End of training
Browse files- README.md +3 -1
- all_results.json +11 -11
- eval_results.json +6 -6
- train_results.json +6 -6
- trainer_state.json +2194 -14
README.md
CHANGED
|
@@ -3,6 +3,8 @@ library_name: transformers
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: google/vit-base-patch16-224-in21k
|
| 5 |
tags:
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
|
@@ -16,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# vit-base-beans
|
| 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
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.0079
|
| 22 |
- Accuracy: 1.0
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: google/vit-base-patch16-224-in21k
|
| 5 |
tags:
|
| 6 |
+
- image-classification
|
| 7 |
+
- vision
|
| 8 |
- generated_from_trainer
|
| 9 |
metrics:
|
| 10 |
- accuracy
|
|
|
|
| 18 |
|
| 19 |
# vit-base-beans
|
| 20 |
|
| 21 |
+
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 beans dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
- Loss: 0.0079
|
| 24 |
- Accuracy: 1.0
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 0.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second":
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_accuracy": 1.0,
|
| 4 |
+
"eval_loss": 0.007947824895381927,
|
| 5 |
+
"eval_runtime": 0.6215,
|
| 6 |
+
"eval_samples_per_second": 213.993,
|
| 7 |
+
"eval_steps_per_second": 14.481,
|
| 8 |
+
"total_flos": 3.6243328994998477e+18,
|
| 9 |
+
"train_loss": 0.03743308119131968,
|
| 10 |
+
"train_runtime": 550.1735,
|
| 11 |
+
"train_samples_per_second": 93.97,
|
| 12 |
+
"train_steps_per_second": 5.907
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 0.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_accuracy": 1.0,
|
| 4 |
+
"eval_loss": 0.007947824895381927,
|
| 5 |
+
"eval_runtime": 0.6215,
|
| 6 |
+
"eval_samples_per_second": 213.993,
|
| 7 |
+
"eval_steps_per_second": 14.481
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"total_flos": 3.6243328994998477e+18,
|
| 4 |
+
"train_loss": 0.03743308119131968,
|
| 5 |
+
"train_runtime": 550.1735,
|
| 6 |
+
"train_samples_per_second": 93.97,
|
| 7 |
+
"train_steps_per_second": 5.907
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "./beans_outputs/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -509,19 +509,2199 @@
|
|
| 509 |
"step": 650
|
| 510 |
},
|
| 511 |
{
|
| 512 |
-
"epoch":
|
| 513 |
-
"
|
| 514 |
-
"
|
| 515 |
-
"
|
| 516 |
-
"
|
| 517 |
-
|
| 518 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 519 |
}
|
| 520 |
],
|
| 521 |
"logging_steps": 10,
|
| 522 |
-
"max_steps":
|
| 523 |
"num_input_tokens_seen": 0,
|
| 524 |
-
"num_train_epochs":
|
| 525 |
"save_steps": 500,
|
| 526 |
"stateful_callbacks": {
|
| 527 |
"TrainerControl": {
|
|
@@ -535,7 +2715,7 @@
|
|
| 535 |
"attributes": {}
|
| 536 |
}
|
| 537 |
},
|
| 538 |
-
"total_flos":
|
| 539 |
"train_batch_size": 8,
|
| 540 |
"trial_name": null,
|
| 541 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.007947824895381927,
|
| 3 |
+
"best_model_checkpoint": "./beans_outputs/checkpoint-3250",
|
| 4 |
+
"epoch": 50.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 3250,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 509 |
"step": 650
|
| 510 |
},
|
| 511 |
{
|
| 512 |
+
"epoch": 10.153846153846153,
|
| 513 |
+
"grad_norm": 2.079591751098633,
|
| 514 |
+
"learning_rate": 1.593846153846154e-05,
|
| 515 |
+
"loss": 0.0956,
|
| 516 |
+
"step": 660
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 10.307692307692308,
|
| 520 |
+
"grad_norm": 0.22365273535251617,
|
| 521 |
+
"learning_rate": 1.587692307692308e-05,
|
| 522 |
+
"loss": 0.0704,
|
| 523 |
+
"step": 670
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 10.461538461538462,
|
| 527 |
+
"grad_norm": 1.2620346546173096,
|
| 528 |
+
"learning_rate": 1.5815384615384616e-05,
|
| 529 |
+
"loss": 0.0927,
|
| 530 |
+
"step": 680
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 10.615384615384615,
|
| 534 |
+
"grad_norm": 0.17169518768787384,
|
| 535 |
+
"learning_rate": 1.5753846153846154e-05,
|
| 536 |
+
"loss": 0.1318,
|
| 537 |
+
"step": 690
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 10.76923076923077,
|
| 541 |
+
"grad_norm": 1.01598060131073,
|
| 542 |
+
"learning_rate": 1.5692307692307693e-05,
|
| 543 |
+
"loss": 0.1102,
|
| 544 |
+
"step": 700
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 10.923076923076923,
|
| 548 |
+
"grad_norm": 1.3003897666931152,
|
| 549 |
+
"learning_rate": 1.5630769230769232e-05,
|
| 550 |
+
"loss": 0.098,
|
| 551 |
+
"step": 710
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 11.0,
|
| 555 |
+
"eval_accuracy": 0.9924812030075187,
|
| 556 |
+
"eval_loss": 0.056982800364494324,
|
| 557 |
+
"eval_runtime": 0.6404,
|
| 558 |
+
"eval_samples_per_second": 207.697,
|
| 559 |
+
"eval_steps_per_second": 14.055,
|
| 560 |
+
"step": 715
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 11.076923076923077,
|
| 564 |
+
"grad_norm": 2.616173505783081,
|
| 565 |
+
"learning_rate": 1.556923076923077e-05,
|
| 566 |
+
"loss": 0.1061,
|
| 567 |
+
"step": 720
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 11.23076923076923,
|
| 571 |
+
"grad_norm": 0.20168966054916382,
|
| 572 |
+
"learning_rate": 1.550769230769231e-05,
|
| 573 |
+
"loss": 0.1275,
|
| 574 |
+
"step": 730
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 11.384615384615385,
|
| 578 |
+
"grad_norm": 2.059192419052124,
|
| 579 |
+
"learning_rate": 1.544615384615385e-05,
|
| 580 |
+
"loss": 0.0463,
|
| 581 |
+
"step": 740
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 11.538461538461538,
|
| 585 |
+
"grad_norm": 0.1697787493467331,
|
| 586 |
+
"learning_rate": 1.5384615384615387e-05,
|
| 587 |
+
"loss": 0.0759,
|
| 588 |
+
"step": 750
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 11.692307692307692,
|
| 592 |
+
"grad_norm": 0.8658211827278137,
|
| 593 |
+
"learning_rate": 1.5323076923076926e-05,
|
| 594 |
+
"loss": 0.1983,
|
| 595 |
+
"step": 760
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 11.846153846153847,
|
| 599 |
+
"grad_norm": 0.4407813251018524,
|
| 600 |
+
"learning_rate": 1.5261538461538465e-05,
|
| 601 |
+
"loss": 0.1181,
|
| 602 |
+
"step": 770
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 12.0,
|
| 606 |
+
"grad_norm": 6.601933479309082,
|
| 607 |
+
"learning_rate": 1.5200000000000002e-05,
|
| 608 |
+
"loss": 0.0935,
|
| 609 |
+
"step": 780
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 12.0,
|
| 613 |
+
"eval_accuracy": 1.0,
|
| 614 |
+
"eval_loss": 0.04177865758538246,
|
| 615 |
+
"eval_runtime": 0.6316,
|
| 616 |
+
"eval_samples_per_second": 210.569,
|
| 617 |
+
"eval_steps_per_second": 14.249,
|
| 618 |
+
"step": 780
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 12.153846153846153,
|
| 622 |
+
"grad_norm": 1.3233654499053955,
|
| 623 |
+
"learning_rate": 1.5138461538461539e-05,
|
| 624 |
+
"loss": 0.0949,
|
| 625 |
+
"step": 790
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 12.307692307692308,
|
| 629 |
+
"grad_norm": 10.496395111083984,
|
| 630 |
+
"learning_rate": 1.5076923076923078e-05,
|
| 631 |
+
"loss": 0.1141,
|
| 632 |
+
"step": 800
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 12.461538461538462,
|
| 636 |
+
"grad_norm": 0.40288811922073364,
|
| 637 |
+
"learning_rate": 1.5015384615384617e-05,
|
| 638 |
+
"loss": 0.0662,
|
| 639 |
+
"step": 810
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 12.615384615384615,
|
| 643 |
+
"grad_norm": 1.3594541549682617,
|
| 644 |
+
"learning_rate": 1.4953846153846154e-05,
|
| 645 |
+
"loss": 0.0672,
|
| 646 |
+
"step": 820
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 12.76923076923077,
|
| 650 |
+
"grad_norm": 0.15957416594028473,
|
| 651 |
+
"learning_rate": 1.4892307692307692e-05,
|
| 652 |
+
"loss": 0.1198,
|
| 653 |
+
"step": 830
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 12.923076923076923,
|
| 657 |
+
"grad_norm": 3.9073538780212402,
|
| 658 |
+
"learning_rate": 1.4830769230769233e-05,
|
| 659 |
+
"loss": 0.0907,
|
| 660 |
+
"step": 840
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 13.0,
|
| 664 |
+
"eval_accuracy": 0.9699248120300752,
|
| 665 |
+
"eval_loss": 0.10930211842060089,
|
| 666 |
+
"eval_runtime": 0.6314,
|
| 667 |
+
"eval_samples_per_second": 210.658,
|
| 668 |
+
"eval_steps_per_second": 14.255,
|
| 669 |
+
"step": 845
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 13.076923076923077,
|
| 673 |
+
"grad_norm": 4.819633483886719,
|
| 674 |
+
"learning_rate": 1.4769230769230772e-05,
|
| 675 |
+
"loss": 0.0678,
|
| 676 |
+
"step": 850
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 13.23076923076923,
|
| 680 |
+
"grad_norm": 0.1750306636095047,
|
| 681 |
+
"learning_rate": 1.4707692307692309e-05,
|
| 682 |
+
"loss": 0.0498,
|
| 683 |
+
"step": 860
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 13.384615384615385,
|
| 687 |
+
"grad_norm": 0.7398102879524231,
|
| 688 |
+
"learning_rate": 1.4646153846153848e-05,
|
| 689 |
+
"loss": 0.112,
|
| 690 |
+
"step": 870
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 13.538461538461538,
|
| 694 |
+
"grad_norm": 1.1426323652267456,
|
| 695 |
+
"learning_rate": 1.4584615384615386e-05,
|
| 696 |
+
"loss": 0.061,
|
| 697 |
+
"step": 880
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 13.692307692307692,
|
| 701 |
+
"grad_norm": 0.16307789087295532,
|
| 702 |
+
"learning_rate": 1.4523076923076923e-05,
|
| 703 |
+
"loss": 0.0405,
|
| 704 |
+
"step": 890
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 13.846153846153847,
|
| 708 |
+
"grad_norm": 3.81508207321167,
|
| 709 |
+
"learning_rate": 1.4461538461538462e-05,
|
| 710 |
+
"loss": 0.0768,
|
| 711 |
+
"step": 900
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 14.0,
|
| 715 |
+
"grad_norm": 0.482637494802475,
|
| 716 |
+
"learning_rate": 1.4400000000000001e-05,
|
| 717 |
+
"loss": 0.0947,
|
| 718 |
+
"step": 910
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 14.0,
|
| 722 |
+
"eval_accuracy": 1.0,
|
| 723 |
+
"eval_loss": 0.03473825752735138,
|
| 724 |
+
"eval_runtime": 0.6324,
|
| 725 |
+
"eval_samples_per_second": 210.312,
|
| 726 |
+
"eval_steps_per_second": 14.232,
|
| 727 |
+
"step": 910
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 14.153846153846153,
|
| 731 |
+
"grad_norm": 0.18653550744056702,
|
| 732 |
+
"learning_rate": 1.4338461538461538e-05,
|
| 733 |
+
"loss": 0.0853,
|
| 734 |
+
"step": 920
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 14.307692307692308,
|
| 738 |
+
"grad_norm": 1.822533130645752,
|
| 739 |
+
"learning_rate": 1.4276923076923077e-05,
|
| 740 |
+
"loss": 0.1008,
|
| 741 |
+
"step": 930
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 14.461538461538462,
|
| 745 |
+
"grad_norm": 0.32205572724342346,
|
| 746 |
+
"learning_rate": 1.4215384615384617e-05,
|
| 747 |
+
"loss": 0.0638,
|
| 748 |
+
"step": 940
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 14.615384615384615,
|
| 752 |
+
"grad_norm": 0.5714139342308044,
|
| 753 |
+
"learning_rate": 1.4153846153846156e-05,
|
| 754 |
+
"loss": 0.1159,
|
| 755 |
+
"step": 950
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 14.76923076923077,
|
| 759 |
+
"grad_norm": 5.901656627655029,
|
| 760 |
+
"learning_rate": 1.4092307692307693e-05,
|
| 761 |
+
"loss": 0.091,
|
| 762 |
+
"step": 960
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 14.923076923076923,
|
| 766 |
+
"grad_norm": 2.119544744491577,
|
| 767 |
+
"learning_rate": 1.4030769230769232e-05,
|
| 768 |
+
"loss": 0.1259,
|
| 769 |
+
"step": 970
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 15.0,
|
| 773 |
+
"eval_accuracy": 0.9849624060150376,
|
| 774 |
+
"eval_loss": 0.07099475711584091,
|
| 775 |
+
"eval_runtime": 0.6277,
|
| 776 |
+
"eval_samples_per_second": 211.88,
|
| 777 |
+
"eval_steps_per_second": 14.338,
|
| 778 |
+
"step": 975
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 15.076923076923077,
|
| 782 |
+
"grad_norm": 0.14739356935024261,
|
| 783 |
+
"learning_rate": 1.3969230769230771e-05,
|
| 784 |
+
"loss": 0.082,
|
| 785 |
+
"step": 980
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 15.23076923076923,
|
| 789 |
+
"grad_norm": 0.12192496657371521,
|
| 790 |
+
"learning_rate": 1.3907692307692308e-05,
|
| 791 |
+
"loss": 0.0558,
|
| 792 |
+
"step": 990
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 15.384615384615385,
|
| 796 |
+
"grad_norm": 0.11841657012701035,
|
| 797 |
+
"learning_rate": 1.3846153846153847e-05,
|
| 798 |
+
"loss": 0.0852,
|
| 799 |
+
"step": 1000
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 15.538461538461538,
|
| 803 |
+
"grad_norm": 0.8062542080879211,
|
| 804 |
+
"learning_rate": 1.3784615384615386e-05,
|
| 805 |
+
"loss": 0.0793,
|
| 806 |
+
"step": 1010
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 15.692307692307692,
|
| 810 |
+
"grad_norm": 0.11781653016805649,
|
| 811 |
+
"learning_rate": 1.3723076923076923e-05,
|
| 812 |
+
"loss": 0.0436,
|
| 813 |
+
"step": 1020
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 15.846153846153847,
|
| 817 |
+
"grad_norm": 0.4233933389186859,
|
| 818 |
+
"learning_rate": 1.3661538461538461e-05,
|
| 819 |
+
"loss": 0.0448,
|
| 820 |
+
"step": 1030
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 16.0,
|
| 824 |
+
"grad_norm": 0.3999457061290741,
|
| 825 |
+
"learning_rate": 1.3600000000000002e-05,
|
| 826 |
+
"loss": 0.0325,
|
| 827 |
+
"step": 1040
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 16.0,
|
| 831 |
+
"eval_accuracy": 0.9774436090225563,
|
| 832 |
+
"eval_loss": 0.05867745727300644,
|
| 833 |
+
"eval_runtime": 0.6418,
|
| 834 |
+
"eval_samples_per_second": 207.244,
|
| 835 |
+
"eval_steps_per_second": 14.024,
|
| 836 |
+
"step": 1040
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 16.153846153846153,
|
| 840 |
+
"grad_norm": 0.36126333475112915,
|
| 841 |
+
"learning_rate": 1.353846153846154e-05,
|
| 842 |
+
"loss": 0.0373,
|
| 843 |
+
"step": 1050
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 16.307692307692307,
|
| 847 |
+
"grad_norm": 2.5870630741119385,
|
| 848 |
+
"learning_rate": 1.3476923076923078e-05,
|
| 849 |
+
"loss": 0.0746,
|
| 850 |
+
"step": 1060
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 16.46153846153846,
|
| 854 |
+
"grad_norm": 10.347090721130371,
|
| 855 |
+
"learning_rate": 1.3415384615384617e-05,
|
| 856 |
+
"loss": 0.047,
|
| 857 |
+
"step": 1070
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 16.615384615384617,
|
| 861 |
+
"grad_norm": 0.5827314257621765,
|
| 862 |
+
"learning_rate": 1.3353846153846155e-05,
|
| 863 |
+
"loss": 0.0832,
|
| 864 |
+
"step": 1080
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 16.76923076923077,
|
| 868 |
+
"grad_norm": 0.1550850123167038,
|
| 869 |
+
"learning_rate": 1.3292307692307692e-05,
|
| 870 |
+
"loss": 0.1211,
|
| 871 |
+
"step": 1090
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 16.923076923076923,
|
| 875 |
+
"grad_norm": 0.8602961897850037,
|
| 876 |
+
"learning_rate": 1.3230769230769231e-05,
|
| 877 |
+
"loss": 0.1397,
|
| 878 |
+
"step": 1100
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 17.0,
|
| 882 |
+
"eval_accuracy": 0.9924812030075187,
|
| 883 |
+
"eval_loss": 0.049453094601631165,
|
| 884 |
+
"eval_runtime": 0.6291,
|
| 885 |
+
"eval_samples_per_second": 211.407,
|
| 886 |
+
"eval_steps_per_second": 14.306,
|
| 887 |
+
"step": 1105
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 17.076923076923077,
|
| 891 |
+
"grad_norm": 1.7802873849868774,
|
| 892 |
+
"learning_rate": 1.316923076923077e-05,
|
| 893 |
+
"loss": 0.0908,
|
| 894 |
+
"step": 1110
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 17.23076923076923,
|
| 898 |
+
"grad_norm": 0.10594117641448975,
|
| 899 |
+
"learning_rate": 1.3107692307692307e-05,
|
| 900 |
+
"loss": 0.0698,
|
| 901 |
+
"step": 1120
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 17.384615384615383,
|
| 905 |
+
"grad_norm": 0.21712522208690643,
|
| 906 |
+
"learning_rate": 1.3046153846153846e-05,
|
| 907 |
+
"loss": 0.1116,
|
| 908 |
+
"step": 1130
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 17.53846153846154,
|
| 912 |
+
"grad_norm": 6.615525245666504,
|
| 913 |
+
"learning_rate": 1.2984615384615386e-05,
|
| 914 |
+
"loss": 0.0767,
|
| 915 |
+
"step": 1140
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 17.692307692307693,
|
| 919 |
+
"grad_norm": 0.10564947128295898,
|
| 920 |
+
"learning_rate": 1.2923076923076925e-05,
|
| 921 |
+
"loss": 0.0522,
|
| 922 |
+
"step": 1150
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 17.846153846153847,
|
| 926 |
+
"grad_norm": 1.2303924560546875,
|
| 927 |
+
"learning_rate": 1.2861538461538462e-05,
|
| 928 |
+
"loss": 0.0558,
|
| 929 |
+
"step": 1160
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 18.0,
|
| 933 |
+
"grad_norm": 0.10486020892858505,
|
| 934 |
+
"learning_rate": 1.2800000000000001e-05,
|
| 935 |
+
"loss": 0.0456,
|
| 936 |
+
"step": 1170
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 18.0,
|
| 940 |
+
"eval_accuracy": 0.9774436090225563,
|
| 941 |
+
"eval_loss": 0.051864467561244965,
|
| 942 |
+
"eval_runtime": 0.6367,
|
| 943 |
+
"eval_samples_per_second": 208.899,
|
| 944 |
+
"eval_steps_per_second": 14.136,
|
| 945 |
+
"step": 1170
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 18.153846153846153,
|
| 949 |
+
"grad_norm": 0.09572620689868927,
|
| 950 |
+
"learning_rate": 1.273846153846154e-05,
|
| 951 |
+
"loss": 0.0446,
|
| 952 |
+
"step": 1180
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 18.307692307692307,
|
| 956 |
+
"grad_norm": 0.12069143354892731,
|
| 957 |
+
"learning_rate": 1.2676923076923077e-05,
|
| 958 |
+
"loss": 0.0612,
|
| 959 |
+
"step": 1190
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 18.46153846153846,
|
| 963 |
+
"grad_norm": 3.15175724029541,
|
| 964 |
+
"learning_rate": 1.2615384615384616e-05,
|
| 965 |
+
"loss": 0.1213,
|
| 966 |
+
"step": 1200
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 18.615384615384617,
|
| 970 |
+
"grad_norm": 0.6527738571166992,
|
| 971 |
+
"learning_rate": 1.2553846153846155e-05,
|
| 972 |
+
"loss": 0.1027,
|
| 973 |
+
"step": 1210
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 18.76923076923077,
|
| 977 |
+
"grad_norm": 0.6189332604408264,
|
| 978 |
+
"learning_rate": 1.2492307692307692e-05,
|
| 979 |
+
"loss": 0.0262,
|
| 980 |
+
"step": 1220
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 18.923076923076923,
|
| 984 |
+
"grad_norm": 0.1209394633769989,
|
| 985 |
+
"learning_rate": 1.243076923076923e-05,
|
| 986 |
+
"loss": 0.0439,
|
| 987 |
+
"step": 1230
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 19.0,
|
| 991 |
+
"eval_accuracy": 1.0,
|
| 992 |
+
"eval_loss": 0.021639494225382805,
|
| 993 |
+
"eval_runtime": 0.6265,
|
| 994 |
+
"eval_samples_per_second": 212.296,
|
| 995 |
+
"eval_steps_per_second": 14.366,
|
| 996 |
+
"step": 1235
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 19.076923076923077,
|
| 1000 |
+
"grad_norm": 6.580589771270752,
|
| 1001 |
+
"learning_rate": 1.2369230769230771e-05,
|
| 1002 |
+
"loss": 0.0469,
|
| 1003 |
+
"step": 1240
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 19.23076923076923,
|
| 1007 |
+
"grad_norm": 8.734590530395508,
|
| 1008 |
+
"learning_rate": 1.230769230769231e-05,
|
| 1009 |
+
"loss": 0.0297,
|
| 1010 |
+
"step": 1250
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 19.384615384615383,
|
| 1014 |
+
"grad_norm": 1.7616609334945679,
|
| 1015 |
+
"learning_rate": 1.2246153846153847e-05,
|
| 1016 |
+
"loss": 0.1147,
|
| 1017 |
+
"step": 1260
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 19.53846153846154,
|
| 1021 |
+
"grad_norm": 0.3767222464084625,
|
| 1022 |
+
"learning_rate": 1.2184615384615386e-05,
|
| 1023 |
+
"loss": 0.0499,
|
| 1024 |
+
"step": 1270
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 19.692307692307693,
|
| 1028 |
+
"grad_norm": 0.1071651503443718,
|
| 1029 |
+
"learning_rate": 1.2123076923076924e-05,
|
| 1030 |
+
"loss": 0.0623,
|
| 1031 |
+
"step": 1280
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 19.846153846153847,
|
| 1035 |
+
"grad_norm": 4.712902545928955,
|
| 1036 |
+
"learning_rate": 1.2061538461538462e-05,
|
| 1037 |
+
"loss": 0.0365,
|
| 1038 |
+
"step": 1290
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 20.0,
|
| 1042 |
+
"grad_norm": 0.08926769345998764,
|
| 1043 |
+
"learning_rate": 1.2e-05,
|
| 1044 |
+
"loss": 0.0484,
|
| 1045 |
+
"step": 1300
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 20.0,
|
| 1049 |
+
"eval_accuracy": 0.9924812030075187,
|
| 1050 |
+
"eval_loss": 0.03160810098052025,
|
| 1051 |
+
"eval_runtime": 0.6412,
|
| 1052 |
+
"eval_samples_per_second": 207.432,
|
| 1053 |
+
"eval_steps_per_second": 14.037,
|
| 1054 |
+
"step": 1300
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 20.153846153846153,
|
| 1058 |
+
"grad_norm": 1.210523009300232,
|
| 1059 |
+
"learning_rate": 1.1938461538461539e-05,
|
| 1060 |
+
"loss": 0.0789,
|
| 1061 |
+
"step": 1310
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 20.307692307692307,
|
| 1065 |
+
"grad_norm": 0.21540193259716034,
|
| 1066 |
+
"learning_rate": 1.1876923076923076e-05,
|
| 1067 |
+
"loss": 0.0214,
|
| 1068 |
+
"step": 1320
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 20.46153846153846,
|
| 1072 |
+
"grad_norm": 0.08566620200872421,
|
| 1073 |
+
"learning_rate": 1.1815384615384617e-05,
|
| 1074 |
+
"loss": 0.0308,
|
| 1075 |
+
"step": 1330
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 20.615384615384617,
|
| 1079 |
+
"grad_norm": 1.3403387069702148,
|
| 1080 |
+
"learning_rate": 1.1753846153846155e-05,
|
| 1081 |
+
"loss": 0.0656,
|
| 1082 |
+
"step": 1340
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 20.76923076923077,
|
| 1086 |
+
"grad_norm": 1.3895039558410645,
|
| 1087 |
+
"learning_rate": 1.1692307692307694e-05,
|
| 1088 |
+
"loss": 0.0651,
|
| 1089 |
+
"step": 1350
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 20.923076923076923,
|
| 1093 |
+
"grad_norm": 2.5756139755249023,
|
| 1094 |
+
"learning_rate": 1.1630769230769231e-05,
|
| 1095 |
+
"loss": 0.0276,
|
| 1096 |
+
"step": 1360
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 21.0,
|
| 1100 |
+
"eval_accuracy": 1.0,
|
| 1101 |
+
"eval_loss": 0.019228629767894745,
|
| 1102 |
+
"eval_runtime": 0.6415,
|
| 1103 |
+
"eval_samples_per_second": 207.316,
|
| 1104 |
+
"eval_steps_per_second": 14.029,
|
| 1105 |
+
"step": 1365
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 21.076923076923077,
|
| 1109 |
+
"grad_norm": 0.11552825570106506,
|
| 1110 |
+
"learning_rate": 1.156923076923077e-05,
|
| 1111 |
+
"loss": 0.0494,
|
| 1112 |
+
"step": 1370
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 21.23076923076923,
|
| 1116 |
+
"grad_norm": 0.08301220834255219,
|
| 1117 |
+
"learning_rate": 1.1507692307692309e-05,
|
| 1118 |
+
"loss": 0.1079,
|
| 1119 |
+
"step": 1380
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 21.384615384615383,
|
| 1123 |
+
"grad_norm": 0.08622205257415771,
|
| 1124 |
+
"learning_rate": 1.1446153846153846e-05,
|
| 1125 |
+
"loss": 0.0397,
|
| 1126 |
+
"step": 1390
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 21.53846153846154,
|
| 1130 |
+
"grad_norm": 1.3071388006210327,
|
| 1131 |
+
"learning_rate": 1.1384615384615385e-05,
|
| 1132 |
+
"loss": 0.1094,
|
| 1133 |
+
"step": 1400
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 21.692307692307693,
|
| 1137 |
+
"grad_norm": 17.1097469329834,
|
| 1138 |
+
"learning_rate": 1.1323076923076924e-05,
|
| 1139 |
+
"loss": 0.0298,
|
| 1140 |
+
"step": 1410
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 21.846153846153847,
|
| 1144 |
+
"grad_norm": 0.08082272112369537,
|
| 1145 |
+
"learning_rate": 1.126153846153846e-05,
|
| 1146 |
+
"loss": 0.0196,
|
| 1147 |
+
"step": 1420
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 22.0,
|
| 1151 |
+
"grad_norm": 0.08441055566072464,
|
| 1152 |
+
"learning_rate": 1.1200000000000001e-05,
|
| 1153 |
+
"loss": 0.0348,
|
| 1154 |
+
"step": 1430
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 22.0,
|
| 1158 |
+
"eval_accuracy": 1.0,
|
| 1159 |
+
"eval_loss": 0.0177127867937088,
|
| 1160 |
+
"eval_runtime": 0.5805,
|
| 1161 |
+
"eval_samples_per_second": 229.097,
|
| 1162 |
+
"eval_steps_per_second": 15.503,
|
| 1163 |
+
"step": 1430
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 22.153846153846153,
|
| 1167 |
+
"grad_norm": 0.6167936325073242,
|
| 1168 |
+
"learning_rate": 1.113846153846154e-05,
|
| 1169 |
+
"loss": 0.0675,
|
| 1170 |
+
"step": 1440
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 22.307692307692307,
|
| 1174 |
+
"grad_norm": 6.9912638664245605,
|
| 1175 |
+
"learning_rate": 1.1076923076923079e-05,
|
| 1176 |
+
"loss": 0.0644,
|
| 1177 |
+
"step": 1450
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 22.46153846153846,
|
| 1181 |
+
"grad_norm": 0.07652874290943146,
|
| 1182 |
+
"learning_rate": 1.1015384615384616e-05,
|
| 1183 |
+
"loss": 0.0572,
|
| 1184 |
+
"step": 1460
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 22.615384615384617,
|
| 1188 |
+
"grad_norm": 0.08351747691631317,
|
| 1189 |
+
"learning_rate": 1.0953846153846155e-05,
|
| 1190 |
+
"loss": 0.051,
|
| 1191 |
+
"step": 1470
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 22.76923076923077,
|
| 1195 |
+
"grad_norm": 0.11051066219806671,
|
| 1196 |
+
"learning_rate": 1.0892307692307693e-05,
|
| 1197 |
+
"loss": 0.032,
|
| 1198 |
+
"step": 1480
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 22.923076923076923,
|
| 1202 |
+
"grad_norm": 10.532815933227539,
|
| 1203 |
+
"learning_rate": 1.083076923076923e-05,
|
| 1204 |
+
"loss": 0.0326,
|
| 1205 |
+
"step": 1490
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 23.0,
|
| 1209 |
+
"eval_accuracy": 1.0,
|
| 1210 |
+
"eval_loss": 0.01754232682287693,
|
| 1211 |
+
"eval_runtime": 0.6431,
|
| 1212 |
+
"eval_samples_per_second": 206.82,
|
| 1213 |
+
"eval_steps_per_second": 13.995,
|
| 1214 |
+
"step": 1495
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 23.076923076923077,
|
| 1218 |
+
"grad_norm": 0.08222197741270065,
|
| 1219 |
+
"learning_rate": 1.076923076923077e-05,
|
| 1220 |
+
"loss": 0.0462,
|
| 1221 |
+
"step": 1500
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 23.23076923076923,
|
| 1225 |
+
"grad_norm": 0.0753277987241745,
|
| 1226 |
+
"learning_rate": 1.0707692307692308e-05,
|
| 1227 |
+
"loss": 0.0516,
|
| 1228 |
+
"step": 1510
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 23.384615384615383,
|
| 1232 |
+
"grad_norm": 0.9446002840995789,
|
| 1233 |
+
"learning_rate": 1.0646153846153845e-05,
|
| 1234 |
+
"loss": 0.0214,
|
| 1235 |
+
"step": 1520
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 23.53846153846154,
|
| 1239 |
+
"grad_norm": 0.0864008441567421,
|
| 1240 |
+
"learning_rate": 1.0584615384615386e-05,
|
| 1241 |
+
"loss": 0.0185,
|
| 1242 |
+
"step": 1530
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 23.692307692307693,
|
| 1246 |
+
"grad_norm": 7.382317543029785,
|
| 1247 |
+
"learning_rate": 1.0523076923076924e-05,
|
| 1248 |
+
"loss": 0.0674,
|
| 1249 |
+
"step": 1540
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 23.846153846153847,
|
| 1253 |
+
"grad_norm": 0.18953648209571838,
|
| 1254 |
+
"learning_rate": 1.0461538461538463e-05,
|
| 1255 |
+
"loss": 0.02,
|
| 1256 |
+
"step": 1550
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 24.0,
|
| 1260 |
+
"grad_norm": 2.4876315593719482,
|
| 1261 |
+
"learning_rate": 1.04e-05,
|
| 1262 |
+
"loss": 0.1014,
|
| 1263 |
+
"step": 1560
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 24.0,
|
| 1267 |
+
"eval_accuracy": 0.9924812030075187,
|
| 1268 |
+
"eval_loss": 0.02354033850133419,
|
| 1269 |
+
"eval_runtime": 0.6442,
|
| 1270 |
+
"eval_samples_per_second": 206.452,
|
| 1271 |
+
"eval_steps_per_second": 13.97,
|
| 1272 |
+
"step": 1560
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 24.153846153846153,
|
| 1276 |
+
"grad_norm": 0.08115736395120621,
|
| 1277 |
+
"learning_rate": 1.033846153846154e-05,
|
| 1278 |
+
"loss": 0.014,
|
| 1279 |
+
"step": 1570
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 24.307692307692307,
|
| 1283 |
+
"grad_norm": 0.06984913349151611,
|
| 1284 |
+
"learning_rate": 1.0276923076923078e-05,
|
| 1285 |
+
"loss": 0.0328,
|
| 1286 |
+
"step": 1580
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 24.46153846153846,
|
| 1290 |
+
"grad_norm": 7.471628665924072,
|
| 1291 |
+
"learning_rate": 1.0215384615384615e-05,
|
| 1292 |
+
"loss": 0.0321,
|
| 1293 |
+
"step": 1590
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 24.615384615384617,
|
| 1297 |
+
"grad_norm": 0.0844321921467781,
|
| 1298 |
+
"learning_rate": 1.0153846153846154e-05,
|
| 1299 |
+
"loss": 0.051,
|
| 1300 |
+
"step": 1600
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 24.76923076923077,
|
| 1304 |
+
"grad_norm": 0.0683414489030838,
|
| 1305 |
+
"learning_rate": 1.0092307692307693e-05,
|
| 1306 |
+
"loss": 0.0428,
|
| 1307 |
+
"step": 1610
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 24.923076923076923,
|
| 1311 |
+
"grad_norm": 0.07370961457490921,
|
| 1312 |
+
"learning_rate": 1.0030769230769231e-05,
|
| 1313 |
+
"loss": 0.0395,
|
| 1314 |
+
"step": 1620
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 25.0,
|
| 1318 |
+
"eval_accuracy": 0.9849624060150376,
|
| 1319 |
+
"eval_loss": 0.04511820524930954,
|
| 1320 |
+
"eval_runtime": 0.6443,
|
| 1321 |
+
"eval_samples_per_second": 206.419,
|
| 1322 |
+
"eval_steps_per_second": 13.968,
|
| 1323 |
+
"step": 1625
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 25.076923076923077,
|
| 1327 |
+
"grad_norm": 0.7060135006904602,
|
| 1328 |
+
"learning_rate": 9.96923076923077e-06,
|
| 1329 |
+
"loss": 0.0197,
|
| 1330 |
+
"step": 1630
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 25.23076923076923,
|
| 1334 |
+
"grad_norm": 0.5647442936897278,
|
| 1335 |
+
"learning_rate": 9.907692307692309e-06,
|
| 1336 |
+
"loss": 0.0636,
|
| 1337 |
+
"step": 1640
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 25.384615384615383,
|
| 1341 |
+
"grad_norm": 0.06951487809419632,
|
| 1342 |
+
"learning_rate": 9.846153846153848e-06,
|
| 1343 |
+
"loss": 0.0338,
|
| 1344 |
+
"step": 1650
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 25.53846153846154,
|
| 1348 |
+
"grad_norm": 0.9578651785850525,
|
| 1349 |
+
"learning_rate": 9.784615384615387e-06,
|
| 1350 |
+
"loss": 0.049,
|
| 1351 |
+
"step": 1660
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 25.692307692307693,
|
| 1355 |
+
"grad_norm": 0.1774640828371048,
|
| 1356 |
+
"learning_rate": 9.723076923076924e-06,
|
| 1357 |
+
"loss": 0.0135,
|
| 1358 |
+
"step": 1670
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 25.846153846153847,
|
| 1362 |
+
"grad_norm": 0.06652193516492844,
|
| 1363 |
+
"learning_rate": 9.661538461538462e-06,
|
| 1364 |
+
"loss": 0.046,
|
| 1365 |
+
"step": 1680
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 26.0,
|
| 1369 |
+
"grad_norm": 0.06518968194723129,
|
| 1370 |
+
"learning_rate": 9.600000000000001e-06,
|
| 1371 |
+
"loss": 0.0265,
|
| 1372 |
+
"step": 1690
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 26.0,
|
| 1376 |
+
"eval_accuracy": 0.9924812030075187,
|
| 1377 |
+
"eval_loss": 0.0296646561473608,
|
| 1378 |
+
"eval_runtime": 0.5911,
|
| 1379 |
+
"eval_samples_per_second": 225.018,
|
| 1380 |
+
"eval_steps_per_second": 15.227,
|
| 1381 |
+
"step": 1690
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 26.153846153846153,
|
| 1385 |
+
"grad_norm": 0.07430905103683472,
|
| 1386 |
+
"learning_rate": 9.53846153846154e-06,
|
| 1387 |
+
"loss": 0.0326,
|
| 1388 |
+
"step": 1700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 26.307692307692307,
|
| 1392 |
+
"grad_norm": 0.07437779009342194,
|
| 1393 |
+
"learning_rate": 9.476923076923079e-06,
|
| 1394 |
+
"loss": 0.0205,
|
| 1395 |
+
"step": 1710
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 26.46153846153846,
|
| 1399 |
+
"grad_norm": 5.995608329772949,
|
| 1400 |
+
"learning_rate": 9.415384615384616e-06,
|
| 1401 |
+
"loss": 0.0725,
|
| 1402 |
+
"step": 1720
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 26.615384615384617,
|
| 1406 |
+
"grad_norm": 0.09473489224910736,
|
| 1407 |
+
"learning_rate": 9.353846153846155e-06,
|
| 1408 |
+
"loss": 0.0142,
|
| 1409 |
+
"step": 1730
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 26.76923076923077,
|
| 1413 |
+
"grad_norm": 0.05937571823596954,
|
| 1414 |
+
"learning_rate": 9.292307692307694e-06,
|
| 1415 |
+
"loss": 0.0212,
|
| 1416 |
+
"step": 1740
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 26.923076923076923,
|
| 1420 |
+
"grad_norm": 19.040122985839844,
|
| 1421 |
+
"learning_rate": 9.230769230769232e-06,
|
| 1422 |
+
"loss": 0.0569,
|
| 1423 |
+
"step": 1750
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 27.0,
|
| 1427 |
+
"eval_accuracy": 0.9924812030075187,
|
| 1428 |
+
"eval_loss": 0.026294343173503876,
|
| 1429 |
+
"eval_runtime": 0.642,
|
| 1430 |
+
"eval_samples_per_second": 207.155,
|
| 1431 |
+
"eval_steps_per_second": 14.018,
|
| 1432 |
+
"step": 1755
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 27.076923076923077,
|
| 1436 |
+
"grad_norm": 0.5836367011070251,
|
| 1437 |
+
"learning_rate": 9.169230769230771e-06,
|
| 1438 |
+
"loss": 0.1035,
|
| 1439 |
+
"step": 1760
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 27.23076923076923,
|
| 1443 |
+
"grad_norm": 0.06087055802345276,
|
| 1444 |
+
"learning_rate": 9.107692307692308e-06,
|
| 1445 |
+
"loss": 0.0518,
|
| 1446 |
+
"step": 1770
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 27.384615384615383,
|
| 1450 |
+
"grad_norm": 0.06178651750087738,
|
| 1451 |
+
"learning_rate": 9.046153846153847e-06,
|
| 1452 |
+
"loss": 0.0477,
|
| 1453 |
+
"step": 1780
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 27.53846153846154,
|
| 1457 |
+
"grad_norm": 0.0936984047293663,
|
| 1458 |
+
"learning_rate": 8.984615384615386e-06,
|
| 1459 |
+
"loss": 0.012,
|
| 1460 |
+
"step": 1790
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 27.692307692307693,
|
| 1464 |
+
"grad_norm": 0.056670334190130234,
|
| 1465 |
+
"learning_rate": 8.923076923076925e-06,
|
| 1466 |
+
"loss": 0.0363,
|
| 1467 |
+
"step": 1800
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 27.846153846153847,
|
| 1471 |
+
"grad_norm": 0.13299456238746643,
|
| 1472 |
+
"learning_rate": 8.861538461538463e-06,
|
| 1473 |
+
"loss": 0.013,
|
| 1474 |
+
"step": 1810
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 28.0,
|
| 1478 |
+
"grad_norm": 0.0713280737400055,
|
| 1479 |
+
"learning_rate": 8.8e-06,
|
| 1480 |
+
"loss": 0.0666,
|
| 1481 |
+
"step": 1820
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 28.0,
|
| 1485 |
+
"eval_accuracy": 0.9849624060150376,
|
| 1486 |
+
"eval_loss": 0.02451479434967041,
|
| 1487 |
+
"eval_runtime": 0.6311,
|
| 1488 |
+
"eval_samples_per_second": 210.727,
|
| 1489 |
+
"eval_steps_per_second": 14.26,
|
| 1490 |
+
"step": 1820
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 28.153846153846153,
|
| 1494 |
+
"grad_norm": 0.08556320518255234,
|
| 1495 |
+
"learning_rate": 8.73846153846154e-06,
|
| 1496 |
+
"loss": 0.0119,
|
| 1497 |
+
"step": 1830
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 28.307692307692307,
|
| 1501 |
+
"grad_norm": 0.05891846865415573,
|
| 1502 |
+
"learning_rate": 8.676923076923078e-06,
|
| 1503 |
+
"loss": 0.0456,
|
| 1504 |
+
"step": 1840
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 28.46153846153846,
|
| 1508 |
+
"grad_norm": 0.05940578132867813,
|
| 1509 |
+
"learning_rate": 8.615384615384617e-06,
|
| 1510 |
+
"loss": 0.0541,
|
| 1511 |
+
"step": 1850
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 28.615384615384617,
|
| 1515 |
+
"grad_norm": 0.06775704026222229,
|
| 1516 |
+
"learning_rate": 8.553846153846156e-06,
|
| 1517 |
+
"loss": 0.0162,
|
| 1518 |
+
"step": 1860
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 28.76923076923077,
|
| 1522 |
+
"grad_norm": 0.058256104588508606,
|
| 1523 |
+
"learning_rate": 8.492307692307693e-06,
|
| 1524 |
+
"loss": 0.0354,
|
| 1525 |
+
"step": 1870
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 28.923076923076923,
|
| 1529 |
+
"grad_norm": 1.286210060119629,
|
| 1530 |
+
"learning_rate": 8.430769230769231e-06,
|
| 1531 |
+
"loss": 0.0285,
|
| 1532 |
+
"step": 1880
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 29.0,
|
| 1536 |
+
"eval_accuracy": 0.9774436090225563,
|
| 1537 |
+
"eval_loss": 0.041793130338191986,
|
| 1538 |
+
"eval_runtime": 0.6391,
|
| 1539 |
+
"eval_samples_per_second": 208.111,
|
| 1540 |
+
"eval_steps_per_second": 14.083,
|
| 1541 |
+
"step": 1885
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 29.076923076923077,
|
| 1545 |
+
"grad_norm": 2.135648727416992,
|
| 1546 |
+
"learning_rate": 8.36923076923077e-06,
|
| 1547 |
+
"loss": 0.0197,
|
| 1548 |
+
"step": 1890
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 29.23076923076923,
|
| 1552 |
+
"grad_norm": 1.5485310554504395,
|
| 1553 |
+
"learning_rate": 8.307692307692309e-06,
|
| 1554 |
+
"loss": 0.0129,
|
| 1555 |
+
"step": 1900
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 29.384615384615383,
|
| 1559 |
+
"grad_norm": 1.2594960927963257,
|
| 1560 |
+
"learning_rate": 8.246153846153848e-06,
|
| 1561 |
+
"loss": 0.0964,
|
| 1562 |
+
"step": 1910
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 29.53846153846154,
|
| 1566 |
+
"grad_norm": 0.13048137724399567,
|
| 1567 |
+
"learning_rate": 8.184615384615385e-06,
|
| 1568 |
+
"loss": 0.0111,
|
| 1569 |
+
"step": 1920
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 29.692307692307693,
|
| 1573 |
+
"grad_norm": 0.38255247473716736,
|
| 1574 |
+
"learning_rate": 8.123076923076924e-06,
|
| 1575 |
+
"loss": 0.0292,
|
| 1576 |
+
"step": 1930
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 29.846153846153847,
|
| 1580 |
+
"grad_norm": 2.1401822566986084,
|
| 1581 |
+
"learning_rate": 8.061538461538463e-06,
|
| 1582 |
+
"loss": 0.0417,
|
| 1583 |
+
"step": 1940
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 30.0,
|
| 1587 |
+
"grad_norm": 0.2889564633369446,
|
| 1588 |
+
"learning_rate": 8.000000000000001e-06,
|
| 1589 |
+
"loss": 0.0892,
|
| 1590 |
+
"step": 1950
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 30.0,
|
| 1594 |
+
"eval_accuracy": 0.9924812030075187,
|
| 1595 |
+
"eval_loss": 0.020448315888643265,
|
| 1596 |
+
"eval_runtime": 0.5776,
|
| 1597 |
+
"eval_samples_per_second": 230.273,
|
| 1598 |
+
"eval_steps_per_second": 15.582,
|
| 1599 |
+
"step": 1950
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 30.153846153846153,
|
| 1603 |
+
"grad_norm": 3.4330976009368896,
|
| 1604 |
+
"learning_rate": 7.93846153846154e-06,
|
| 1605 |
+
"loss": 0.0466,
|
| 1606 |
+
"step": 1960
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 30.307692307692307,
|
| 1610 |
+
"grad_norm": 0.05678678676486015,
|
| 1611 |
+
"learning_rate": 7.876923076923077e-06,
|
| 1612 |
+
"loss": 0.0701,
|
| 1613 |
+
"step": 1970
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 30.46153846153846,
|
| 1617 |
+
"grad_norm": 0.05310628563165665,
|
| 1618 |
+
"learning_rate": 7.815384615384616e-06,
|
| 1619 |
+
"loss": 0.0118,
|
| 1620 |
+
"step": 1980
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 30.615384615384617,
|
| 1624 |
+
"grad_norm": 0.07134439796209335,
|
| 1625 |
+
"learning_rate": 7.753846153846155e-06,
|
| 1626 |
+
"loss": 0.0412,
|
| 1627 |
+
"step": 1990
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 30.76923076923077,
|
| 1631 |
+
"grad_norm": 0.06206020340323448,
|
| 1632 |
+
"learning_rate": 7.692307692307694e-06,
|
| 1633 |
+
"loss": 0.0254,
|
| 1634 |
+
"step": 2000
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 30.923076923076923,
|
| 1638 |
+
"grad_norm": 0.0549427792429924,
|
| 1639 |
+
"learning_rate": 7.630769230769232e-06,
|
| 1640 |
+
"loss": 0.0371,
|
| 1641 |
+
"step": 2010
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 31.0,
|
| 1645 |
+
"eval_accuracy": 0.9849624060150376,
|
| 1646 |
+
"eval_loss": 0.03390338271856308,
|
| 1647 |
+
"eval_runtime": 0.6391,
|
| 1648 |
+
"eval_samples_per_second": 208.12,
|
| 1649 |
+
"eval_steps_per_second": 14.083,
|
| 1650 |
+
"step": 2015
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 31.076923076923077,
|
| 1654 |
+
"grad_norm": 0.05257405340671539,
|
| 1655 |
+
"learning_rate": 7.5692307692307695e-06,
|
| 1656 |
+
"loss": 0.0119,
|
| 1657 |
+
"step": 2020
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 31.23076923076923,
|
| 1661 |
+
"grad_norm": 1.0072262287139893,
|
| 1662 |
+
"learning_rate": 7.507692307692308e-06,
|
| 1663 |
+
"loss": 0.0131,
|
| 1664 |
+
"step": 2030
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 31.384615384615383,
|
| 1668 |
+
"grad_norm": 0.054051704704761505,
|
| 1669 |
+
"learning_rate": 7.446153846153846e-06,
|
| 1670 |
+
"loss": 0.0309,
|
| 1671 |
+
"step": 2040
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 31.53846153846154,
|
| 1675 |
+
"grad_norm": 0.059842586517333984,
|
| 1676 |
+
"learning_rate": 7.384615384615386e-06,
|
| 1677 |
+
"loss": 0.0699,
|
| 1678 |
+
"step": 2050
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 31.692307692307693,
|
| 1682 |
+
"grad_norm": 0.4310505986213684,
|
| 1683 |
+
"learning_rate": 7.323076923076924e-06,
|
| 1684 |
+
"loss": 0.0105,
|
| 1685 |
+
"step": 2060
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 31.846153846153847,
|
| 1689 |
+
"grad_norm": 0.05525004491209984,
|
| 1690 |
+
"learning_rate": 7.261538461538462e-06,
|
| 1691 |
+
"loss": 0.0654,
|
| 1692 |
+
"step": 2070
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 32.0,
|
| 1696 |
+
"grad_norm": 0.06051107123494148,
|
| 1697 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 1698 |
+
"loss": 0.0105,
|
| 1699 |
+
"step": 2080
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 32.0,
|
| 1703 |
+
"eval_accuracy": 1.0,
|
| 1704 |
+
"eval_loss": 0.01434730738401413,
|
| 1705 |
+
"eval_runtime": 0.5855,
|
| 1706 |
+
"eval_samples_per_second": 227.144,
|
| 1707 |
+
"eval_steps_per_second": 15.371,
|
| 1708 |
+
"step": 2080
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 32.15384615384615,
|
| 1712 |
+
"grad_norm": 0.04868883639574051,
|
| 1713 |
+
"learning_rate": 7.1384615384615385e-06,
|
| 1714 |
+
"loss": 0.032,
|
| 1715 |
+
"step": 2090
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 32.30769230769231,
|
| 1719 |
+
"grad_norm": 0.10859151929616928,
|
| 1720 |
+
"learning_rate": 7.076923076923078e-06,
|
| 1721 |
+
"loss": 0.084,
|
| 1722 |
+
"step": 2100
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 32.46153846153846,
|
| 1726 |
+
"grad_norm": 0.05709298700094223,
|
| 1727 |
+
"learning_rate": 7.015384615384616e-06,
|
| 1728 |
+
"loss": 0.0189,
|
| 1729 |
+
"step": 2110
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 32.61538461538461,
|
| 1733 |
+
"grad_norm": 0.08583523333072662,
|
| 1734 |
+
"learning_rate": 6.953846153846154e-06,
|
| 1735 |
+
"loss": 0.0124,
|
| 1736 |
+
"step": 2120
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 32.76923076923077,
|
| 1740 |
+
"grad_norm": 0.7491576671600342,
|
| 1741 |
+
"learning_rate": 6.892307692307693e-06,
|
| 1742 |
+
"loss": 0.0213,
|
| 1743 |
+
"step": 2130
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 32.92307692307692,
|
| 1747 |
+
"grad_norm": 0.3305934965610504,
|
| 1748 |
+
"learning_rate": 6.830769230769231e-06,
|
| 1749 |
+
"loss": 0.0563,
|
| 1750 |
+
"step": 2140
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 33.0,
|
| 1754 |
+
"eval_accuracy": 1.0,
|
| 1755 |
+
"eval_loss": 0.014035705476999283,
|
| 1756 |
+
"eval_runtime": 0.6445,
|
| 1757 |
+
"eval_samples_per_second": 206.373,
|
| 1758 |
+
"eval_steps_per_second": 13.965,
|
| 1759 |
+
"step": 2145
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 33.07692307692308,
|
| 1763 |
+
"grad_norm": 0.049748744815588,
|
| 1764 |
+
"learning_rate": 6.76923076923077e-06,
|
| 1765 |
+
"loss": 0.0104,
|
| 1766 |
+
"step": 2150
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 33.23076923076923,
|
| 1770 |
+
"grad_norm": 0.05033630132675171,
|
| 1771 |
+
"learning_rate": 6.707692307692308e-06,
|
| 1772 |
+
"loss": 0.0548,
|
| 1773 |
+
"step": 2160
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 33.38461538461539,
|
| 1777 |
+
"grad_norm": 0.054612692445516586,
|
| 1778 |
+
"learning_rate": 6.646153846153846e-06,
|
| 1779 |
+
"loss": 0.0356,
|
| 1780 |
+
"step": 2170
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 33.53846153846154,
|
| 1784 |
+
"grad_norm": 0.05341866612434387,
|
| 1785 |
+
"learning_rate": 6.584615384615385e-06,
|
| 1786 |
+
"loss": 0.0133,
|
| 1787 |
+
"step": 2180
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 33.69230769230769,
|
| 1791 |
+
"grad_norm": 0.04895515367388725,
|
| 1792 |
+
"learning_rate": 6.523076923076923e-06,
|
| 1793 |
+
"loss": 0.0213,
|
| 1794 |
+
"step": 2190
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 33.84615384615385,
|
| 1798 |
+
"grad_norm": 0.063107430934906,
|
| 1799 |
+
"learning_rate": 6.461538461538463e-06,
|
| 1800 |
+
"loss": 0.0112,
|
| 1801 |
+
"step": 2200
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 34.0,
|
| 1805 |
+
"grad_norm": 7.929271221160889,
|
| 1806 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 1807 |
+
"loss": 0.0573,
|
| 1808 |
+
"step": 2210
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 34.0,
|
| 1812 |
+
"eval_accuracy": 1.0,
|
| 1813 |
+
"eval_loss": 0.010156131349503994,
|
| 1814 |
+
"eval_runtime": 0.6308,
|
| 1815 |
+
"eval_samples_per_second": 210.855,
|
| 1816 |
+
"eval_steps_per_second": 14.268,
|
| 1817 |
+
"step": 2210
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 34.15384615384615,
|
| 1821 |
+
"grad_norm": 0.11311448365449905,
|
| 1822 |
+
"learning_rate": 6.3384615384615385e-06,
|
| 1823 |
+
"loss": 0.0271,
|
| 1824 |
+
"step": 2220
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 34.30769230769231,
|
| 1828 |
+
"grad_norm": 0.17935284972190857,
|
| 1829 |
+
"learning_rate": 6.276923076923077e-06,
|
| 1830 |
+
"loss": 0.0444,
|
| 1831 |
+
"step": 2230
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 34.46153846153846,
|
| 1835 |
+
"grad_norm": 0.07903819531202316,
|
| 1836 |
+
"learning_rate": 6.215384615384615e-06,
|
| 1837 |
+
"loss": 0.039,
|
| 1838 |
+
"step": 2240
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 34.61538461538461,
|
| 1842 |
+
"grad_norm": 0.07042822986841202,
|
| 1843 |
+
"learning_rate": 6.153846153846155e-06,
|
| 1844 |
+
"loss": 0.0617,
|
| 1845 |
+
"step": 2250
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 34.76923076923077,
|
| 1849 |
+
"grad_norm": 0.05035420507192612,
|
| 1850 |
+
"learning_rate": 6.092307692307693e-06,
|
| 1851 |
+
"loss": 0.0505,
|
| 1852 |
+
"step": 2260
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 34.92307692307692,
|
| 1856 |
+
"grad_norm": 0.04776820167899132,
|
| 1857 |
+
"learning_rate": 6.030769230769231e-06,
|
| 1858 |
+
"loss": 0.0409,
|
| 1859 |
+
"step": 2270
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 35.0,
|
| 1863 |
+
"eval_accuracy": 1.0,
|
| 1864 |
+
"eval_loss": 0.009572061710059643,
|
| 1865 |
+
"eval_runtime": 0.6399,
|
| 1866 |
+
"eval_samples_per_second": 207.852,
|
| 1867 |
+
"eval_steps_per_second": 14.065,
|
| 1868 |
+
"step": 2275
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 35.07692307692308,
|
| 1872 |
+
"grad_norm": 20.547210693359375,
|
| 1873 |
+
"learning_rate": 5.9692307692307695e-06,
|
| 1874 |
+
"loss": 0.0212,
|
| 1875 |
+
"step": 2280
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 35.23076923076923,
|
| 1879 |
+
"grad_norm": 16.86725616455078,
|
| 1880 |
+
"learning_rate": 5.907692307692308e-06,
|
| 1881 |
+
"loss": 0.0421,
|
| 1882 |
+
"step": 2290
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 35.38461538461539,
|
| 1886 |
+
"grad_norm": 1.5179917812347412,
|
| 1887 |
+
"learning_rate": 5.846153846153847e-06,
|
| 1888 |
+
"loss": 0.1226,
|
| 1889 |
+
"step": 2300
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 35.53846153846154,
|
| 1893 |
+
"grad_norm": 0.045909151434898376,
|
| 1894 |
+
"learning_rate": 5.784615384615385e-06,
|
| 1895 |
+
"loss": 0.0092,
|
| 1896 |
+
"step": 2310
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 35.69230769230769,
|
| 1900 |
+
"grad_norm": 0.04946780204772949,
|
| 1901 |
+
"learning_rate": 5.723076923076923e-06,
|
| 1902 |
+
"loss": 0.0505,
|
| 1903 |
+
"step": 2320
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 35.84615384615385,
|
| 1907 |
+
"grad_norm": 0.055896684527397156,
|
| 1908 |
+
"learning_rate": 5.661538461538462e-06,
|
| 1909 |
+
"loss": 0.0159,
|
| 1910 |
+
"step": 2330
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 36.0,
|
| 1914 |
+
"grad_norm": 0.04570392891764641,
|
| 1915 |
+
"learning_rate": 5.600000000000001e-06,
|
| 1916 |
+
"loss": 0.0523,
|
| 1917 |
+
"step": 2340
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 36.0,
|
| 1921 |
+
"eval_accuracy": 0.9924812030075187,
|
| 1922 |
+
"eval_loss": 0.01487450860440731,
|
| 1923 |
+
"eval_runtime": 0.6368,
|
| 1924 |
+
"eval_samples_per_second": 208.852,
|
| 1925 |
+
"eval_steps_per_second": 14.133,
|
| 1926 |
+
"step": 2340
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 36.15384615384615,
|
| 1930 |
+
"grad_norm": 0.06980939954519272,
|
| 1931 |
+
"learning_rate": 5.538461538461539e-06,
|
| 1932 |
+
"loss": 0.0219,
|
| 1933 |
+
"step": 2350
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 36.30769230769231,
|
| 1937 |
+
"grad_norm": 0.05014393478631973,
|
| 1938 |
+
"learning_rate": 5.476923076923077e-06,
|
| 1939 |
+
"loss": 0.0352,
|
| 1940 |
+
"step": 2360
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 36.46153846153846,
|
| 1944 |
+
"grad_norm": 0.04635272175073624,
|
| 1945 |
+
"learning_rate": 5.415384615384615e-06,
|
| 1946 |
+
"loss": 0.0328,
|
| 1947 |
+
"step": 2370
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 36.61538461538461,
|
| 1951 |
+
"grad_norm": 0.04787183925509453,
|
| 1952 |
+
"learning_rate": 5.353846153846154e-06,
|
| 1953 |
+
"loss": 0.0106,
|
| 1954 |
+
"step": 2380
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 36.76923076923077,
|
| 1958 |
+
"grad_norm": 0.05444607138633728,
|
| 1959 |
+
"learning_rate": 5.292307692307693e-06,
|
| 1960 |
+
"loss": 0.0443,
|
| 1961 |
+
"step": 2390
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 36.92307692307692,
|
| 1965 |
+
"grad_norm": 0.046256501227617264,
|
| 1966 |
+
"learning_rate": 5.230769230769232e-06,
|
| 1967 |
+
"loss": 0.0131,
|
| 1968 |
+
"step": 2400
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 37.0,
|
| 1972 |
+
"eval_accuracy": 0.9924812030075187,
|
| 1973 |
+
"eval_loss": 0.0196556244045496,
|
| 1974 |
+
"eval_runtime": 0.641,
|
| 1975 |
+
"eval_samples_per_second": 207.498,
|
| 1976 |
+
"eval_steps_per_second": 14.041,
|
| 1977 |
+
"step": 2405
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 37.07692307692308,
|
| 1981 |
+
"grad_norm": 0.045623164623975754,
|
| 1982 |
+
"learning_rate": 5.16923076923077e-06,
|
| 1983 |
+
"loss": 0.0112,
|
| 1984 |
+
"step": 2410
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 37.23076923076923,
|
| 1988 |
+
"grad_norm": 0.04630188271403313,
|
| 1989 |
+
"learning_rate": 5.1076923076923075e-06,
|
| 1990 |
+
"loss": 0.0129,
|
| 1991 |
+
"step": 2420
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 37.38461538461539,
|
| 1995 |
+
"grad_norm": 1.0303974151611328,
|
| 1996 |
+
"learning_rate": 5.046153846153846e-06,
|
| 1997 |
+
"loss": 0.0523,
|
| 1998 |
+
"step": 2430
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 37.53846153846154,
|
| 2002 |
+
"grad_norm": 0.8218058347702026,
|
| 2003 |
+
"learning_rate": 4.984615384615385e-06,
|
| 2004 |
+
"loss": 0.0532,
|
| 2005 |
+
"step": 2440
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 37.69230769230769,
|
| 2009 |
+
"grad_norm": 1.5203208923339844,
|
| 2010 |
+
"learning_rate": 4.923076923076924e-06,
|
| 2011 |
+
"loss": 0.0345,
|
| 2012 |
+
"step": 2450
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 37.84615384615385,
|
| 2016 |
+
"grad_norm": 0.05802327021956444,
|
| 2017 |
+
"learning_rate": 4.861538461538462e-06,
|
| 2018 |
+
"loss": 0.0251,
|
| 2019 |
+
"step": 2460
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 38.0,
|
| 2023 |
+
"grad_norm": 0.06536999344825745,
|
| 2024 |
+
"learning_rate": 4.800000000000001e-06,
|
| 2025 |
+
"loss": 0.0329,
|
| 2026 |
+
"step": 2470
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 38.0,
|
| 2030 |
+
"eval_accuracy": 1.0,
|
| 2031 |
+
"eval_loss": 0.010934116318821907,
|
| 2032 |
+
"eval_runtime": 0.6397,
|
| 2033 |
+
"eval_samples_per_second": 207.921,
|
| 2034 |
+
"eval_steps_per_second": 14.07,
|
| 2035 |
+
"step": 2470
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 38.15384615384615,
|
| 2039 |
+
"grad_norm": 0.07708264887332916,
|
| 2040 |
+
"learning_rate": 4.738461538461539e-06,
|
| 2041 |
+
"loss": 0.0339,
|
| 2042 |
+
"step": 2480
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 38.30769230769231,
|
| 2046 |
+
"grad_norm": 0.05018337070941925,
|
| 2047 |
+
"learning_rate": 4.676923076923077e-06,
|
| 2048 |
+
"loss": 0.0371,
|
| 2049 |
+
"step": 2490
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 38.46153846153846,
|
| 2053 |
+
"grad_norm": 2.005122423171997,
|
| 2054 |
+
"learning_rate": 4.615384615384616e-06,
|
| 2055 |
+
"loss": 0.0493,
|
| 2056 |
+
"step": 2500
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 38.61538461538461,
|
| 2060 |
+
"grad_norm": 0.04191539064049721,
|
| 2061 |
+
"learning_rate": 4.553846153846154e-06,
|
| 2062 |
+
"loss": 0.056,
|
| 2063 |
+
"step": 2510
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 38.76923076923077,
|
| 2067 |
+
"grad_norm": 0.08912540227174759,
|
| 2068 |
+
"learning_rate": 4.492307692307693e-06,
|
| 2069 |
+
"loss": 0.0675,
|
| 2070 |
+
"step": 2520
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 38.92307692307692,
|
| 2074 |
+
"grad_norm": 4.123304843902588,
|
| 2075 |
+
"learning_rate": 4.430769230769232e-06,
|
| 2076 |
+
"loss": 0.0577,
|
| 2077 |
+
"step": 2530
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 39.0,
|
| 2081 |
+
"eval_accuracy": 1.0,
|
| 2082 |
+
"eval_loss": 0.00963473692536354,
|
| 2083 |
+
"eval_runtime": 0.6269,
|
| 2084 |
+
"eval_samples_per_second": 212.155,
|
| 2085 |
+
"eval_steps_per_second": 14.356,
|
| 2086 |
+
"step": 2535
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 39.07692307692308,
|
| 2090 |
+
"grad_norm": 0.12956681847572327,
|
| 2091 |
+
"learning_rate": 4.36923076923077e-06,
|
| 2092 |
+
"loss": 0.0348,
|
| 2093 |
+
"step": 2540
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 39.23076923076923,
|
| 2097 |
+
"grad_norm": 0.0469551756978035,
|
| 2098 |
+
"learning_rate": 4.307692307692308e-06,
|
| 2099 |
+
"loss": 0.047,
|
| 2100 |
+
"step": 2550
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 39.38461538461539,
|
| 2104 |
+
"grad_norm": 0.0566897876560688,
|
| 2105 |
+
"learning_rate": 4.246153846153846e-06,
|
| 2106 |
+
"loss": 0.0305,
|
| 2107 |
+
"step": 2560
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 39.53846153846154,
|
| 2111 |
+
"grad_norm": 0.04538924619555473,
|
| 2112 |
+
"learning_rate": 4.184615384615385e-06,
|
| 2113 |
+
"loss": 0.0083,
|
| 2114 |
+
"step": 2570
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 39.69230769230769,
|
| 2118 |
+
"grad_norm": 0.1393657773733139,
|
| 2119 |
+
"learning_rate": 4.123076923076924e-06,
|
| 2120 |
+
"loss": 0.0087,
|
| 2121 |
+
"step": 2580
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 39.84615384615385,
|
| 2125 |
+
"grad_norm": 0.04170211777091026,
|
| 2126 |
+
"learning_rate": 4.061538461538462e-06,
|
| 2127 |
+
"loss": 0.008,
|
| 2128 |
+
"step": 2590
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 40.0,
|
| 2132 |
+
"grad_norm": 0.04205217584967613,
|
| 2133 |
+
"learning_rate": 4.000000000000001e-06,
|
| 2134 |
+
"loss": 0.0085,
|
| 2135 |
+
"step": 2600
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 40.0,
|
| 2139 |
+
"eval_accuracy": 0.9924812030075187,
|
| 2140 |
+
"eval_loss": 0.014666187576949596,
|
| 2141 |
+
"eval_runtime": 0.5786,
|
| 2142 |
+
"eval_samples_per_second": 229.849,
|
| 2143 |
+
"eval_steps_per_second": 15.554,
|
| 2144 |
+
"step": 2600
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 40.15384615384615,
|
| 2148 |
+
"grad_norm": 0.04466895014047623,
|
| 2149 |
+
"learning_rate": 3.938461538461539e-06,
|
| 2150 |
+
"loss": 0.0376,
|
| 2151 |
+
"step": 2610
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 40.30769230769231,
|
| 2155 |
+
"grad_norm": 0.04949569329619408,
|
| 2156 |
+
"learning_rate": 3.876923076923077e-06,
|
| 2157 |
+
"loss": 0.0621,
|
| 2158 |
+
"step": 2620
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 40.46153846153846,
|
| 2162 |
+
"grad_norm": 0.0461997464299202,
|
| 2163 |
+
"learning_rate": 3.815384615384616e-06,
|
| 2164 |
+
"loss": 0.0107,
|
| 2165 |
+
"step": 2630
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 40.61538461538461,
|
| 2169 |
+
"grad_norm": 0.048004575073719025,
|
| 2170 |
+
"learning_rate": 3.753846153846154e-06,
|
| 2171 |
+
"loss": 0.0093,
|
| 2172 |
+
"step": 2640
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 40.76923076923077,
|
| 2176 |
+
"grad_norm": 0.04209740087389946,
|
| 2177 |
+
"learning_rate": 3.692307692307693e-06,
|
| 2178 |
+
"loss": 0.0342,
|
| 2179 |
+
"step": 2650
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 40.92307692307692,
|
| 2183 |
+
"grad_norm": 0.15323477983474731,
|
| 2184 |
+
"learning_rate": 3.630769230769231e-06,
|
| 2185 |
+
"loss": 0.0618,
|
| 2186 |
+
"step": 2660
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 41.0,
|
| 2190 |
+
"eval_accuracy": 1.0,
|
| 2191 |
+
"eval_loss": 0.009433195926249027,
|
| 2192 |
+
"eval_runtime": 0.6376,
|
| 2193 |
+
"eval_samples_per_second": 208.608,
|
| 2194 |
+
"eval_steps_per_second": 14.116,
|
| 2195 |
+
"step": 2665
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 41.07692307692308,
|
| 2199 |
+
"grad_norm": 0.040465518832206726,
|
| 2200 |
+
"learning_rate": 3.5692307692307692e-06,
|
| 2201 |
+
"loss": 0.0079,
|
| 2202 |
+
"step": 2670
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 41.23076923076923,
|
| 2206 |
+
"grad_norm": 0.06956275552511215,
|
| 2207 |
+
"learning_rate": 3.507692307692308e-06,
|
| 2208 |
+
"loss": 0.0278,
|
| 2209 |
+
"step": 2680
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 41.38461538461539,
|
| 2213 |
+
"grad_norm": 0.04409582540392876,
|
| 2214 |
+
"learning_rate": 3.4461538461538464e-06,
|
| 2215 |
+
"loss": 0.0079,
|
| 2216 |
+
"step": 2690
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 41.53846153846154,
|
| 2220 |
+
"grad_norm": 13.665828704833984,
|
| 2221 |
+
"learning_rate": 3.384615384615385e-06,
|
| 2222 |
+
"loss": 0.0187,
|
| 2223 |
+
"step": 2700
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 41.69230769230769,
|
| 2227 |
+
"grad_norm": 1.3187448978424072,
|
| 2228 |
+
"learning_rate": 3.323076923076923e-06,
|
| 2229 |
+
"loss": 0.1204,
|
| 2230 |
+
"step": 2710
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 41.84615384615385,
|
| 2234 |
+
"grad_norm": 0.04300126060843468,
|
| 2235 |
+
"learning_rate": 3.2615384615384615e-06,
|
| 2236 |
+
"loss": 0.0198,
|
| 2237 |
+
"step": 2720
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 42.0,
|
| 2241 |
+
"grad_norm": 0.0438438281416893,
|
| 2242 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 2243 |
+
"loss": 0.0847,
|
| 2244 |
+
"step": 2730
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 42.0,
|
| 2248 |
+
"eval_accuracy": 0.9924812030075187,
|
| 2249 |
+
"eval_loss": 0.019689705222845078,
|
| 2250 |
+
"eval_runtime": 0.634,
|
| 2251 |
+
"eval_samples_per_second": 209.767,
|
| 2252 |
+
"eval_steps_per_second": 14.195,
|
| 2253 |
+
"step": 2730
|
| 2254 |
+
},
|
| 2255 |
+
{
|
| 2256 |
+
"epoch": 42.15384615384615,
|
| 2257 |
+
"grad_norm": 0.055344920605421066,
|
| 2258 |
+
"learning_rate": 3.1384615384615386e-06,
|
| 2259 |
+
"loss": 0.0343,
|
| 2260 |
+
"step": 2740
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"epoch": 42.30769230769231,
|
| 2264 |
+
"grad_norm": 0.04840540513396263,
|
| 2265 |
+
"learning_rate": 3.0769230769230774e-06,
|
| 2266 |
+
"loss": 0.0091,
|
| 2267 |
+
"step": 2750
|
| 2268 |
+
},
|
| 2269 |
+
{
|
| 2270 |
+
"epoch": 42.46153846153846,
|
| 2271 |
+
"grad_norm": 0.0416325181722641,
|
| 2272 |
+
"learning_rate": 3.0153846153846154e-06,
|
| 2273 |
+
"loss": 0.0379,
|
| 2274 |
+
"step": 2760
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 42.61538461538461,
|
| 2278 |
+
"grad_norm": 0.04047630727291107,
|
| 2279 |
+
"learning_rate": 2.953846153846154e-06,
|
| 2280 |
+
"loss": 0.0344,
|
| 2281 |
+
"step": 2770
|
| 2282 |
+
},
|
| 2283 |
+
{
|
| 2284 |
+
"epoch": 42.76923076923077,
|
| 2285 |
+
"grad_norm": 0.039694271981716156,
|
| 2286 |
+
"learning_rate": 2.8923076923076925e-06,
|
| 2287 |
+
"loss": 0.0556,
|
| 2288 |
+
"step": 2780
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 42.92307692307692,
|
| 2292 |
+
"grad_norm": 0.0425509512424469,
|
| 2293 |
+
"learning_rate": 2.830769230769231e-06,
|
| 2294 |
+
"loss": 0.0291,
|
| 2295 |
+
"step": 2790
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 43.0,
|
| 2299 |
+
"eval_accuracy": 1.0,
|
| 2300 |
+
"eval_loss": 0.008893251419067383,
|
| 2301 |
+
"eval_runtime": 0.6271,
|
| 2302 |
+
"eval_samples_per_second": 212.079,
|
| 2303 |
+
"eval_steps_per_second": 14.351,
|
| 2304 |
+
"step": 2795
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 43.07692307692308,
|
| 2308 |
+
"grad_norm": 0.04988468438386917,
|
| 2309 |
+
"learning_rate": 2.7692307692307697e-06,
|
| 2310 |
+
"loss": 0.0369,
|
| 2311 |
+
"step": 2800
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 43.23076923076923,
|
| 2315 |
+
"grad_norm": 0.07137361913919449,
|
| 2316 |
+
"learning_rate": 2.7076923076923076e-06,
|
| 2317 |
+
"loss": 0.0434,
|
| 2318 |
+
"step": 2810
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 43.38461538461539,
|
| 2322 |
+
"grad_norm": 7.0051679611206055,
|
| 2323 |
+
"learning_rate": 2.6461538461538464e-06,
|
| 2324 |
+
"loss": 0.0291,
|
| 2325 |
+
"step": 2820
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 43.53846153846154,
|
| 2329 |
+
"grad_norm": 0.045469850301742554,
|
| 2330 |
+
"learning_rate": 2.584615384615385e-06,
|
| 2331 |
+
"loss": 0.0338,
|
| 2332 |
+
"step": 2830
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 43.69230769230769,
|
| 2336 |
+
"grad_norm": 0.08003593236207962,
|
| 2337 |
+
"learning_rate": 2.523076923076923e-06,
|
| 2338 |
+
"loss": 0.0099,
|
| 2339 |
+
"step": 2840
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 43.84615384615385,
|
| 2343 |
+
"grad_norm": 0.04380409047007561,
|
| 2344 |
+
"learning_rate": 2.461538461538462e-06,
|
| 2345 |
+
"loss": 0.0111,
|
| 2346 |
+
"step": 2850
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 44.0,
|
| 2350 |
+
"grad_norm": 13.31029224395752,
|
| 2351 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 2352 |
+
"loss": 0.0568,
|
| 2353 |
+
"step": 2860
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 44.0,
|
| 2357 |
+
"eval_accuracy": 1.0,
|
| 2358 |
+
"eval_loss": 0.008692615665495396,
|
| 2359 |
+
"eval_runtime": 0.585,
|
| 2360 |
+
"eval_samples_per_second": 227.347,
|
| 2361 |
+
"eval_steps_per_second": 15.384,
|
| 2362 |
+
"step": 2860
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 44.15384615384615,
|
| 2366 |
+
"grad_norm": 0.04654600843787193,
|
| 2367 |
+
"learning_rate": 2.3384615384615387e-06,
|
| 2368 |
+
"loss": 0.0087,
|
| 2369 |
+
"step": 2870
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 44.30769230769231,
|
| 2373 |
+
"grad_norm": 7.452500343322754,
|
| 2374 |
+
"learning_rate": 2.276923076923077e-06,
|
| 2375 |
+
"loss": 0.0108,
|
| 2376 |
+
"step": 2880
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 44.46153846153846,
|
| 2380 |
+
"grad_norm": 13.458589553833008,
|
| 2381 |
+
"learning_rate": 2.215384615384616e-06,
|
| 2382 |
+
"loss": 0.0274,
|
| 2383 |
+
"step": 2890
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 44.61538461538461,
|
| 2387 |
+
"grad_norm": 0.044014327228069305,
|
| 2388 |
+
"learning_rate": 2.153846153846154e-06,
|
| 2389 |
+
"loss": 0.0299,
|
| 2390 |
+
"step": 2900
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 44.76923076923077,
|
| 2394 |
+
"grad_norm": 0.041860181838274,
|
| 2395 |
+
"learning_rate": 2.0923076923076926e-06,
|
| 2396 |
+
"loss": 0.0112,
|
| 2397 |
+
"step": 2910
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 44.92307692307692,
|
| 2401 |
+
"grad_norm": 0.04078350216150284,
|
| 2402 |
+
"learning_rate": 2.030769230769231e-06,
|
| 2403 |
+
"loss": 0.0077,
|
| 2404 |
+
"step": 2920
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 45.0,
|
| 2408 |
+
"eval_accuracy": 1.0,
|
| 2409 |
+
"eval_loss": 0.010402214713394642,
|
| 2410 |
+
"eval_runtime": 0.6383,
|
| 2411 |
+
"eval_samples_per_second": 208.376,
|
| 2412 |
+
"eval_steps_per_second": 14.101,
|
| 2413 |
+
"step": 2925
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 45.07692307692308,
|
| 2417 |
+
"grad_norm": 0.7966273427009583,
|
| 2418 |
+
"learning_rate": 1.9692307692307693e-06,
|
| 2419 |
+
"loss": 0.0432,
|
| 2420 |
+
"step": 2930
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 45.23076923076923,
|
| 2424 |
+
"grad_norm": 0.04070662334561348,
|
| 2425 |
+
"learning_rate": 1.907692307692308e-06,
|
| 2426 |
+
"loss": 0.0549,
|
| 2427 |
+
"step": 2940
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 45.38461538461539,
|
| 2431 |
+
"grad_norm": 0.042289331555366516,
|
| 2432 |
+
"learning_rate": 1.8461538461538465e-06,
|
| 2433 |
+
"loss": 0.0364,
|
| 2434 |
+
"step": 2950
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 45.53846153846154,
|
| 2438 |
+
"grad_norm": 0.04655339941382408,
|
| 2439 |
+
"learning_rate": 1.7846153846153846e-06,
|
| 2440 |
+
"loss": 0.0114,
|
| 2441 |
+
"step": 2960
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 45.69230769230769,
|
| 2445 |
+
"grad_norm": 0.04026507958769798,
|
| 2446 |
+
"learning_rate": 1.7230769230769232e-06,
|
| 2447 |
+
"loss": 0.0078,
|
| 2448 |
+
"step": 2970
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 45.84615384615385,
|
| 2452 |
+
"grad_norm": 0.048073675483465195,
|
| 2453 |
+
"learning_rate": 1.6615384615384616e-06,
|
| 2454 |
+
"loss": 0.0155,
|
| 2455 |
+
"step": 2980
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 46.0,
|
| 2459 |
+
"grad_norm": 0.1250167340040207,
|
| 2460 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 2461 |
+
"loss": 0.008,
|
| 2462 |
+
"step": 2990
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 46.0,
|
| 2466 |
+
"eval_accuracy": 1.0,
|
| 2467 |
+
"eval_loss": 0.013788605108857155,
|
| 2468 |
+
"eval_runtime": 0.5799,
|
| 2469 |
+
"eval_samples_per_second": 229.355,
|
| 2470 |
+
"eval_steps_per_second": 15.52,
|
| 2471 |
+
"step": 2990
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 46.15384615384615,
|
| 2475 |
+
"grad_norm": 0.3202461302280426,
|
| 2476 |
+
"learning_rate": 1.5384615384615387e-06,
|
| 2477 |
+
"loss": 0.0352,
|
| 2478 |
+
"step": 3000
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 46.30769230769231,
|
| 2482 |
+
"grad_norm": 2.8946588039398193,
|
| 2483 |
+
"learning_rate": 1.476923076923077e-06,
|
| 2484 |
+
"loss": 0.0196,
|
| 2485 |
+
"step": 3010
|
| 2486 |
+
},
|
| 2487 |
+
{
|
| 2488 |
+
"epoch": 46.46153846153846,
|
| 2489 |
+
"grad_norm": 0.0777769535779953,
|
| 2490 |
+
"learning_rate": 1.4153846153846155e-06,
|
| 2491 |
+
"loss": 0.0079,
|
| 2492 |
+
"step": 3020
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 46.61538461538461,
|
| 2496 |
+
"grad_norm": 1.6607468128204346,
|
| 2497 |
+
"learning_rate": 1.3538461538461538e-06,
|
| 2498 |
+
"loss": 0.0102,
|
| 2499 |
+
"step": 3030
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 46.76923076923077,
|
| 2503 |
+
"grad_norm": 0.04541005194187164,
|
| 2504 |
+
"learning_rate": 1.2923076923076924e-06,
|
| 2505 |
+
"loss": 0.0085,
|
| 2506 |
+
"step": 3040
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 46.92307692307692,
|
| 2510 |
+
"grad_norm": 0.041475191712379456,
|
| 2511 |
+
"learning_rate": 1.230769230769231e-06,
|
| 2512 |
+
"loss": 0.0272,
|
| 2513 |
+
"step": 3050
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 47.0,
|
| 2517 |
+
"eval_accuracy": 1.0,
|
| 2518 |
+
"eval_loss": 0.00810349639505148,
|
| 2519 |
+
"eval_runtime": 0.6353,
|
| 2520 |
+
"eval_samples_per_second": 209.365,
|
| 2521 |
+
"eval_steps_per_second": 14.168,
|
| 2522 |
+
"step": 3055
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 47.07692307692308,
|
| 2526 |
+
"grad_norm": 0.048281800001859665,
|
| 2527 |
+
"learning_rate": 1.1692307692307693e-06,
|
| 2528 |
+
"loss": 0.008,
|
| 2529 |
+
"step": 3060
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 47.23076923076923,
|
| 2533 |
+
"grad_norm": 0.03975387290120125,
|
| 2534 |
+
"learning_rate": 1.107692307692308e-06,
|
| 2535 |
+
"loss": 0.0431,
|
| 2536 |
+
"step": 3070
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 47.38461538461539,
|
| 2540 |
+
"grad_norm": 0.040405042469501495,
|
| 2541 |
+
"learning_rate": 1.0461538461538463e-06,
|
| 2542 |
+
"loss": 0.0355,
|
| 2543 |
+
"step": 3080
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 47.53846153846154,
|
| 2547 |
+
"grad_norm": 0.04081344977021217,
|
| 2548 |
+
"learning_rate": 9.846153846153847e-07,
|
| 2549 |
+
"loss": 0.0078,
|
| 2550 |
+
"step": 3090
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 47.69230769230769,
|
| 2554 |
+
"grad_norm": 0.045139458030462265,
|
| 2555 |
+
"learning_rate": 9.230769230769232e-07,
|
| 2556 |
+
"loss": 0.0096,
|
| 2557 |
+
"step": 3100
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 47.84615384615385,
|
| 2561 |
+
"grad_norm": 0.06699339300394058,
|
| 2562 |
+
"learning_rate": 8.615384615384616e-07,
|
| 2563 |
+
"loss": 0.0402,
|
| 2564 |
+
"step": 3110
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 48.0,
|
| 2568 |
+
"grad_norm": 0.04077847674489021,
|
| 2569 |
+
"learning_rate": 8.000000000000001e-07,
|
| 2570 |
+
"loss": 0.008,
|
| 2571 |
+
"step": 3120
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 48.0,
|
| 2575 |
+
"eval_accuracy": 1.0,
|
| 2576 |
+
"eval_loss": 0.008442863821983337,
|
| 2577 |
+
"eval_runtime": 0.6316,
|
| 2578 |
+
"eval_samples_per_second": 210.576,
|
| 2579 |
+
"eval_steps_per_second": 14.25,
|
| 2580 |
+
"step": 3120
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 48.15384615384615,
|
| 2584 |
+
"grad_norm": 0.14330124855041504,
|
| 2585 |
+
"learning_rate": 7.384615384615385e-07,
|
| 2586 |
+
"loss": 0.0396,
|
| 2587 |
+
"step": 3130
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 48.30769230769231,
|
| 2591 |
+
"grad_norm": 0.04051917791366577,
|
| 2592 |
+
"learning_rate": 6.769230769230769e-07,
|
| 2593 |
+
"loss": 0.0613,
|
| 2594 |
+
"step": 3140
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 48.46153846153846,
|
| 2598 |
+
"grad_norm": 0.03945121914148331,
|
| 2599 |
+
"learning_rate": 6.153846153846155e-07,
|
| 2600 |
+
"loss": 0.0092,
|
| 2601 |
+
"step": 3150
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 48.61538461538461,
|
| 2605 |
+
"grad_norm": 0.04850227013230324,
|
| 2606 |
+
"learning_rate": 5.53846153846154e-07,
|
| 2607 |
+
"loss": 0.0395,
|
| 2608 |
+
"step": 3160
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 48.76923076923077,
|
| 2612 |
+
"grad_norm": 0.611132800579071,
|
| 2613 |
+
"learning_rate": 4.923076923076923e-07,
|
| 2614 |
+
"loss": 0.015,
|
| 2615 |
+
"step": 3170
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 48.92307692307692,
|
| 2619 |
+
"grad_norm": 0.039736129343509674,
|
| 2620 |
+
"learning_rate": 4.307692307692308e-07,
|
| 2621 |
+
"loss": 0.0112,
|
| 2622 |
+
"step": 3180
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 49.0,
|
| 2626 |
+
"eval_accuracy": 1.0,
|
| 2627 |
+
"eval_loss": 0.008192874491214752,
|
| 2628 |
+
"eval_runtime": 0.594,
|
| 2629 |
+
"eval_samples_per_second": 223.907,
|
| 2630 |
+
"eval_steps_per_second": 15.152,
|
| 2631 |
+
"step": 3185
|
| 2632 |
+
},
|
| 2633 |
+
{
|
| 2634 |
+
"epoch": 49.07692307692308,
|
| 2635 |
+
"grad_norm": 0.0747324600815773,
|
| 2636 |
+
"learning_rate": 3.6923076923076927e-07,
|
| 2637 |
+
"loss": 0.018,
|
| 2638 |
+
"step": 3190
|
| 2639 |
+
},
|
| 2640 |
+
{
|
| 2641 |
+
"epoch": 49.23076923076923,
|
| 2642 |
+
"grad_norm": 0.03965931013226509,
|
| 2643 |
+
"learning_rate": 3.0769230769230774e-07,
|
| 2644 |
+
"loss": 0.0948,
|
| 2645 |
+
"step": 3200
|
| 2646 |
+
},
|
| 2647 |
+
{
|
| 2648 |
+
"epoch": 49.38461538461539,
|
| 2649 |
+
"grad_norm": 3.6046483516693115,
|
| 2650 |
+
"learning_rate": 2.4615384615384616e-07,
|
| 2651 |
+
"loss": 0.0302,
|
| 2652 |
+
"step": 3210
|
| 2653 |
+
},
|
| 2654 |
+
{
|
| 2655 |
+
"epoch": 49.53846153846154,
|
| 2656 |
+
"grad_norm": 0.03949074074625969,
|
| 2657 |
+
"learning_rate": 1.8461538461538464e-07,
|
| 2658 |
+
"loss": 0.0094,
|
| 2659 |
+
"step": 3220
|
| 2660 |
+
},
|
| 2661 |
+
{
|
| 2662 |
+
"epoch": 49.69230769230769,
|
| 2663 |
+
"grad_norm": 0.03980456292629242,
|
| 2664 |
+
"learning_rate": 1.2307692307692308e-07,
|
| 2665 |
+
"loss": 0.0388,
|
| 2666 |
+
"step": 3230
|
| 2667 |
+
},
|
| 2668 |
+
{
|
| 2669 |
+
"epoch": 49.84615384615385,
|
| 2670 |
+
"grad_norm": 8.264440536499023,
|
| 2671 |
+
"learning_rate": 6.153846153846154e-08,
|
| 2672 |
+
"loss": 0.0143,
|
| 2673 |
+
"step": 3240
|
| 2674 |
+
},
|
| 2675 |
+
{
|
| 2676 |
+
"epoch": 50.0,
|
| 2677 |
+
"grad_norm": 0.043733663856983185,
|
| 2678 |
+
"learning_rate": 0.0,
|
| 2679 |
+
"loss": 0.013,
|
| 2680 |
+
"step": 3250
|
| 2681 |
+
},
|
| 2682 |
+
{
|
| 2683 |
+
"epoch": 50.0,
|
| 2684 |
+
"eval_accuracy": 1.0,
|
| 2685 |
+
"eval_loss": 0.007947824895381927,
|
| 2686 |
+
"eval_runtime": 0.6285,
|
| 2687 |
+
"eval_samples_per_second": 211.608,
|
| 2688 |
+
"eval_steps_per_second": 14.319,
|
| 2689 |
+
"step": 3250
|
| 2690 |
+
},
|
| 2691 |
+
{
|
| 2692 |
+
"epoch": 50.0,
|
| 2693 |
+
"step": 3250,
|
| 2694 |
+
"total_flos": 3.6243328994998477e+18,
|
| 2695 |
+
"train_loss": 0.03743308119131968,
|
| 2696 |
+
"train_runtime": 550.1735,
|
| 2697 |
+
"train_samples_per_second": 93.97,
|
| 2698 |
+
"train_steps_per_second": 5.907
|
| 2699 |
}
|
| 2700 |
],
|
| 2701 |
"logging_steps": 10,
|
| 2702 |
+
"max_steps": 3250,
|
| 2703 |
"num_input_tokens_seen": 0,
|
| 2704 |
+
"num_train_epochs": 50,
|
| 2705 |
"save_steps": 500,
|
| 2706 |
"stateful_callbacks": {
|
| 2707 |
"TrainerControl": {
|
|
|
|
| 2715 |
"attributes": {}
|
| 2716 |
}
|
| 2717 |
},
|
| 2718 |
+
"total_flos": 3.6243328994998477e+18,
|
| 2719 |
"train_batch_size": 8,
|
| 2720 |
"trial_name": null,
|
| 2721 |
"trial_params": null
|