cheers again
Browse files- README.md +4 -3
- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +1686 -0
README.md
CHANGED
|
@@ -2,6 +2,7 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224-in21k
|
| 4 |
tags:
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
|
@@ -15,10 +16,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# test-hasy-5
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224-in21k
|
| 4 |
tags:
|
| 5 |
+
- image-classification
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
|
|
|
| 16 |
|
| 17 |
# test-hasy-5
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the HASY dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.6861
|
| 22 |
+
- Accuracy: 0.8067
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"eval_accuracy": 0.8066528066528067,
|
| 4 |
+
"eval_loss": 0.6860660910606384,
|
| 5 |
+
"eval_runtime": 2.0046,
|
| 6 |
+
"eval_samples_per_second": 239.949,
|
| 7 |
+
"eval_steps_per_second": 30.43,
|
| 8 |
+
"total_flos": 3.355193271048192e+19,
|
| 9 |
+
"train_loss": 1.2517024893769495,
|
| 10 |
+
"train_runtime": 5380.7558,
|
| 11 |
+
"train_samples_per_second": 80.379,
|
| 12 |
+
"train_steps_per_second": 10.054
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"eval_accuracy": 0.8066528066528067,
|
| 4 |
+
"eval_loss": 0.6860660910606384,
|
| 5 |
+
"eval_runtime": 2.0046,
|
| 6 |
+
"eval_samples_per_second": 239.949,
|
| 7 |
+
"eval_steps_per_second": 30.43
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"total_flos": 3.355193271048192e+19,
|
| 4 |
+
"train_loss": 1.2517024893769495,
|
| 5 |
+
"train_runtime": 5380.7558,
|
| 6 |
+
"train_samples_per_second": 80.379,
|
| 7 |
+
"train_steps_per_second": 10.054
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1686 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8066528066528067,
|
| 3 |
+
"best_model_checkpoint": "test-hasy-5/checkpoint-18935",
|
| 4 |
+
"epoch": 100.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 54100,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.92,
|
| 13 |
+
"grad_norm": 4.6960930824279785,
|
| 14 |
+
"learning_rate": 1.9815157116451017e-05,
|
| 15 |
+
"loss": 3.9645,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.0,
|
| 20 |
+
"eval_accuracy": 0.3970893970893971,
|
| 21 |
+
"eval_loss": 3.429507255554199,
|
| 22 |
+
"eval_runtime": 2.0488,
|
| 23 |
+
"eval_samples_per_second": 234.77,
|
| 24 |
+
"eval_steps_per_second": 29.773,
|
| 25 |
+
"step": 541
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 1.85,
|
| 29 |
+
"grad_norm": 3.364806652069092,
|
| 30 |
+
"learning_rate": 1.9630314232902035e-05,
|
| 31 |
+
"loss": 3.4258,
|
| 32 |
+
"step": 1000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.0,
|
| 36 |
+
"eval_accuracy": 0.4781704781704782,
|
| 37 |
+
"eval_loss": 2.879011392593384,
|
| 38 |
+
"eval_runtime": 1.9708,
|
| 39 |
+
"eval_samples_per_second": 244.059,
|
| 40 |
+
"eval_steps_per_second": 30.951,
|
| 41 |
+
"step": 1082
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 2.77,
|
| 45 |
+
"grad_norm": 3.35432767868042,
|
| 46 |
+
"learning_rate": 1.944547134935305e-05,
|
| 47 |
+
"loss": 3.04,
|
| 48 |
+
"step": 1500
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 3.0,
|
| 52 |
+
"eval_accuracy": 0.5467775467775468,
|
| 53 |
+
"eval_loss": 2.4893012046813965,
|
| 54 |
+
"eval_runtime": 2.0175,
|
| 55 |
+
"eval_samples_per_second": 238.411,
|
| 56 |
+
"eval_steps_per_second": 30.235,
|
| 57 |
+
"step": 1623
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 3.7,
|
| 61 |
+
"grad_norm": 4.459615230560303,
|
| 62 |
+
"learning_rate": 1.9260628465804068e-05,
|
| 63 |
+
"loss": 2.793,
|
| 64 |
+
"step": 2000
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 4.0,
|
| 68 |
+
"eval_accuracy": 0.5738045738045738,
|
| 69 |
+
"eval_loss": 2.2005958557128906,
|
| 70 |
+
"eval_runtime": 1.962,
|
| 71 |
+
"eval_samples_per_second": 245.161,
|
| 72 |
+
"eval_steps_per_second": 31.091,
|
| 73 |
+
"step": 2164
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 4.62,
|
| 77 |
+
"grad_norm": 4.481846332550049,
|
| 78 |
+
"learning_rate": 1.9075785582255083e-05,
|
| 79 |
+
"loss": 2.5551,
|
| 80 |
+
"step": 2500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 5.0,
|
| 84 |
+
"eval_accuracy": 0.6340956340956341,
|
| 85 |
+
"eval_loss": 1.9055824279785156,
|
| 86 |
+
"eval_runtime": 2.0221,
|
| 87 |
+
"eval_samples_per_second": 237.875,
|
| 88 |
+
"eval_steps_per_second": 30.167,
|
| 89 |
+
"step": 2705
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 5.55,
|
| 93 |
+
"grad_norm": 4.1784281730651855,
|
| 94 |
+
"learning_rate": 1.88909426987061e-05,
|
| 95 |
+
"loss": 2.3662,
|
| 96 |
+
"step": 3000
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 6.0,
|
| 100 |
+
"eval_accuracy": 0.6632016632016632,
|
| 101 |
+
"eval_loss": 1.7023240327835083,
|
| 102 |
+
"eval_runtime": 1.9666,
|
| 103 |
+
"eval_samples_per_second": 244.588,
|
| 104 |
+
"eval_steps_per_second": 31.018,
|
| 105 |
+
"step": 3246
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 6.47,
|
| 109 |
+
"grad_norm": 6.06294059753418,
|
| 110 |
+
"learning_rate": 1.8706099815157116e-05,
|
| 111 |
+
"loss": 2.1965,
|
| 112 |
+
"step": 3500
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 7.0,
|
| 116 |
+
"eval_accuracy": 0.6798336798336798,
|
| 117 |
+
"eval_loss": 1.5739575624465942,
|
| 118 |
+
"eval_runtime": 1.9941,
|
| 119 |
+
"eval_samples_per_second": 241.216,
|
| 120 |
+
"eval_steps_per_second": 30.591,
|
| 121 |
+
"step": 3787
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 7.39,
|
| 125 |
+
"grad_norm": 4.912960052490234,
|
| 126 |
+
"learning_rate": 1.8521256931608135e-05,
|
| 127 |
+
"loss": 2.1397,
|
| 128 |
+
"step": 4000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.0,
|
| 132 |
+
"eval_accuracy": 0.6943866943866944,
|
| 133 |
+
"eval_loss": 1.4560521841049194,
|
| 134 |
+
"eval_runtime": 1.9835,
|
| 135 |
+
"eval_samples_per_second": 242.507,
|
| 136 |
+
"eval_steps_per_second": 30.754,
|
| 137 |
+
"step": 4328
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 8.32,
|
| 141 |
+
"grad_norm": 5.236889362335205,
|
| 142 |
+
"learning_rate": 1.833641404805915e-05,
|
| 143 |
+
"loss": 1.9955,
|
| 144 |
+
"step": 4500
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 9.0,
|
| 148 |
+
"eval_accuracy": 0.7234927234927235,
|
| 149 |
+
"eval_loss": 1.3202540874481201,
|
| 150 |
+
"eval_runtime": 2.0536,
|
| 151 |
+
"eval_samples_per_second": 234.218,
|
| 152 |
+
"eval_steps_per_second": 29.703,
|
| 153 |
+
"step": 4869
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 9.24,
|
| 157 |
+
"grad_norm": 5.675503253936768,
|
| 158 |
+
"learning_rate": 1.8151571164510168e-05,
|
| 159 |
+
"loss": 1.9282,
|
| 160 |
+
"step": 5000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 10.0,
|
| 164 |
+
"eval_accuracy": 0.738045738045738,
|
| 165 |
+
"eval_loss": 1.2246184349060059,
|
| 166 |
+
"eval_runtime": 2.0017,
|
| 167 |
+
"eval_samples_per_second": 240.293,
|
| 168 |
+
"eval_steps_per_second": 30.474,
|
| 169 |
+
"step": 5410
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 10.17,
|
| 173 |
+
"grad_norm": 4.67825174331665,
|
| 174 |
+
"learning_rate": 1.7966728280961186e-05,
|
| 175 |
+
"loss": 1.8368,
|
| 176 |
+
"step": 5500
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 11.0,
|
| 180 |
+
"eval_accuracy": 0.738045738045738,
|
| 181 |
+
"eval_loss": 1.1823257207870483,
|
| 182 |
+
"eval_runtime": 1.9774,
|
| 183 |
+
"eval_samples_per_second": 243.246,
|
| 184 |
+
"eval_steps_per_second": 30.848,
|
| 185 |
+
"step": 5951
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 11.09,
|
| 189 |
+
"grad_norm": 4.809859275817871,
|
| 190 |
+
"learning_rate": 1.77818853974122e-05,
|
| 191 |
+
"loss": 1.812,
|
| 192 |
+
"step": 6000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 12.0,
|
| 196 |
+
"eval_accuracy": 0.7214137214137214,
|
| 197 |
+
"eval_loss": 1.1297953128814697,
|
| 198 |
+
"eval_runtime": 2.0307,
|
| 199 |
+
"eval_samples_per_second": 236.864,
|
| 200 |
+
"eval_steps_per_second": 30.039,
|
| 201 |
+
"step": 6492
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 12.01,
|
| 205 |
+
"grad_norm": 5.255190849304199,
|
| 206 |
+
"learning_rate": 1.759704251386322e-05,
|
| 207 |
+
"loss": 1.7353,
|
| 208 |
+
"step": 6500
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 12.94,
|
| 212 |
+
"grad_norm": 8.597217559814453,
|
| 213 |
+
"learning_rate": 1.7412199630314234e-05,
|
| 214 |
+
"loss": 1.7195,
|
| 215 |
+
"step": 7000
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 13.0,
|
| 219 |
+
"eval_accuracy": 0.7484407484407485,
|
| 220 |
+
"eval_loss": 1.0423070192337036,
|
| 221 |
+
"eval_runtime": 2.0193,
|
| 222 |
+
"eval_samples_per_second": 238.201,
|
| 223 |
+
"eval_steps_per_second": 30.208,
|
| 224 |
+
"step": 7033
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 13.86,
|
| 228 |
+
"grad_norm": 6.453842639923096,
|
| 229 |
+
"learning_rate": 1.7227356746765253e-05,
|
| 230 |
+
"loss": 1.6314,
|
| 231 |
+
"step": 7500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 14.0,
|
| 235 |
+
"eval_accuracy": 0.7422037422037422,
|
| 236 |
+
"eval_loss": 1.0077309608459473,
|
| 237 |
+
"eval_runtime": 2.0783,
|
| 238 |
+
"eval_samples_per_second": 231.439,
|
| 239 |
+
"eval_steps_per_second": 29.351,
|
| 240 |
+
"step": 7574
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 14.79,
|
| 244 |
+
"grad_norm": 8.70645523071289,
|
| 245 |
+
"learning_rate": 1.7042513863216268e-05,
|
| 246 |
+
"loss": 1.5979,
|
| 247 |
+
"step": 8000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 15.0,
|
| 251 |
+
"eval_accuracy": 0.7463617463617463,
|
| 252 |
+
"eval_loss": 1.00509512424469,
|
| 253 |
+
"eval_runtime": 1.9889,
|
| 254 |
+
"eval_samples_per_second": 241.847,
|
| 255 |
+
"eval_steps_per_second": 30.671,
|
| 256 |
+
"step": 8115
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 15.71,
|
| 260 |
+
"grad_norm": 7.348147392272949,
|
| 261 |
+
"learning_rate": 1.6857670979667286e-05,
|
| 262 |
+
"loss": 1.5656,
|
| 263 |
+
"step": 8500
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 16.0,
|
| 267 |
+
"eval_accuracy": 0.7692307692307693,
|
| 268 |
+
"eval_loss": 0.9325113296508789,
|
| 269 |
+
"eval_runtime": 1.9923,
|
| 270 |
+
"eval_samples_per_second": 241.43,
|
| 271 |
+
"eval_steps_per_second": 30.618,
|
| 272 |
+
"step": 8656
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 16.64,
|
| 276 |
+
"grad_norm": 6.420931816101074,
|
| 277 |
+
"learning_rate": 1.66728280961183e-05,
|
| 278 |
+
"loss": 1.5414,
|
| 279 |
+
"step": 9000
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 17.0,
|
| 283 |
+
"eval_accuracy": 0.7733887733887734,
|
| 284 |
+
"eval_loss": 0.8889437913894653,
|
| 285 |
+
"eval_runtime": 1.9727,
|
| 286 |
+
"eval_samples_per_second": 243.822,
|
| 287 |
+
"eval_steps_per_second": 30.921,
|
| 288 |
+
"step": 9197
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 17.56,
|
| 292 |
+
"grad_norm": 8.127350807189941,
|
| 293 |
+
"learning_rate": 1.6487985212569316e-05,
|
| 294 |
+
"loss": 1.5342,
|
| 295 |
+
"step": 9500
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 18.0,
|
| 299 |
+
"eval_accuracy": 0.7484407484407485,
|
| 300 |
+
"eval_loss": 0.9072721600532532,
|
| 301 |
+
"eval_runtime": 1.9933,
|
| 302 |
+
"eval_samples_per_second": 241.303,
|
| 303 |
+
"eval_steps_per_second": 30.602,
|
| 304 |
+
"step": 9738
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 18.48,
|
| 308 |
+
"grad_norm": 6.122061252593994,
|
| 309 |
+
"learning_rate": 1.6303142329020334e-05,
|
| 310 |
+
"loss": 1.4898,
|
| 311 |
+
"step": 10000
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 19.0,
|
| 315 |
+
"eval_accuracy": 0.7713097713097713,
|
| 316 |
+
"eval_loss": 0.8425627946853638,
|
| 317 |
+
"eval_runtime": 1.9868,
|
| 318 |
+
"eval_samples_per_second": 242.099,
|
| 319 |
+
"eval_steps_per_second": 30.703,
|
| 320 |
+
"step": 10279
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 19.41,
|
| 324 |
+
"grad_norm": 6.640945911407471,
|
| 325 |
+
"learning_rate": 1.611829944547135e-05,
|
| 326 |
+
"loss": 1.4731,
|
| 327 |
+
"step": 10500
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 20.0,
|
| 331 |
+
"eval_accuracy": 0.7442827442827443,
|
| 332 |
+
"eval_loss": 0.862506091594696,
|
| 333 |
+
"eval_runtime": 1.9786,
|
| 334 |
+
"eval_samples_per_second": 243.096,
|
| 335 |
+
"eval_steps_per_second": 30.829,
|
| 336 |
+
"step": 10820
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 20.33,
|
| 340 |
+
"grad_norm": 6.019400119781494,
|
| 341 |
+
"learning_rate": 1.5933456561922367e-05,
|
| 342 |
+
"loss": 1.451,
|
| 343 |
+
"step": 11000
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 21.0,
|
| 347 |
+
"eval_accuracy": 0.762993762993763,
|
| 348 |
+
"eval_loss": 0.8015209436416626,
|
| 349 |
+
"eval_runtime": 1.9644,
|
| 350 |
+
"eval_samples_per_second": 244.864,
|
| 351 |
+
"eval_steps_per_second": 31.053,
|
| 352 |
+
"step": 11361
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 21.26,
|
| 356 |
+
"grad_norm": 5.140503406524658,
|
| 357 |
+
"learning_rate": 1.5748613678373382e-05,
|
| 358 |
+
"loss": 1.4578,
|
| 359 |
+
"step": 11500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 22.0,
|
| 363 |
+
"eval_accuracy": 0.7588357588357588,
|
| 364 |
+
"eval_loss": 0.8520306944847107,
|
| 365 |
+
"eval_runtime": 2.0001,
|
| 366 |
+
"eval_samples_per_second": 240.484,
|
| 367 |
+
"eval_steps_per_second": 30.498,
|
| 368 |
+
"step": 11902
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 22.18,
|
| 372 |
+
"grad_norm": 15.190984725952148,
|
| 373 |
+
"learning_rate": 1.55637707948244e-05,
|
| 374 |
+
"loss": 1.4126,
|
| 375 |
+
"step": 12000
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 23.0,
|
| 379 |
+
"eval_accuracy": 0.7713097713097713,
|
| 380 |
+
"eval_loss": 0.7928301692008972,
|
| 381 |
+
"eval_runtime": 1.9822,
|
| 382 |
+
"eval_samples_per_second": 242.66,
|
| 383 |
+
"eval_steps_per_second": 30.774,
|
| 384 |
+
"step": 12443
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 23.11,
|
| 388 |
+
"grad_norm": 11.220525741577148,
|
| 389 |
+
"learning_rate": 1.5378927911275416e-05,
|
| 390 |
+
"loss": 1.3626,
|
| 391 |
+
"step": 12500
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 24.0,
|
| 395 |
+
"eval_accuracy": 0.7837837837837838,
|
| 396 |
+
"eval_loss": 0.754388689994812,
|
| 397 |
+
"eval_runtime": 1.975,
|
| 398 |
+
"eval_samples_per_second": 243.545,
|
| 399 |
+
"eval_steps_per_second": 30.886,
|
| 400 |
+
"step": 12984
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 24.03,
|
| 404 |
+
"grad_norm": 3.5185582637786865,
|
| 405 |
+
"learning_rate": 1.5194085027726432e-05,
|
| 406 |
+
"loss": 1.3905,
|
| 407 |
+
"step": 13000
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 24.95,
|
| 411 |
+
"grad_norm": 8.19352912902832,
|
| 412 |
+
"learning_rate": 1.5009242144177449e-05,
|
| 413 |
+
"loss": 1.3694,
|
| 414 |
+
"step": 13500
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 25.0,
|
| 418 |
+
"eval_accuracy": 0.7775467775467776,
|
| 419 |
+
"eval_loss": 0.7698755860328674,
|
| 420 |
+
"eval_runtime": 2.0179,
|
| 421 |
+
"eval_samples_per_second": 238.368,
|
| 422 |
+
"eval_steps_per_second": 30.23,
|
| 423 |
+
"step": 13525
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 25.88,
|
| 427 |
+
"grad_norm": 6.003907680511475,
|
| 428 |
+
"learning_rate": 1.4824399260628467e-05,
|
| 429 |
+
"loss": 1.3612,
|
| 430 |
+
"step": 14000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 26.0,
|
| 434 |
+
"eval_accuracy": 0.7775467775467776,
|
| 435 |
+
"eval_loss": 0.7602183818817139,
|
| 436 |
+
"eval_runtime": 1.9833,
|
| 437 |
+
"eval_samples_per_second": 242.521,
|
| 438 |
+
"eval_steps_per_second": 30.756,
|
| 439 |
+
"step": 14066
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 26.8,
|
| 443 |
+
"grad_norm": 6.613931655883789,
|
| 444 |
+
"learning_rate": 1.4639556377079484e-05,
|
| 445 |
+
"loss": 1.2963,
|
| 446 |
+
"step": 14500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 27.0,
|
| 450 |
+
"eval_accuracy": 0.7713097713097713,
|
| 451 |
+
"eval_loss": 0.7532169818878174,
|
| 452 |
+
"eval_runtime": 2.0706,
|
| 453 |
+
"eval_samples_per_second": 232.305,
|
| 454 |
+
"eval_steps_per_second": 29.461,
|
| 455 |
+
"step": 14607
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 27.73,
|
| 459 |
+
"grad_norm": 7.66683292388916,
|
| 460 |
+
"learning_rate": 1.44547134935305e-05,
|
| 461 |
+
"loss": 1.3009,
|
| 462 |
+
"step": 15000
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 28.0,
|
| 466 |
+
"eval_accuracy": 0.7920997920997921,
|
| 467 |
+
"eval_loss": 0.7012535929679871,
|
| 468 |
+
"eval_runtime": 1.9606,
|
| 469 |
+
"eval_samples_per_second": 245.335,
|
| 470 |
+
"eval_steps_per_second": 31.113,
|
| 471 |
+
"step": 15148
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 28.65,
|
| 475 |
+
"grad_norm": 7.342077255249023,
|
| 476 |
+
"learning_rate": 1.4269870609981517e-05,
|
| 477 |
+
"loss": 1.2598,
|
| 478 |
+
"step": 15500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 29.0,
|
| 482 |
+
"eval_accuracy": 0.7796257796257796,
|
| 483 |
+
"eval_loss": 0.7084705233573914,
|
| 484 |
+
"eval_runtime": 1.9824,
|
| 485 |
+
"eval_samples_per_second": 242.632,
|
| 486 |
+
"eval_steps_per_second": 30.77,
|
| 487 |
+
"step": 15689
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 29.57,
|
| 491 |
+
"grad_norm": 5.679790019989014,
|
| 492 |
+
"learning_rate": 1.4085027726432534e-05,
|
| 493 |
+
"loss": 1.2565,
|
| 494 |
+
"step": 16000
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 30.0,
|
| 498 |
+
"eval_accuracy": 0.7775467775467776,
|
| 499 |
+
"eval_loss": 0.7023281455039978,
|
| 500 |
+
"eval_runtime": 1.9659,
|
| 501 |
+
"eval_samples_per_second": 244.668,
|
| 502 |
+
"eval_steps_per_second": 31.029,
|
| 503 |
+
"step": 16230
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 30.5,
|
| 507 |
+
"grad_norm": 5.493412971496582,
|
| 508 |
+
"learning_rate": 1.390018484288355e-05,
|
| 509 |
+
"loss": 1.2735,
|
| 510 |
+
"step": 16500
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 31.0,
|
| 514 |
+
"eval_accuracy": 0.7775467775467776,
|
| 515 |
+
"eval_loss": 0.7047860026359558,
|
| 516 |
+
"eval_runtime": 1.9718,
|
| 517 |
+
"eval_samples_per_second": 243.937,
|
| 518 |
+
"eval_steps_per_second": 30.936,
|
| 519 |
+
"step": 16771
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 31.42,
|
| 523 |
+
"grad_norm": 6.2688093185424805,
|
| 524 |
+
"learning_rate": 1.3715341959334567e-05,
|
| 525 |
+
"loss": 1.2743,
|
| 526 |
+
"step": 17000
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 32.0,
|
| 530 |
+
"eval_accuracy": 0.7920997920997921,
|
| 531 |
+
"eval_loss": 0.6794067621231079,
|
| 532 |
+
"eval_runtime": 1.9764,
|
| 533 |
+
"eval_samples_per_second": 243.372,
|
| 534 |
+
"eval_steps_per_second": 30.864,
|
| 535 |
+
"step": 17312
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 32.35,
|
| 539 |
+
"grad_norm": 10.169917106628418,
|
| 540 |
+
"learning_rate": 1.3530499075785584e-05,
|
| 541 |
+
"loss": 1.2441,
|
| 542 |
+
"step": 17500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 33.0,
|
| 546 |
+
"eval_accuracy": 0.7858627858627859,
|
| 547 |
+
"eval_loss": 0.693196713924408,
|
| 548 |
+
"eval_runtime": 1.972,
|
| 549 |
+
"eval_samples_per_second": 243.92,
|
| 550 |
+
"eval_steps_per_second": 30.934,
|
| 551 |
+
"step": 17853
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 33.27,
|
| 555 |
+
"grad_norm": 8.05045223236084,
|
| 556 |
+
"learning_rate": 1.33456561922366e-05,
|
| 557 |
+
"loss": 1.2282,
|
| 558 |
+
"step": 18000
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 34.0,
|
| 562 |
+
"eval_accuracy": 0.7941787941787942,
|
| 563 |
+
"eval_loss": 0.7038751840591431,
|
| 564 |
+
"eval_runtime": 1.992,
|
| 565 |
+
"eval_samples_per_second": 241.466,
|
| 566 |
+
"eval_steps_per_second": 30.623,
|
| 567 |
+
"step": 18394
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 34.2,
|
| 571 |
+
"grad_norm": 5.410665035247803,
|
| 572 |
+
"learning_rate": 1.3160813308687617e-05,
|
| 573 |
+
"loss": 1.2204,
|
| 574 |
+
"step": 18500
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 35.0,
|
| 578 |
+
"eval_accuracy": 0.8066528066528067,
|
| 579 |
+
"eval_loss": 0.6860660910606384,
|
| 580 |
+
"eval_runtime": 1.9773,
|
| 581 |
+
"eval_samples_per_second": 243.258,
|
| 582 |
+
"eval_steps_per_second": 30.85,
|
| 583 |
+
"step": 18935
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 35.12,
|
| 587 |
+
"grad_norm": 11.123208045959473,
|
| 588 |
+
"learning_rate": 1.2975970425138634e-05,
|
| 589 |
+
"loss": 1.1808,
|
| 590 |
+
"step": 19000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 36.0,
|
| 594 |
+
"eval_accuracy": 0.7962577962577962,
|
| 595 |
+
"eval_loss": 0.6589930057525635,
|
| 596 |
+
"eval_runtime": 1.9769,
|
| 597 |
+
"eval_samples_per_second": 243.305,
|
| 598 |
+
"eval_steps_per_second": 30.856,
|
| 599 |
+
"step": 19476
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 36.04,
|
| 603 |
+
"grad_norm": 6.165465354919434,
|
| 604 |
+
"learning_rate": 1.279112754158965e-05,
|
| 605 |
+
"loss": 1.1933,
|
| 606 |
+
"step": 19500
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 36.97,
|
| 610 |
+
"grad_norm": 6.407803535461426,
|
| 611 |
+
"learning_rate": 1.2606284658040667e-05,
|
| 612 |
+
"loss": 1.1928,
|
| 613 |
+
"step": 20000
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 37.0,
|
| 617 |
+
"eval_accuracy": 0.7817047817047817,
|
| 618 |
+
"eval_loss": 0.678415834903717,
|
| 619 |
+
"eval_runtime": 1.9844,
|
| 620 |
+
"eval_samples_per_second": 242.388,
|
| 621 |
+
"eval_steps_per_second": 30.739,
|
| 622 |
+
"step": 20017
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 37.89,
|
| 626 |
+
"grad_norm": 4.849668979644775,
|
| 627 |
+
"learning_rate": 1.2421441774491683e-05,
|
| 628 |
+
"loss": 1.1914,
|
| 629 |
+
"step": 20500
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 38.0,
|
| 633 |
+
"eval_accuracy": 0.7962577962577962,
|
| 634 |
+
"eval_loss": 0.6559053659439087,
|
| 635 |
+
"eval_runtime": 1.9912,
|
| 636 |
+
"eval_samples_per_second": 241.56,
|
| 637 |
+
"eval_steps_per_second": 30.634,
|
| 638 |
+
"step": 20558
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 38.82,
|
| 642 |
+
"grad_norm": 9.1309232711792,
|
| 643 |
+
"learning_rate": 1.2236598890942698e-05,
|
| 644 |
+
"loss": 1.1856,
|
| 645 |
+
"step": 21000
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 39.0,
|
| 649 |
+
"eval_accuracy": 0.7962577962577962,
|
| 650 |
+
"eval_loss": 0.6769025325775146,
|
| 651 |
+
"eval_runtime": 2.0066,
|
| 652 |
+
"eval_samples_per_second": 239.713,
|
| 653 |
+
"eval_steps_per_second": 30.4,
|
| 654 |
+
"step": 21099
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 39.74,
|
| 658 |
+
"grad_norm": 5.001546382904053,
|
| 659 |
+
"learning_rate": 1.2051756007393715e-05,
|
| 660 |
+
"loss": 1.1585,
|
| 661 |
+
"step": 21500
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 40.0,
|
| 665 |
+
"eval_accuracy": 0.8004158004158004,
|
| 666 |
+
"eval_loss": 0.64976966381073,
|
| 667 |
+
"eval_runtime": 1.9804,
|
| 668 |
+
"eval_samples_per_second": 242.874,
|
| 669 |
+
"eval_steps_per_second": 30.801,
|
| 670 |
+
"step": 21640
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 40.67,
|
| 674 |
+
"grad_norm": 14.044866561889648,
|
| 675 |
+
"learning_rate": 1.1866913123844732e-05,
|
| 676 |
+
"loss": 1.1713,
|
| 677 |
+
"step": 22000
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 41.0,
|
| 681 |
+
"eval_accuracy": 0.7920997920997921,
|
| 682 |
+
"eval_loss": 0.6447434425354004,
|
| 683 |
+
"eval_runtime": 1.9973,
|
| 684 |
+
"eval_samples_per_second": 240.829,
|
| 685 |
+
"eval_steps_per_second": 30.542,
|
| 686 |
+
"step": 22181
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 41.59,
|
| 690 |
+
"grad_norm": 10.289350509643555,
|
| 691 |
+
"learning_rate": 1.1682070240295748e-05,
|
| 692 |
+
"loss": 1.1183,
|
| 693 |
+
"step": 22500
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 42.0,
|
| 697 |
+
"eval_accuracy": 0.7713097713097713,
|
| 698 |
+
"eval_loss": 0.6748064756393433,
|
| 699 |
+
"eval_runtime": 1.9672,
|
| 700 |
+
"eval_samples_per_second": 244.509,
|
| 701 |
+
"eval_steps_per_second": 31.008,
|
| 702 |
+
"step": 22722
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 42.51,
|
| 706 |
+
"grad_norm": 12.7116117477417,
|
| 707 |
+
"learning_rate": 1.1497227356746765e-05,
|
| 708 |
+
"loss": 1.1564,
|
| 709 |
+
"step": 23000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 43.0,
|
| 713 |
+
"eval_accuracy": 0.7920997920997921,
|
| 714 |
+
"eval_loss": 0.6545261740684509,
|
| 715 |
+
"eval_runtime": 1.9659,
|
| 716 |
+
"eval_samples_per_second": 244.669,
|
| 717 |
+
"eval_steps_per_second": 31.029,
|
| 718 |
+
"step": 23263
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 43.44,
|
| 722 |
+
"grad_norm": 3.0720624923706055,
|
| 723 |
+
"learning_rate": 1.1312384473197783e-05,
|
| 724 |
+
"loss": 1.1215,
|
| 725 |
+
"step": 23500
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 44.0,
|
| 729 |
+
"eval_accuracy": 0.7879417879417879,
|
| 730 |
+
"eval_loss": 0.6690270900726318,
|
| 731 |
+
"eval_runtime": 1.9635,
|
| 732 |
+
"eval_samples_per_second": 244.971,
|
| 733 |
+
"eval_steps_per_second": 31.067,
|
| 734 |
+
"step": 23804
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 44.36,
|
| 738 |
+
"grad_norm": 7.927094459533691,
|
| 739 |
+
"learning_rate": 1.11275415896488e-05,
|
| 740 |
+
"loss": 1.1008,
|
| 741 |
+
"step": 24000
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 45.0,
|
| 745 |
+
"eval_accuracy": 0.7879417879417879,
|
| 746 |
+
"eval_loss": 0.659792423248291,
|
| 747 |
+
"eval_runtime": 1.9747,
|
| 748 |
+
"eval_samples_per_second": 243.578,
|
| 749 |
+
"eval_steps_per_second": 30.89,
|
| 750 |
+
"step": 24345
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 45.29,
|
| 754 |
+
"grad_norm": 8.912357330322266,
|
| 755 |
+
"learning_rate": 1.0942698706099817e-05,
|
| 756 |
+
"loss": 1.1344,
|
| 757 |
+
"step": 24500
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 46.0,
|
| 761 |
+
"eval_accuracy": 0.8024948024948025,
|
| 762 |
+
"eval_loss": 0.6550182104110718,
|
| 763 |
+
"eval_runtime": 2.0112,
|
| 764 |
+
"eval_samples_per_second": 239.156,
|
| 765 |
+
"eval_steps_per_second": 30.33,
|
| 766 |
+
"step": 24886
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 46.21,
|
| 770 |
+
"grad_norm": 9.598004341125488,
|
| 771 |
+
"learning_rate": 1.0757855822550833e-05,
|
| 772 |
+
"loss": 1.126,
|
| 773 |
+
"step": 25000
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 47.0,
|
| 777 |
+
"eval_accuracy": 0.7858627858627859,
|
| 778 |
+
"eval_loss": 0.6521425247192383,
|
| 779 |
+
"eval_runtime": 1.9713,
|
| 780 |
+
"eval_samples_per_second": 244.004,
|
| 781 |
+
"eval_steps_per_second": 30.944,
|
| 782 |
+
"step": 25427
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 47.13,
|
| 786 |
+
"grad_norm": 4.670881271362305,
|
| 787 |
+
"learning_rate": 1.057301293900185e-05,
|
| 788 |
+
"loss": 1.125,
|
| 789 |
+
"step": 25500
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 48.0,
|
| 793 |
+
"eval_accuracy": 0.7817047817047817,
|
| 794 |
+
"eval_loss": 0.6812848448753357,
|
| 795 |
+
"eval_runtime": 2.016,
|
| 796 |
+
"eval_samples_per_second": 238.588,
|
| 797 |
+
"eval_steps_per_second": 30.258,
|
| 798 |
+
"step": 25968
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 48.06,
|
| 802 |
+
"grad_norm": 8.11451244354248,
|
| 803 |
+
"learning_rate": 1.0388170055452866e-05,
|
| 804 |
+
"loss": 1.0682,
|
| 805 |
+
"step": 26000
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 48.98,
|
| 809 |
+
"grad_norm": 8.960821151733398,
|
| 810 |
+
"learning_rate": 1.0203327171903883e-05,
|
| 811 |
+
"loss": 1.0855,
|
| 812 |
+
"step": 26500
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 49.0,
|
| 816 |
+
"eval_accuracy": 0.7858627858627859,
|
| 817 |
+
"eval_loss": 0.6419298052787781,
|
| 818 |
+
"eval_runtime": 1.974,
|
| 819 |
+
"eval_samples_per_second": 243.673,
|
| 820 |
+
"eval_steps_per_second": 30.902,
|
| 821 |
+
"step": 26509
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 49.91,
|
| 825 |
+
"grad_norm": 3.053118944168091,
|
| 826 |
+
"learning_rate": 1.00184842883549e-05,
|
| 827 |
+
"loss": 1.0452,
|
| 828 |
+
"step": 27000
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 50.0,
|
| 832 |
+
"eval_accuracy": 0.8004158004158004,
|
| 833 |
+
"eval_loss": 0.6550863981246948,
|
| 834 |
+
"eval_runtime": 2.0504,
|
| 835 |
+
"eval_samples_per_second": 234.587,
|
| 836 |
+
"eval_steps_per_second": 29.75,
|
| 837 |
+
"step": 27050
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 50.83,
|
| 841 |
+
"grad_norm": 5.4594340324401855,
|
| 842 |
+
"learning_rate": 9.833641404805916e-06,
|
| 843 |
+
"loss": 1.0626,
|
| 844 |
+
"step": 27500
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 51.0,
|
| 848 |
+
"eval_accuracy": 0.7920997920997921,
|
| 849 |
+
"eval_loss": 0.6675499081611633,
|
| 850 |
+
"eval_runtime": 1.9991,
|
| 851 |
+
"eval_samples_per_second": 240.605,
|
| 852 |
+
"eval_steps_per_second": 30.513,
|
| 853 |
+
"step": 27591
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 51.76,
|
| 857 |
+
"grad_norm": 8.158236503601074,
|
| 858 |
+
"learning_rate": 9.648798521256933e-06,
|
| 859 |
+
"loss": 1.0155,
|
| 860 |
+
"step": 28000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 52.0,
|
| 864 |
+
"eval_accuracy": 0.7920997920997921,
|
| 865 |
+
"eval_loss": 0.6945971846580505,
|
| 866 |
+
"eval_runtime": 1.9873,
|
| 867 |
+
"eval_samples_per_second": 242.042,
|
| 868 |
+
"eval_steps_per_second": 30.696,
|
| 869 |
+
"step": 28132
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 52.68,
|
| 873 |
+
"grad_norm": 5.626604080200195,
|
| 874 |
+
"learning_rate": 9.46395563770795e-06,
|
| 875 |
+
"loss": 1.0319,
|
| 876 |
+
"step": 28500
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 53.0,
|
| 880 |
+
"eval_accuracy": 0.7796257796257796,
|
| 881 |
+
"eval_loss": 0.6942130923271179,
|
| 882 |
+
"eval_runtime": 1.966,
|
| 883 |
+
"eval_samples_per_second": 244.665,
|
| 884 |
+
"eval_steps_per_second": 31.028,
|
| 885 |
+
"step": 28673
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 53.6,
|
| 889 |
+
"grad_norm": 6.82182502746582,
|
| 890 |
+
"learning_rate": 9.279112754158966e-06,
|
| 891 |
+
"loss": 1.0488,
|
| 892 |
+
"step": 29000
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 54.0,
|
| 896 |
+
"eval_accuracy": 0.7983367983367984,
|
| 897 |
+
"eval_loss": 0.6496003866195679,
|
| 898 |
+
"eval_runtime": 2.0069,
|
| 899 |
+
"eval_samples_per_second": 239.673,
|
| 900 |
+
"eval_steps_per_second": 30.395,
|
| 901 |
+
"step": 29214
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 54.53,
|
| 905 |
+
"grad_norm": 7.865675926208496,
|
| 906 |
+
"learning_rate": 9.094269870609981e-06,
|
| 907 |
+
"loss": 1.0558,
|
| 908 |
+
"step": 29500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 55.0,
|
| 912 |
+
"eval_accuracy": 0.8045738045738046,
|
| 913 |
+
"eval_loss": 0.6465332508087158,
|
| 914 |
+
"eval_runtime": 1.9938,
|
| 915 |
+
"eval_samples_per_second": 241.25,
|
| 916 |
+
"eval_steps_per_second": 30.595,
|
| 917 |
+
"step": 29755
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 55.45,
|
| 921 |
+
"grad_norm": 7.172035217285156,
|
| 922 |
+
"learning_rate": 8.909426987060998e-06,
|
| 923 |
+
"loss": 0.9913,
|
| 924 |
+
"step": 30000
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 56.0,
|
| 928 |
+
"eval_accuracy": 0.7920997920997921,
|
| 929 |
+
"eval_loss": 0.6654109954833984,
|
| 930 |
+
"eval_runtime": 1.9693,
|
| 931 |
+
"eval_samples_per_second": 244.248,
|
| 932 |
+
"eval_steps_per_second": 30.975,
|
| 933 |
+
"step": 30296
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 56.38,
|
| 937 |
+
"grad_norm": 6.30518102645874,
|
| 938 |
+
"learning_rate": 8.724584103512016e-06,
|
| 939 |
+
"loss": 1.0555,
|
| 940 |
+
"step": 30500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 57.0,
|
| 944 |
+
"eval_accuracy": 0.7962577962577962,
|
| 945 |
+
"eval_loss": 0.656141996383667,
|
| 946 |
+
"eval_runtime": 1.9741,
|
| 947 |
+
"eval_samples_per_second": 243.66,
|
| 948 |
+
"eval_steps_per_second": 30.901,
|
| 949 |
+
"step": 30837
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 57.3,
|
| 953 |
+
"grad_norm": 3.0917370319366455,
|
| 954 |
+
"learning_rate": 8.539741219963033e-06,
|
| 955 |
+
"loss": 0.9803,
|
| 956 |
+
"step": 31000
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 58.0,
|
| 960 |
+
"eval_accuracy": 0.7941787941787942,
|
| 961 |
+
"eval_loss": 0.673220157623291,
|
| 962 |
+
"eval_runtime": 1.993,
|
| 963 |
+
"eval_samples_per_second": 241.346,
|
| 964 |
+
"eval_steps_per_second": 30.607,
|
| 965 |
+
"step": 31378
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 58.23,
|
| 969 |
+
"grad_norm": 8.285308837890625,
|
| 970 |
+
"learning_rate": 8.35489833641405e-06,
|
| 971 |
+
"loss": 1.0393,
|
| 972 |
+
"step": 31500
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 59.0,
|
| 976 |
+
"eval_accuracy": 0.7817047817047817,
|
| 977 |
+
"eval_loss": 0.6892696619033813,
|
| 978 |
+
"eval_runtime": 1.9798,
|
| 979 |
+
"eval_samples_per_second": 242.949,
|
| 980 |
+
"eval_steps_per_second": 30.811,
|
| 981 |
+
"step": 31919
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 59.15,
|
| 985 |
+
"grad_norm": 3.1396327018737793,
|
| 986 |
+
"learning_rate": 8.170055452865066e-06,
|
| 987 |
+
"loss": 0.9677,
|
| 988 |
+
"step": 32000
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 60.0,
|
| 992 |
+
"eval_accuracy": 0.8045738045738046,
|
| 993 |
+
"eval_loss": 0.6823599934577942,
|
| 994 |
+
"eval_runtime": 2.0127,
|
| 995 |
+
"eval_samples_per_second": 238.985,
|
| 996 |
+
"eval_steps_per_second": 30.308,
|
| 997 |
+
"step": 32460
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 60.07,
|
| 1001 |
+
"grad_norm": 12.875879287719727,
|
| 1002 |
+
"learning_rate": 7.985212569316083e-06,
|
| 1003 |
+
"loss": 1.0366,
|
| 1004 |
+
"step": 32500
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 61.0,
|
| 1008 |
+
"grad_norm": 8.770364761352539,
|
| 1009 |
+
"learning_rate": 7.8003696857671e-06,
|
| 1010 |
+
"loss": 1.0082,
|
| 1011 |
+
"step": 33000
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 61.0,
|
| 1015 |
+
"eval_accuracy": 0.7941787941787942,
|
| 1016 |
+
"eval_loss": 0.6618274450302124,
|
| 1017 |
+
"eval_runtime": 2.0008,
|
| 1018 |
+
"eval_samples_per_second": 240.403,
|
| 1019 |
+
"eval_steps_per_second": 30.488,
|
| 1020 |
+
"step": 33001
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 61.92,
|
| 1024 |
+
"grad_norm": 6.0600972175598145,
|
| 1025 |
+
"learning_rate": 7.615526802218115e-06,
|
| 1026 |
+
"loss": 1.0096,
|
| 1027 |
+
"step": 33500
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 62.0,
|
| 1031 |
+
"eval_accuracy": 0.7837837837837838,
|
| 1032 |
+
"eval_loss": 0.6691136360168457,
|
| 1033 |
+
"eval_runtime": 1.9817,
|
| 1034 |
+
"eval_samples_per_second": 242.719,
|
| 1035 |
+
"eval_steps_per_second": 30.781,
|
| 1036 |
+
"step": 33542
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 62.85,
|
| 1040 |
+
"grad_norm": 10.777630805969238,
|
| 1041 |
+
"learning_rate": 7.430683918669132e-06,
|
| 1042 |
+
"loss": 0.9685,
|
| 1043 |
+
"step": 34000
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 63.0,
|
| 1047 |
+
"eval_accuracy": 0.8024948024948025,
|
| 1048 |
+
"eval_loss": 0.6792653203010559,
|
| 1049 |
+
"eval_runtime": 2.0645,
|
| 1050 |
+
"eval_samples_per_second": 232.985,
|
| 1051 |
+
"eval_steps_per_second": 29.547,
|
| 1052 |
+
"step": 34083
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 63.77,
|
| 1056 |
+
"grad_norm": 3.9615447521209717,
|
| 1057 |
+
"learning_rate": 7.245841035120148e-06,
|
| 1058 |
+
"loss": 0.9847,
|
| 1059 |
+
"step": 34500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 64.0,
|
| 1063 |
+
"eval_accuracy": 0.7837837837837838,
|
| 1064 |
+
"eval_loss": 0.6894533634185791,
|
| 1065 |
+
"eval_runtime": 2.0054,
|
| 1066 |
+
"eval_samples_per_second": 239.847,
|
| 1067 |
+
"eval_steps_per_second": 30.417,
|
| 1068 |
+
"step": 34624
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 64.7,
|
| 1072 |
+
"grad_norm": 9.38687801361084,
|
| 1073 |
+
"learning_rate": 7.060998151571166e-06,
|
| 1074 |
+
"loss": 0.9639,
|
| 1075 |
+
"step": 35000
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 65.0,
|
| 1079 |
+
"eval_accuracy": 0.7733887733887734,
|
| 1080 |
+
"eval_loss": 0.7297117114067078,
|
| 1081 |
+
"eval_runtime": 2.0147,
|
| 1082 |
+
"eval_samples_per_second": 238.744,
|
| 1083 |
+
"eval_steps_per_second": 30.277,
|
| 1084 |
+
"step": 35165
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 65.62,
|
| 1088 |
+
"grad_norm": 12.292973518371582,
|
| 1089 |
+
"learning_rate": 6.876155268022182e-06,
|
| 1090 |
+
"loss": 0.9776,
|
| 1091 |
+
"step": 35500
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 66.0,
|
| 1095 |
+
"eval_accuracy": 0.7920997920997921,
|
| 1096 |
+
"eval_loss": 0.6561179757118225,
|
| 1097 |
+
"eval_runtime": 1.9845,
|
| 1098 |
+
"eval_samples_per_second": 242.381,
|
| 1099 |
+
"eval_steps_per_second": 30.739,
|
| 1100 |
+
"step": 35706
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 66.54,
|
| 1104 |
+
"grad_norm": 14.023015022277832,
|
| 1105 |
+
"learning_rate": 6.691312384473199e-06,
|
| 1106 |
+
"loss": 1.0074,
|
| 1107 |
+
"step": 36000
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 67.0,
|
| 1111 |
+
"eval_accuracy": 0.7775467775467776,
|
| 1112 |
+
"eval_loss": 0.6998913884162903,
|
| 1113 |
+
"eval_runtime": 1.9686,
|
| 1114 |
+
"eval_samples_per_second": 244.338,
|
| 1115 |
+
"eval_steps_per_second": 30.987,
|
| 1116 |
+
"step": 36247
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 67.47,
|
| 1120 |
+
"grad_norm": 13.870222091674805,
|
| 1121 |
+
"learning_rate": 6.506469500924215e-06,
|
| 1122 |
+
"loss": 0.9466,
|
| 1123 |
+
"step": 36500
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 68.0,
|
| 1127 |
+
"eval_accuracy": 0.7941787941787942,
|
| 1128 |
+
"eval_loss": 0.6880961656570435,
|
| 1129 |
+
"eval_runtime": 1.9686,
|
| 1130 |
+
"eval_samples_per_second": 244.34,
|
| 1131 |
+
"eval_steps_per_second": 30.987,
|
| 1132 |
+
"step": 36788
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 68.39,
|
| 1136 |
+
"grad_norm": 6.1949639320373535,
|
| 1137 |
+
"learning_rate": 6.321626617375231e-06,
|
| 1138 |
+
"loss": 0.9425,
|
| 1139 |
+
"step": 37000
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 69.0,
|
| 1143 |
+
"eval_accuracy": 0.7962577962577962,
|
| 1144 |
+
"eval_loss": 0.6805587410926819,
|
| 1145 |
+
"eval_runtime": 1.9709,
|
| 1146 |
+
"eval_samples_per_second": 244.052,
|
| 1147 |
+
"eval_steps_per_second": 30.95,
|
| 1148 |
+
"step": 37329
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 69.32,
|
| 1152 |
+
"grad_norm": 7.145143508911133,
|
| 1153 |
+
"learning_rate": 6.136783733826248e-06,
|
| 1154 |
+
"loss": 0.9594,
|
| 1155 |
+
"step": 37500
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 70.0,
|
| 1159 |
+
"eval_accuracy": 0.7900207900207901,
|
| 1160 |
+
"eval_loss": 0.7202461361885071,
|
| 1161 |
+
"eval_runtime": 2.0125,
|
| 1162 |
+
"eval_samples_per_second": 239.001,
|
| 1163 |
+
"eval_steps_per_second": 30.31,
|
| 1164 |
+
"step": 37870
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 70.24,
|
| 1168 |
+
"grad_norm": 9.215810775756836,
|
| 1169 |
+
"learning_rate": 5.951940850277265e-06,
|
| 1170 |
+
"loss": 0.9311,
|
| 1171 |
+
"step": 38000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 71.0,
|
| 1175 |
+
"eval_accuracy": 0.7754677754677755,
|
| 1176 |
+
"eval_loss": 0.7161967754364014,
|
| 1177 |
+
"eval_runtime": 1.977,
|
| 1178 |
+
"eval_samples_per_second": 243.297,
|
| 1179 |
+
"eval_steps_per_second": 30.855,
|
| 1180 |
+
"step": 38411
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 71.16,
|
| 1184 |
+
"grad_norm": 6.461187362670898,
|
| 1185 |
+
"learning_rate": 5.767097966728281e-06,
|
| 1186 |
+
"loss": 0.9429,
|
| 1187 |
+
"step": 38500
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 72.0,
|
| 1191 |
+
"eval_accuracy": 0.7920997920997921,
|
| 1192 |
+
"eval_loss": 0.7284368276596069,
|
| 1193 |
+
"eval_runtime": 2.0663,
|
| 1194 |
+
"eval_samples_per_second": 232.783,
|
| 1195 |
+
"eval_steps_per_second": 29.521,
|
| 1196 |
+
"step": 38952
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 72.09,
|
| 1200 |
+
"grad_norm": 11.850204467773438,
|
| 1201 |
+
"learning_rate": 5.582255083179298e-06,
|
| 1202 |
+
"loss": 0.9666,
|
| 1203 |
+
"step": 39000
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 73.0,
|
| 1207 |
+
"eval_accuracy": 0.7962577962577962,
|
| 1208 |
+
"eval_loss": 0.6871474981307983,
|
| 1209 |
+
"eval_runtime": 1.97,
|
| 1210 |
+
"eval_samples_per_second": 244.162,
|
| 1211 |
+
"eval_steps_per_second": 30.964,
|
| 1212 |
+
"step": 39493
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 73.01,
|
| 1216 |
+
"grad_norm": 8.0579252243042,
|
| 1217 |
+
"learning_rate": 5.3974121996303146e-06,
|
| 1218 |
+
"loss": 0.932,
|
| 1219 |
+
"step": 39500
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 73.94,
|
| 1223 |
+
"grad_norm": 1.1972132921218872,
|
| 1224 |
+
"learning_rate": 5.212569316081332e-06,
|
| 1225 |
+
"loss": 0.945,
|
| 1226 |
+
"step": 40000
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 74.0,
|
| 1230 |
+
"eval_accuracy": 0.7941787941787942,
|
| 1231 |
+
"eval_loss": 0.6778899431228638,
|
| 1232 |
+
"eval_runtime": 2.0376,
|
| 1233 |
+
"eval_samples_per_second": 236.067,
|
| 1234 |
+
"eval_steps_per_second": 29.938,
|
| 1235 |
+
"step": 40034
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 74.86,
|
| 1239 |
+
"grad_norm": 5.484439849853516,
|
| 1240 |
+
"learning_rate": 5.027726432532349e-06,
|
| 1241 |
+
"loss": 0.9387,
|
| 1242 |
+
"step": 40500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 75.0,
|
| 1246 |
+
"eval_accuracy": 0.7941787941787942,
|
| 1247 |
+
"eval_loss": 0.735752522945404,
|
| 1248 |
+
"eval_runtime": 1.9762,
|
| 1249 |
+
"eval_samples_per_second": 243.395,
|
| 1250 |
+
"eval_steps_per_second": 30.867,
|
| 1251 |
+
"step": 40575
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 75.79,
|
| 1255 |
+
"grad_norm": 2.0908420085906982,
|
| 1256 |
+
"learning_rate": 4.8428835489833645e-06,
|
| 1257 |
+
"loss": 0.9132,
|
| 1258 |
+
"step": 41000
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 76.0,
|
| 1262 |
+
"eval_accuracy": 0.7941787941787942,
|
| 1263 |
+
"eval_loss": 0.7043733596801758,
|
| 1264 |
+
"eval_runtime": 1.9764,
|
| 1265 |
+
"eval_samples_per_second": 243.367,
|
| 1266 |
+
"eval_steps_per_second": 30.864,
|
| 1267 |
+
"step": 41116
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 76.71,
|
| 1271 |
+
"grad_norm": 10.380330085754395,
|
| 1272 |
+
"learning_rate": 4.658040665434381e-06,
|
| 1273 |
+
"loss": 0.9181,
|
| 1274 |
+
"step": 41500
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 77.0,
|
| 1278 |
+
"eval_accuracy": 0.7962577962577962,
|
| 1279 |
+
"eval_loss": 0.7041053771972656,
|
| 1280 |
+
"eval_runtime": 2.0006,
|
| 1281 |
+
"eval_samples_per_second": 240.43,
|
| 1282 |
+
"eval_steps_per_second": 30.491,
|
| 1283 |
+
"step": 41657
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 77.63,
|
| 1287 |
+
"grad_norm": 9.135781288146973,
|
| 1288 |
+
"learning_rate": 4.473197781885398e-06,
|
| 1289 |
+
"loss": 0.9218,
|
| 1290 |
+
"step": 42000
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 78.0,
|
| 1294 |
+
"eval_accuracy": 0.7941787941787942,
|
| 1295 |
+
"eval_loss": 0.6986111998558044,
|
| 1296 |
+
"eval_runtime": 1.9688,
|
| 1297 |
+
"eval_samples_per_second": 244.309,
|
| 1298 |
+
"eval_steps_per_second": 30.983,
|
| 1299 |
+
"step": 42198
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 78.56,
|
| 1303 |
+
"grad_norm": 17.338001251220703,
|
| 1304 |
+
"learning_rate": 4.288354898336414e-06,
|
| 1305 |
+
"loss": 0.8621,
|
| 1306 |
+
"step": 42500
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 79.0,
|
| 1310 |
+
"eval_accuracy": 0.8004158004158004,
|
| 1311 |
+
"eval_loss": 0.6909247040748596,
|
| 1312 |
+
"eval_runtime": 1.997,
|
| 1313 |
+
"eval_samples_per_second": 240.86,
|
| 1314 |
+
"eval_steps_per_second": 30.546,
|
| 1315 |
+
"step": 42739
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 79.48,
|
| 1319 |
+
"grad_norm": 6.793923854827881,
|
| 1320 |
+
"learning_rate": 4.103512014787431e-06,
|
| 1321 |
+
"loss": 0.9236,
|
| 1322 |
+
"step": 43000
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 80.0,
|
| 1326 |
+
"eval_accuracy": 0.7983367983367984,
|
| 1327 |
+
"eval_loss": 0.7135599851608276,
|
| 1328 |
+
"eval_runtime": 1.9949,
|
| 1329 |
+
"eval_samples_per_second": 241.121,
|
| 1330 |
+
"eval_steps_per_second": 30.579,
|
| 1331 |
+
"step": 43280
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 80.41,
|
| 1335 |
+
"grad_norm": 3.9345781803131104,
|
| 1336 |
+
"learning_rate": 3.918669131238448e-06,
|
| 1337 |
+
"loss": 0.8667,
|
| 1338 |
+
"step": 43500
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 81.0,
|
| 1342 |
+
"eval_accuracy": 0.8024948024948025,
|
| 1343 |
+
"eval_loss": 0.7008742094039917,
|
| 1344 |
+
"eval_runtime": 1.992,
|
| 1345 |
+
"eval_samples_per_second": 241.461,
|
| 1346 |
+
"eval_steps_per_second": 30.622,
|
| 1347 |
+
"step": 43821
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 81.33,
|
| 1351 |
+
"grad_norm": 16.883420944213867,
|
| 1352 |
+
"learning_rate": 3.7338262476894642e-06,
|
| 1353 |
+
"loss": 0.8856,
|
| 1354 |
+
"step": 44000
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 82.0,
|
| 1358 |
+
"eval_accuracy": 0.7920997920997921,
|
| 1359 |
+
"eval_loss": 0.7127683162689209,
|
| 1360 |
+
"eval_runtime": 2.0451,
|
| 1361 |
+
"eval_samples_per_second": 235.201,
|
| 1362 |
+
"eval_steps_per_second": 29.828,
|
| 1363 |
+
"step": 44362
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 82.26,
|
| 1367 |
+
"grad_norm": 7.969069480895996,
|
| 1368 |
+
"learning_rate": 3.548983364140481e-06,
|
| 1369 |
+
"loss": 0.917,
|
| 1370 |
+
"step": 44500
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 83.0,
|
| 1374 |
+
"eval_accuracy": 0.7983367983367984,
|
| 1375 |
+
"eval_loss": 0.7134777307510376,
|
| 1376 |
+
"eval_runtime": 2.0529,
|
| 1377 |
+
"eval_samples_per_second": 234.298,
|
| 1378 |
+
"eval_steps_per_second": 29.714,
|
| 1379 |
+
"step": 44903
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 83.18,
|
| 1383 |
+
"grad_norm": 1.545163631439209,
|
| 1384 |
+
"learning_rate": 3.3641404805914975e-06,
|
| 1385 |
+
"loss": 0.8835,
|
| 1386 |
+
"step": 45000
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 84.0,
|
| 1390 |
+
"eval_accuracy": 0.7900207900207901,
|
| 1391 |
+
"eval_loss": 0.7295302748680115,
|
| 1392 |
+
"eval_runtime": 1.9747,
|
| 1393 |
+
"eval_samples_per_second": 243.583,
|
| 1394 |
+
"eval_steps_per_second": 30.891,
|
| 1395 |
+
"step": 45444
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 84.1,
|
| 1399 |
+
"grad_norm": 5.072544097900391,
|
| 1400 |
+
"learning_rate": 3.1792975970425146e-06,
|
| 1401 |
+
"loss": 0.8879,
|
| 1402 |
+
"step": 45500
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 85.0,
|
| 1406 |
+
"eval_accuracy": 0.7900207900207901,
|
| 1407 |
+
"eval_loss": 0.7449509501457214,
|
| 1408 |
+
"eval_runtime": 2.0759,
|
| 1409 |
+
"eval_samples_per_second": 231.71,
|
| 1410 |
+
"eval_steps_per_second": 29.385,
|
| 1411 |
+
"step": 45985
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 85.03,
|
| 1415 |
+
"grad_norm": 5.645694732666016,
|
| 1416 |
+
"learning_rate": 2.9944547134935308e-06,
|
| 1417 |
+
"loss": 0.9114,
|
| 1418 |
+
"step": 46000
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 85.95,
|
| 1422 |
+
"grad_norm": 5.065194129943848,
|
| 1423 |
+
"learning_rate": 2.8096118299445474e-06,
|
| 1424 |
+
"loss": 0.8764,
|
| 1425 |
+
"step": 46500
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 86.0,
|
| 1429 |
+
"eval_accuracy": 0.7941787941787942,
|
| 1430 |
+
"eval_loss": 0.7362204194068909,
|
| 1431 |
+
"eval_runtime": 1.9869,
|
| 1432 |
+
"eval_samples_per_second": 242.081,
|
| 1433 |
+
"eval_steps_per_second": 30.7,
|
| 1434 |
+
"step": 46526
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 86.88,
|
| 1438 |
+
"grad_norm": 5.654088020324707,
|
| 1439 |
+
"learning_rate": 2.624768946395564e-06,
|
| 1440 |
+
"loss": 0.8674,
|
| 1441 |
+
"step": 47000
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 87.0,
|
| 1445 |
+
"eval_accuracy": 0.7941787941787942,
|
| 1446 |
+
"eval_loss": 0.7232093811035156,
|
| 1447 |
+
"eval_runtime": 2.0232,
|
| 1448 |
+
"eval_samples_per_second": 237.746,
|
| 1449 |
+
"eval_steps_per_second": 30.151,
|
| 1450 |
+
"step": 47067
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 87.8,
|
| 1454 |
+
"grad_norm": 12.72859001159668,
|
| 1455 |
+
"learning_rate": 2.4399260628465807e-06,
|
| 1456 |
+
"loss": 0.8583,
|
| 1457 |
+
"step": 47500
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 88.0,
|
| 1461 |
+
"eval_accuracy": 0.7941787941787942,
|
| 1462 |
+
"eval_loss": 0.7407870888710022,
|
| 1463 |
+
"eval_runtime": 1.9934,
|
| 1464 |
+
"eval_samples_per_second": 241.296,
|
| 1465 |
+
"eval_steps_per_second": 30.601,
|
| 1466 |
+
"step": 47608
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 88.72,
|
| 1470 |
+
"grad_norm": 6.526777744293213,
|
| 1471 |
+
"learning_rate": 2.2550831792975973e-06,
|
| 1472 |
+
"loss": 0.881,
|
| 1473 |
+
"step": 48000
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 89.0,
|
| 1477 |
+
"eval_accuracy": 0.8004158004158004,
|
| 1478 |
+
"eval_loss": 0.7377821803092957,
|
| 1479 |
+
"eval_runtime": 1.9802,
|
| 1480 |
+
"eval_samples_per_second": 242.901,
|
| 1481 |
+
"eval_steps_per_second": 30.804,
|
| 1482 |
+
"step": 48149
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 89.65,
|
| 1486 |
+
"grad_norm": 8.497318267822266,
|
| 1487 |
+
"learning_rate": 2.070240295748614e-06,
|
| 1488 |
+
"loss": 0.8668,
|
| 1489 |
+
"step": 48500
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 90.0,
|
| 1493 |
+
"eval_accuracy": 0.7900207900207901,
|
| 1494 |
+
"eval_loss": 0.7473007440567017,
|
| 1495 |
+
"eval_runtime": 2.0137,
|
| 1496 |
+
"eval_samples_per_second": 238.867,
|
| 1497 |
+
"eval_steps_per_second": 30.293,
|
| 1498 |
+
"step": 48690
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 90.57,
|
| 1502 |
+
"grad_norm": 6.455136775970459,
|
| 1503 |
+
"learning_rate": 1.8853974121996305e-06,
|
| 1504 |
+
"loss": 0.8779,
|
| 1505 |
+
"step": 49000
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 91.0,
|
| 1509 |
+
"eval_accuracy": 0.7983367983367984,
|
| 1510 |
+
"eval_loss": 0.7438368201255798,
|
| 1511 |
+
"eval_runtime": 1.9731,
|
| 1512 |
+
"eval_samples_per_second": 243.774,
|
| 1513 |
+
"eval_steps_per_second": 30.915,
|
| 1514 |
+
"step": 49231
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 91.5,
|
| 1518 |
+
"grad_norm": 5.713993072509766,
|
| 1519 |
+
"learning_rate": 1.700554528650647e-06,
|
| 1520 |
+
"loss": 0.8717,
|
| 1521 |
+
"step": 49500
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 92.0,
|
| 1525 |
+
"eval_accuracy": 0.8004158004158004,
|
| 1526 |
+
"eval_loss": 0.7389739751815796,
|
| 1527 |
+
"eval_runtime": 1.9686,
|
| 1528 |
+
"eval_samples_per_second": 244.34,
|
| 1529 |
+
"eval_steps_per_second": 30.987,
|
| 1530 |
+
"step": 49772
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 92.42,
|
| 1534 |
+
"grad_norm": 5.342690467834473,
|
| 1535 |
+
"learning_rate": 1.5157116451016638e-06,
|
| 1536 |
+
"loss": 0.8781,
|
| 1537 |
+
"step": 50000
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 93.0,
|
| 1541 |
+
"eval_accuracy": 0.7983367983367984,
|
| 1542 |
+
"eval_loss": 0.7473535537719727,
|
| 1543 |
+
"eval_runtime": 1.98,
|
| 1544 |
+
"eval_samples_per_second": 242.926,
|
| 1545 |
+
"eval_steps_per_second": 30.808,
|
| 1546 |
+
"step": 50313
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 93.35,
|
| 1550 |
+
"grad_norm": 9.870634078979492,
|
| 1551 |
+
"learning_rate": 1.3308687615526802e-06,
|
| 1552 |
+
"loss": 0.8845,
|
| 1553 |
+
"step": 50500
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 94.0,
|
| 1557 |
+
"eval_accuracy": 0.7900207900207901,
|
| 1558 |
+
"eval_loss": 0.7445840835571289,
|
| 1559 |
+
"eval_runtime": 1.9776,
|
| 1560 |
+
"eval_samples_per_second": 243.222,
|
| 1561 |
+
"eval_steps_per_second": 30.845,
|
| 1562 |
+
"step": 50854
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 94.27,
|
| 1566 |
+
"grad_norm": 8.909347534179688,
|
| 1567 |
+
"learning_rate": 1.1460258780036969e-06,
|
| 1568 |
+
"loss": 0.8623,
|
| 1569 |
+
"step": 51000
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 95.0,
|
| 1573 |
+
"eval_accuracy": 0.7920997920997921,
|
| 1574 |
+
"eval_loss": 0.7315581440925598,
|
| 1575 |
+
"eval_runtime": 1.9728,
|
| 1576 |
+
"eval_samples_per_second": 243.814,
|
| 1577 |
+
"eval_steps_per_second": 30.92,
|
| 1578 |
+
"step": 51395
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 95.19,
|
| 1582 |
+
"grad_norm": 10.748625755310059,
|
| 1583 |
+
"learning_rate": 9.611829944547135e-07,
|
| 1584 |
+
"loss": 0.8341,
|
| 1585 |
+
"step": 51500
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 96.0,
|
| 1589 |
+
"eval_accuracy": 0.7879417879417879,
|
| 1590 |
+
"eval_loss": 0.7457364201545715,
|
| 1591 |
+
"eval_runtime": 2.0017,
|
| 1592 |
+
"eval_samples_per_second": 240.29,
|
| 1593 |
+
"eval_steps_per_second": 30.473,
|
| 1594 |
+
"step": 51936
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 96.12,
|
| 1598 |
+
"grad_norm": 3.179774761199951,
|
| 1599 |
+
"learning_rate": 7.763401109057302e-07,
|
| 1600 |
+
"loss": 0.8766,
|
| 1601 |
+
"step": 52000
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 97.0,
|
| 1605 |
+
"eval_accuracy": 0.7920997920997921,
|
| 1606 |
+
"eval_loss": 0.743617832660675,
|
| 1607 |
+
"eval_runtime": 1.999,
|
| 1608 |
+
"eval_samples_per_second": 240.625,
|
| 1609 |
+
"eval_steps_per_second": 30.516,
|
| 1610 |
+
"step": 52477
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 97.04,
|
| 1614 |
+
"grad_norm": 12.243720054626465,
|
| 1615 |
+
"learning_rate": 5.914972273567468e-07,
|
| 1616 |
+
"loss": 0.8101,
|
| 1617 |
+
"step": 52500
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 97.97,
|
| 1621 |
+
"grad_norm": 18.670886993408203,
|
| 1622 |
+
"learning_rate": 4.066543438077634e-07,
|
| 1623 |
+
"loss": 0.8681,
|
| 1624 |
+
"step": 53000
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 98.0,
|
| 1628 |
+
"eval_accuracy": 0.7900207900207901,
|
| 1629 |
+
"eval_loss": 0.7483807802200317,
|
| 1630 |
+
"eval_runtime": 2.0039,
|
| 1631 |
+
"eval_samples_per_second": 240.035,
|
| 1632 |
+
"eval_steps_per_second": 30.441,
|
| 1633 |
+
"step": 53018
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 98.89,
|
| 1637 |
+
"grad_norm": 8.483085632324219,
|
| 1638 |
+
"learning_rate": 2.2181146025878005e-07,
|
| 1639 |
+
"loss": 0.8635,
|
| 1640 |
+
"step": 53500
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 99.0,
|
| 1644 |
+
"eval_accuracy": 0.7941787941787942,
|
| 1645 |
+
"eval_loss": 0.7391884922981262,
|
| 1646 |
+
"eval_runtime": 1.9875,
|
| 1647 |
+
"eval_samples_per_second": 242.013,
|
| 1648 |
+
"eval_steps_per_second": 30.692,
|
| 1649 |
+
"step": 53559
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 99.82,
|
| 1653 |
+
"grad_norm": 10.068202018737793,
|
| 1654 |
+
"learning_rate": 3.696857670979668e-08,
|
| 1655 |
+
"loss": 0.8091,
|
| 1656 |
+
"step": 54000
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 100.0,
|
| 1660 |
+
"eval_accuracy": 0.7920997920997921,
|
| 1661 |
+
"eval_loss": 0.7390549182891846,
|
| 1662 |
+
"eval_runtime": 2.0448,
|
| 1663 |
+
"eval_samples_per_second": 235.228,
|
| 1664 |
+
"eval_steps_per_second": 29.831,
|
| 1665 |
+
"step": 54100
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 100.0,
|
| 1669 |
+
"step": 54100,
|
| 1670 |
+
"total_flos": 3.355193271048192e+19,
|
| 1671 |
+
"train_loss": 1.2517024893769495,
|
| 1672 |
+
"train_runtime": 5380.7558,
|
| 1673 |
+
"train_samples_per_second": 80.379,
|
| 1674 |
+
"train_steps_per_second": 10.054
|
| 1675 |
+
}
|
| 1676 |
+
],
|
| 1677 |
+
"logging_steps": 500,
|
| 1678 |
+
"max_steps": 54100,
|
| 1679 |
+
"num_input_tokens_seen": 0,
|
| 1680 |
+
"num_train_epochs": 100,
|
| 1681 |
+
"save_steps": 500,
|
| 1682 |
+
"total_flos": 3.355193271048192e+19,
|
| 1683 |
+
"train_batch_size": 8,
|
| 1684 |
+
"trial_name": null,
|
| 1685 |
+
"trial_params": null
|
| 1686 |
+
}
|