05,
|
| 644 |
+
"loss": 0.606,
|
| 645 |
+
"step": 1030
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 2.65,
|
| 649 |
+
"learning_rate": 1.2741390591056964e-05,
|
| 650 |
+
"loss": 1.253,
|
| 651 |
+
"step": 1040
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 2.67,
|
| 655 |
+
"learning_rate": 1.2606248685911856e-05,
|
| 656 |
+
"loss": 1.1392,
|
| 657 |
+
"step": 1050
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.7,
|
| 661 |
+
"learning_rate": 1.2470594140822008e-05,
|
| 662 |
+
"loss": 0.9999,
|
| 663 |
+
"step": 1060
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 2.72,
|
| 667 |
+
"learning_rate": 1.2334453638559057e-05,
|
| 668 |
+
"loss": 1.0181,
|
| 669 |
+
"step": 1070
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 2.75,
|
| 673 |
+
"learning_rate": 1.2197853957480689e-05,
|
| 674 |
+
"loss": 0.9913,
|
| 675 |
+
"step": 1080
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 2.77,
|
| 679 |
+
"learning_rate": 1.2060821966263465e-05,
|
| 680 |
+
"loss": 0.7443,
|
| 681 |
+
"step": 1090
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.8,
|
| 685 |
+
"learning_rate": 1.1923384618617819e-05,
|
| 686 |
+
"loss": 0.9462,
|
| 687 |
+
"step": 1100
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 2.82,
|
| 691 |
+
"learning_rate": 1.1785568947986368e-05,
|
| 692 |
+
"loss": 1.6673,
|
| 693 |
+
"step": 1110
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 2.85,
|
| 697 |
+
"learning_rate": 1.1647402062226534e-05,
|
| 698 |
+
"loss": 1.494,
|
| 699 |
+
"step": 1120
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 2.88,
|
| 703 |
+
"learning_rate": 1.1508911138278516e-05,
|
| 704 |
+
"loss": 1.3606,
|
| 705 |
+
"step": 1130
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 2.9,
|
| 709 |
+
"learning_rate": 1.1370123416819683e-05,
|
| 710 |
+
"loss": 0.5196,
|
| 711 |
+
"step": 1140
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 2.93,
|
| 715 |
+
"learning_rate": 1.123106619690643e-05,
|
| 716 |
+
"loss": 1.0671,
|
| 717 |
+
"step": 1150
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 2.95,
|
| 721 |
+
"learning_rate": 1.1091766830604552e-05,
|
| 722 |
+
"loss": 1.1892,
|
| 723 |
+
"step": 1160
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.98,
|
| 727 |
+
"learning_rate": 1.0952252717609188e-05,
|
| 728 |
+
"loss": 1.5041,
|
| 729 |
+
"step": 1170
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.0,
|
| 733 |
+
"eval_accuracy": 0.3597122302158273,
|
| 734 |
+
"eval_loss": 2.1702427864074707,
|
| 735 |
+
"eval_runtime": 14.7195,
|
| 736 |
+
"eval_samples_per_second": 9.443,
|
| 737 |
+
"eval_steps_per_second": 2.378,
|
| 738 |
+
"step": 1179
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 3.0,
|
| 742 |
+
"learning_rate": 1.0812551299855389e-05,
|
| 743 |
+
"loss": 1.1424,
|
| 744 |
+
"step": 1180
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.03,
|
| 748 |
+
"learning_rate": 1.0672690056120398e-05,
|
| 749 |
+
"loss": 0.9707,
|
| 750 |
+
"step": 1190
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.05,
|
| 754 |
+
"learning_rate": 1.0532696496618664e-05,
|
| 755 |
+
"loss": 0.6855,
|
| 756 |
+
"step": 1200
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 3.08,
|
| 760 |
+
"learning_rate": 1.0392598157590687e-05,
|
| 761 |
+
"loss": 0.362,
|
| 762 |
+
"step": 1210
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 3.1,
|
| 766 |
+
"learning_rate": 1.0252422595886723e-05,
|
| 767 |
+
"loss": 1.2138,
|
| 768 |
+
"step": 1220
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 3.13,
|
| 772 |
+
"learning_rate": 1.011219738354646e-05,
|
| 773 |
+
"loss": 0.4679,
|
| 774 |
+
"step": 1230
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 3.16,
|
| 778 |
+
"learning_rate": 9.971950102375696e-06,
|
| 779 |
+
"loss": 0.4326,
|
| 780 |
+
"step": 1240
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 3.18,
|
| 784 |
+
"learning_rate": 9.83170833852108e-06,
|
| 785 |
+
"loss": 0.4858,
|
| 786 |
+
"step": 1250
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.21,
|
| 790 |
+
"learning_rate": 9.691499677044032e-06,
|
| 791 |
+
"loss": 0.6772,
|
| 792 |
+
"step": 1260
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.23,
|
| 796 |
+
"learning_rate": 9.551351696494854e-06,
|
| 797 |
+
"loss": 0.5304,
|
| 798 |
+
"step": 1270
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 3.26,
|
| 802 |
+
"learning_rate": 9.41129196348811e-06,
|
| 803 |
+
"loss": 0.6597,
|
| 804 |
+
"step": 1280
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 3.28,
|
| 808 |
+
"learning_rate": 9.2713480272804e-06,
|
| 809 |
+
"loss": 0.8794,
|
| 810 |
+
"step": 1290
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 3.31,
|
| 814 |
+
"learning_rate": 9.131547414351486e-06,
|
| 815 |
+
"loss": 0.4008,
|
| 816 |
+
"step": 1300
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 3.33,
|
| 820 |
+
"learning_rate": 8.991917622989957e-06,
|
| 821 |
+
"loss": 0.9578,
|
| 822 |
+
"step": 1310
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 3.36,
|
| 826 |
+
"learning_rate": 8.852486117884391e-06,
|
| 827 |
+
"loss": 0.8909,
|
| 828 |
+
"step": 1320
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.38,
|
| 832 |
+
"learning_rate": 8.713280324721163e-06,
|
| 833 |
+
"loss": 0.7782,
|
| 834 |
+
"step": 1330
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.41,
|
| 838 |
+
"learning_rate": 8.574327624789904e-06,
|
| 839 |
+
"loss": 0.6601,
|
| 840 |
+
"step": 1340
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 3.44,
|
| 844 |
+
"learning_rate": 8.43565534959769e-06,
|
| 845 |
+
"loss": 0.597,
|
| 846 |
+
"step": 1350
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 3.46,
|
| 850 |
+
"learning_rate": 8.297290775493051e-06,
|
| 851 |
+
"loss": 0.6266,
|
| 852 |
+
"step": 1360
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 3.49,
|
| 856 |
+
"learning_rate": 8.159261118300784e-06,
|
| 857 |
+
"loss": 0.4856,
|
| 858 |
+
"step": 1370
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 3.51,
|
| 862 |
+
"learning_rate": 8.02159352796871e-06,
|
| 863 |
+
"loss": 0.5143,
|
| 864 |
+
"step": 1380
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 3.54,
|
| 868 |
+
"learning_rate": 7.884315083227373e-06,
|
| 869 |
+
"loss": 0.4511,
|
| 870 |
+
"step": 1390
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.56,
|
| 874 |
+
"learning_rate": 7.747452786263755e-06,
|
| 875 |
+
"loss": 0.7128,
|
| 876 |
+
"step": 1400
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.59,
|
| 880 |
+
"learning_rate": 7.611033557410027e-06,
|
| 881 |
+
"loss": 0.6028,
|
| 882 |
+
"step": 1410
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 3.61,
|
| 886 |
+
"learning_rate": 7.4750842298484205e-06,
|
| 887 |
+
"loss": 0.7882,
|
| 888 |
+
"step": 1420
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 3.64,
|
| 892 |
+
"learning_rate": 7.33963154433325e-06,
|
| 893 |
+
"loss": 0.6686,
|
| 894 |
+
"step": 1430
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 3.66,
|
| 898 |
+
"learning_rate": 7.204702143931072e-06,
|
| 899 |
+
"loss": 0.7159,
|
| 900 |
+
"step": 1440
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 3.69,
|
| 904 |
+
"learning_rate": 7.070322568780111e-06,
|
| 905 |
+
"loss": 1.4071,
|
| 906 |
+
"step": 1450
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 3.72,
|
| 910 |
+
"learning_rate": 6.9365192508698975e-06,
|
| 911 |
+
"loss": 0.9269,
|
| 912 |
+
"step": 1460
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.74,
|
| 916 |
+
"learning_rate": 6.8033185088421874e-06,
|
| 917 |
+
"loss": 1.0381,
|
| 918 |
+
"step": 1470
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 3.77,
|
| 922 |
+
"learning_rate": 6.670746542814187e-06,
|
| 923 |
+
"loss": 0.9117,
|
| 924 |
+
"step": 1480
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 3.79,
|
| 928 |
+
"learning_rate": 6.538829429225068e-06,
|
| 929 |
+
"loss": 0.6995,
|
| 930 |
+
"step": 1490
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 3.82,
|
| 934 |
+
"learning_rate": 6.407593115706849e-06,
|
| 935 |
+
"loss": 1.5987,
|
| 936 |
+
"step": 1500
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 3.84,
|
| 940 |
+
"learning_rate": 6.277063415980549e-06,
|
| 941 |
+
"loss": 0.7295,
|
| 942 |
+
"step": 1510
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 3.87,
|
| 946 |
+
"learning_rate": 6.147266004778755e-06,
|
| 947 |
+
"loss": 0.1654,
|
| 948 |
+
"step": 1520
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 3.89,
|
| 952 |
+
"learning_rate": 6.018226412795469e-06,
|
| 953 |
+
"loss": 0.7961,
|
| 954 |
+
"step": 1530
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.92,
|
| 958 |
+
"learning_rate": 5.889970021664329e-06,
|
| 959 |
+
"loss": 0.6862,
|
| 960 |
+
"step": 1540
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 3.94,
|
| 964 |
+
"learning_rate": 5.7625220589661136e-06,
|
| 965 |
+
"loss": 0.3667,
|
| 966 |
+
"step": 1550
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 3.97,
|
| 970 |
+
"learning_rate": 5.635907593266578e-06,
|
| 971 |
+
"loss": 0.8349,
|
| 972 |
+
"step": 1560
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 3.99,
|
| 976 |
+
"learning_rate": 5.510151529185554e-06,
|
| 977 |
+
"loss": 1.0457,
|
| 978 |
+
"step": 1570
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 4.0,
|
| 982 |
+
"eval_accuracy": 0.5683453237410072,
|
| 983 |
+
"eval_loss": 1.5413200855255127,
|
| 984 |
+
"eval_runtime": 14.7224,
|
| 985 |
+
"eval_samples_per_second": 9.441,
|
| 986 |
+
"eval_steps_per_second": 2.377,
|
| 987 |
+
"step": 1572
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 4.02,
|
| 991 |
+
"learning_rate": 5.385278602498288e-06,
|
| 992 |
+
"loss": 0.762,
|
| 993 |
+
"step": 1580
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 4.05,
|
| 997 |
+
"learning_rate": 5.2613133752700145e-06,
|
| 998 |
+
"loss": 0.4267,
|
| 999 |
+
"step": 1590
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 4.07,
|
| 1003 |
+
"learning_rate": 5.138280231024676e-06,
|
| 1004 |
+
"loss": 0.6985,
|
| 1005 |
+
"step": 1600
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 4.1,
|
| 1009 |
+
"learning_rate": 5.016203369948774e-06,
|
| 1010 |
+
"loss": 0.5179,
|
| 1011 |
+
"step": 1610
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 4.12,
|
| 1015 |
+
"learning_rate": 4.895106804131274e-06,
|
| 1016 |
+
"loss": 0.7994,
|
| 1017 |
+
"step": 1620
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.15,
|
| 1021 |
+
"learning_rate": 4.775014352840512e-06,
|
| 1022 |
+
"loss": 0.3598,
|
| 1023 |
+
"step": 1630
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 4.17,
|
| 1027 |
+
"learning_rate": 4.655949637839042e-06,
|
| 1028 |
+
"loss": 0.5081,
|
| 1029 |
+
"step": 1640
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 4.2,
|
| 1033 |
+
"learning_rate": 4.537936078737291e-06,
|
| 1034 |
+
"loss": 0.956,
|
| 1035 |
+
"step": 1650
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 4.22,
|
| 1039 |
+
"learning_rate": 4.420996888387033e-06,
|
| 1040 |
+
"loss": 1.2839,
|
| 1041 |
+
"step": 1660
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 4.25,
|
| 1045 |
+
"learning_rate": 4.305155068315481e-06,
|
| 1046 |
+
"loss": 0.6122,
|
| 1047 |
+
"step": 1670
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 4.27,
|
| 1051 |
+
"learning_rate": 4.190433404200977e-06,
|
| 1052 |
+
"loss": 1.3103,
|
| 1053 |
+
"step": 1680
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 4.3,
|
| 1057 |
+
"learning_rate": 4.07685446139113e-06,
|
| 1058 |
+
"loss": 0.6164,
|
| 1059 |
+
"step": 1690
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.33,
|
| 1063 |
+
"learning_rate": 3.964440580464286e-06,
|
| 1064 |
+
"loss": 0.9398,
|
| 1065 |
+
"step": 1700
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 4.35,
|
| 1069 |
+
"learning_rate": 3.853213872835229e-06,
|
| 1070 |
+
"loss": 0.1861,
|
| 1071 |
+
"step": 1710
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 4.38,
|
| 1075 |
+
"learning_rate": 3.743196216405922e-06,
|
| 1076 |
+
"loss": 0.543,
|
| 1077 |
+
"step": 1720
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 4.4,
|
| 1081 |
+
"learning_rate": 3.6344092512622173e-06,
|
| 1082 |
+
"loss": 0.6261,
|
| 1083 |
+
"step": 1730
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 4.43,
|
| 1087 |
+
"learning_rate": 3.5268743754173174e-06,
|
| 1088 |
+
"loss": 0.3409,
|
| 1089 |
+
"step": 1740
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 4.45,
|
| 1093 |
+
"learning_rate": 3.4206127406028744e-06,
|
| 1094 |
+
"loss": 0.4565,
|
| 1095 |
+
"step": 1750
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 4.48,
|
| 1099 |
+
"learning_rate": 3.3156452481085056e-06,
|
| 1100 |
+
"loss": 0.4853,
|
| 1101 |
+
"step": 1760
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.5,
|
| 1105 |
+
"learning_rate": 3.2119925446705824e-06,
|
| 1106 |
+
"loss": 0.3187,
|
| 1107 |
+
"step": 1770
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 4.53,
|
| 1111 |
+
"learning_rate": 3.109675018411108e-06,
|
| 1112 |
+
"loss": 0.5515,
|
| 1113 |
+
"step": 1780
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 4.55,
|
| 1117 |
+
"learning_rate": 3.0087127948274264e-06,
|
| 1118 |
+
"loss": 0.8492,
|
| 1119 |
+
"step": 1790
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 4.58,
|
| 1123 |
+
"learning_rate": 2.909125732833612e-06,
|
| 1124 |
+
"loss": 0.8423,
|
| 1125 |
+
"step": 1800
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 4.61,
|
| 1129 |
+
"learning_rate": 2.8109334208542994e-06,
|
| 1130 |
+
"loss": 0.7927,
|
| 1131 |
+
"step": 1810
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 4.63,
|
| 1135 |
+
"learning_rate": 2.7141551729716974e-06,
|
| 1136 |
+
"loss": 0.2295,
|
| 1137 |
+
"step": 1820
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 4.66,
|
| 1141 |
+
"learning_rate": 2.6188100251265947e-06,
|
| 1142 |
+
"loss": 0.6877,
|
| 1143 |
+
"step": 1830
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.68,
|
| 1147 |
+
"learning_rate": 2.5249167313740307e-06,
|
| 1148 |
+
"loss": 0.6192,
|
| 1149 |
+
"step": 1840
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 4.71,
|
| 1153 |
+
"learning_rate": 2.432493760194469e-06,
|
| 1154 |
+
"loss": 0.4079,
|
| 1155 |
+
"step": 1850
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 4.73,
|
| 1159 |
+
"learning_rate": 2.3415592908610874e-06,
|
| 1160 |
+
"loss": 0.1516,
|
| 1161 |
+
"step": 1860
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 4.76,
|
| 1165 |
+
"learning_rate": 2.2521312098639914e-06,
|
| 1166 |
+
"loss": 0.8265,
|
| 1167 |
+
"step": 1870
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 4.78,
|
| 1171 |
+
"learning_rate": 2.164227107392002e-06,
|
| 1172 |
+
"loss": 0.4464,
|
| 1173 |
+
"step": 1880
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 4.81,
|
| 1177 |
+
"learning_rate": 2.077864273872727e-06,
|
| 1178 |
+
"loss": 0.8132,
|
| 1179 |
+
"step": 1890
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 4.83,
|
| 1183 |
+
"learning_rate": 1.9930596965715985e-06,
|
| 1184 |
+
"loss": 0.6039,
|
| 1185 |
+
"step": 1900
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 4.86,
|
| 1189 |
+
"learning_rate": 1.9098300562505266e-06,
|
| 1190 |
+
"loss": 0.3869,
|
| 1191 |
+
"step": 1910
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 4.89,
|
| 1195 |
+
"learning_rate": 1.8281917238868707e-06,
|
| 1196 |
+
"loss": 0.2218,
|
| 1197 |
+
"step": 1920
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 4.91,
|
| 1201 |
+
"learning_rate": 1.7481607574533044e-06,
|
| 1202 |
+
"loss": 0.0394,
|
| 1203 |
+
"step": 1930
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 4.94,
|
| 1207 |
+
"learning_rate": 1.6697528987592893e-06,
|
| 1208 |
+
"loss": 0.5701,
|
| 1209 |
+
"step": 1940
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 4.96,
|
| 1213 |
+
"learning_rate": 1.5929835703546992e-06,
|
| 1214 |
+
"loss": 0.8708,
|
| 1215 |
+
"step": 1950
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 4.99,
|
| 1219 |
+
"learning_rate": 1.517867872496267e-06,
|
| 1220 |
+
"loss": 0.3315,
|
| 1221 |
+
"step": 1960
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 5.0,
|
| 1225 |
+
"eval_accuracy": 0.697841726618705,
|
| 1226 |
+
"eval_loss": 1.0769166946411133,
|
| 1227 |
+
"eval_runtime": 14.7066,
|
| 1228 |
+
"eval_samples_per_second": 9.452,
|
| 1229 |
+
"eval_steps_per_second": 2.38,
|
| 1230 |
+
"step": 1965
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 5.01,
|
| 1234 |
+
"learning_rate": 1.4444205801774204e-06,
|
| 1235 |
+
"loss": 0.8844,
|
| 1236 |
+
"step": 1970
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 5.04,
|
| 1240 |
+
"learning_rate": 1.3726561402220818e-06,
|
| 1241 |
+
"loss": 0.4769,
|
| 1242 |
+
"step": 1980
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 5.06,
|
| 1246 |
+
"learning_rate": 1.3025886684430467e-06,
|
| 1247 |
+
"loss": 0.4141,
|
| 1248 |
+
"step": 1990
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.09,
|
| 1252 |
+
"learning_rate": 1.2342319468654329e-06,
|
| 1253 |
+
"loss": 0.0528,
|
| 1254 |
+
"step": 2000
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 5.11,
|
| 1258 |
+
"learning_rate": 1.167599421015818e-06,
|
| 1259 |
+
"loss": 0.3157,
|
| 1260 |
+
"step": 2010
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 5.14,
|
| 1264 |
+
"learning_rate": 1.1027041972775442e-06,
|
| 1265 |
+
"loss": 0.2718,
|
| 1266 |
+
"step": 2020
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 5.17,
|
| 1270 |
+
"learning_rate": 1.0395590403127487e-06,
|
| 1271 |
+
"loss": 0.5592,
|
| 1272 |
+
"step": 2030
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 5.19,
|
| 1276 |
+
"learning_rate": 9.78176370551588e-07,
|
| 1277 |
+
"loss": 0.4728,
|
| 1278 |
+
"step": 2040
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 5.22,
|
| 1282 |
+
"learning_rate": 9.185682617491865e-07,
|
| 1283 |
+
"loss": 0.3589,
|
| 1284 |
+
"step": 2050
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 5.24,
|
| 1288 |
+
"learning_rate": 8.6074643861078e-07,
|
| 1289 |
+
"loss": 0.6848,
|
| 1290 |
+
"step": 2060
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.27,
|
| 1294 |
+
"learning_rate": 8.047222744854943e-07,
|
| 1295 |
+
"loss": 0.3668,
|
| 1296 |
+
"step": 2070
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 5.29,
|
| 1300 |
+
"learning_rate": 7.50506789129255e-07,
|
| 1301 |
+
"loss": 0.8621,
|
| 1302 |
+
"step": 2080
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 5.32,
|
| 1306 |
+
"learning_rate": 6.981106465372389e-07,
|
| 1307 |
+
"loss": 0.6104,
|
| 1308 |
+
"step": 2090
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 5.34,
|
| 1312 |
+
"learning_rate": 6.475441528463033e-07,
|
| 1313 |
+
"loss": 0.4872,
|
| 1314 |
+
"step": 2100
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 5.37,
|
| 1318 |
+
"learning_rate": 5.988172543078097e-07,
|
| 1319 |
+
"loss": 0.3606,
|
| 1320 |
+
"step": 2110
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 5.39,
|
| 1324 |
+
"learning_rate": 5.519395353312195e-07,
|
| 1325 |
+
"loss": 0.4107,
|
| 1326 |
+
"step": 2120
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 5.42,
|
| 1330 |
+
"learning_rate": 5.069202165988874e-07,
|
| 1331 |
+
"loss": 0.3428,
|
| 1332 |
+
"step": 2130
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.45,
|
| 1336 |
+
"learning_rate": 4.6376815325237966e-07,
|
| 1337 |
+
"loss": 0.6439,
|
| 1338 |
+
"step": 2140
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 5.47,
|
| 1342 |
+
"learning_rate": 4.224918331506955e-07,
|
| 1343 |
+
"loss": 0.2382,
|
| 1344 |
+
"step": 2150
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 5.5,
|
| 1348 |
+
"learning_rate": 3.8309937520074283e-07,
|
| 1349 |
+
"loss": 0.4123,
|
| 1350 |
+
"step": 2160
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 5.52,
|
| 1354 |
+
"learning_rate": 3.455985277603713e-07,
|
| 1355 |
+
"loss": 0.4865,
|
| 1356 |
+
"step": 2170
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 5.55,
|
| 1360 |
+
"learning_rate": 3.099966671143018e-07,
|
| 1361 |
+
"loss": 0.378,
|
| 1362 |
+
"step": 2180
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 5.57,
|
| 1366 |
+
"learning_rate": 2.7630079602323447e-07,
|
| 1367 |
+
"loss": 0.3095,
|
| 1368 |
+
"step": 2190
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 5.6,
|
| 1372 |
+
"learning_rate": 2.445175423464341e-07,
|
| 1373 |
+
"loss": 0.113,
|
| 1374 |
+
"step": 2200
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.62,
|
| 1378 |
+
"learning_rate": 2.1465315773804617e-07,
|
| 1379 |
+
"loss": 0.43,
|
| 1380 |
+
"step": 2210
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 5.65,
|
| 1384 |
+
"learning_rate": 1.867135164174283e-07,
|
| 1385 |
+
"loss": 0.5848,
|
| 1386 |
+
"step": 2220
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 5.67,
|
| 1390 |
+
"learning_rate": 1.6070411401370335e-07,
|
| 1391 |
+
"loss": 0.4236,
|
| 1392 |
+
"step": 2230
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 5.7,
|
| 1396 |
+
"learning_rate": 1.3663006648479415e-07,
|
| 1397 |
+
"loss": 0.3211,
|
| 1398 |
+
"step": 2240
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 5.73,
|
| 1402 |
+
"learning_rate": 1.1449610911112741e-07,
|
| 1403 |
+
"loss": 0.4408,
|
| 1404 |
+
"step": 2250
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 5.75,
|
| 1408 |
+
"learning_rate": 9.43065955642275e-08,
|
| 1409 |
+
"loss": 0.1455,
|
| 1410 |
+
"step": 2260
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 5.78,
|
| 1414 |
+
"learning_rate": 7.606549705035937e-08,
|
| 1415 |
+
"loss": 0.055,
|
| 1416 |
+
"step": 2270
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 5.8,
|
| 1420 |
+
"learning_rate": 5.977640152940556e-08,
|
| 1421 |
+
"loss": 0.8437,
|
| 1422 |
+
"step": 2280
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 5.83,
|
| 1426 |
+
"learning_rate": 4.544251300913405e-08,
|
| 1427 |
+
"loss": 0.7393,
|
| 1428 |
+
"step": 2290
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 5.85,
|
| 1432 |
+
"learning_rate": 3.3066650914970186e-08,
|
| 1433 |
+
"loss": 0.0791,
|
| 1434 |
+
"step": 2300
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 5.88,
|
| 1438 |
+
"learning_rate": 2.265124953543918e-08,
|
| 1439 |
+
"loss": 0.2199,
|
| 1440 |
+
"step": 2310
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 5.9,
|
| 1444 |
+
"learning_rate": 1.4198357543336871e-08,
|
| 1445 |
+
"loss": 0.4947,
|
| 1446 |
+
"step": 2320
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 5.93,
|
| 1450 |
+
"learning_rate": 7.70963759277099e-09,
|
| 1451 |
+
"loss": 0.2422,
|
| 1452 |
+
"step": 2330
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 5.95,
|
| 1456 |
+
"learning_rate": 3.1863659921216316e-09,
|
| 1457 |
+
"loss": 0.7831,
|
| 1458 |
+
"step": 2340
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 5.98,
|
| 1462 |
+
"learning_rate": 6.294324529942942e-10,
|
| 1463 |
+
"loss": 0.2178,
|
| 1464 |
+
"step": 2350
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 6.0,
|
| 1468 |
+
"eval_accuracy": 0.6762589928057554,
|
| 1469 |
+
"eval_loss": 1.233504295349121,
|
| 1470 |
+
"eval_runtime": 14.7199,
|
| 1471 |
+
"eval_samples_per_second": 9.443,
|
| 1472 |
+
"eval_steps_per_second": 2.378,
|
| 1473 |
+
"step": 2358
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 6.0,
|
| 1477 |
+
"step": 2358,
|
| 1478 |
+
"total_flos": 6.779516273836425e+18,
|
| 1479 |
+
"train_loss": 0.9992087238718637,
|
| 1480 |
+
"train_runtime": 3103.0559,
|
| 1481 |
+
"train_samples_per_second": 1.52,
|
| 1482 |
+
"train_steps_per_second": 0.76
|
| 1483 |
+
}
|
| 1484 |
+
],
|
| 1485 |
+
"logging_steps": 10,
|
| 1486 |
+
"max_steps": 2358,
|
| 1487 |
+
"num_input_tokens_seen": 0,
|
| 1488 |
+
"num_train_epochs": 6,
|
| 1489 |
+
"save_steps": 500,
|
| 1490 |
+
"total_flos": 6.779516273836425e+18,
|
| 1491 |
+
"train_batch_size": 2,
|
| 1492 |
+
"trial_name": null,
|
| 1493 |
+
"trial_params": null
|
| 1494 |
+
}
|