End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2590 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: open-thoughts/OpenThinker2-7B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: OpenR1-Math-Raw-all-correct-5k_OT7B
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# OpenR1-Math-Raw-all-correct-5k_OT7B
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [open-thoughts/OpenThinker2-7B](https://huggingface.co/open-thoughts/OpenThinker2-7B) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: open-thoughts/OpenThinker2-7B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: OpenR1-Math-Raw-all-correct-5k_OT7B
|
|
|
|
| 16 |
|
| 17 |
# OpenR1-Math-Raw-all-correct-5k_OT7B
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [open-thoughts/OpenThinker2-7B](https://huggingface.co/open-thoughts/OpenThinker2-7B) on the neginr/OpenR1-Math-Raw-all-correct-5k dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.955414012738854,
|
| 3 |
+
"total_flos": 3.947982283988664e+17,
|
| 4 |
+
"train_loss": 0.25456836733680505,
|
| 5 |
+
"train_runtime": 5658.0504,
|
| 6 |
+
"train_samples_per_second": 6.186,
|
| 7 |
+
"train_steps_per_second": 0.064
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.955414012738854,
|
| 3 |
+
"total_flos": 3.947982283988664e+17,
|
| 4 |
+
"train_loss": 0.25456836733680505,
|
| 5 |
+
"train_runtime": 5658.0504,
|
| 6 |
+
"train_samples_per_second": 6.186,
|
| 7 |
+
"train_steps_per_second": 0.064
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2590 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.955414012738854,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 364,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01910828025477707,
|
| 13 |
+
"grad_norm": 0.3258431400870058,
|
| 14 |
+
"learning_rate": 5.405405405405406e-09,
|
| 15 |
+
"loss": 0.2596,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03821656050955414,
|
| 20 |
+
"grad_norm": 0.2865253531051695,
|
| 21 |
+
"learning_rate": 1.0810810810810811e-08,
|
| 22 |
+
"loss": 0.2512,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05732484076433121,
|
| 27 |
+
"grad_norm": 0.30892863941613385,
|
| 28 |
+
"learning_rate": 1.6216216216216218e-08,
|
| 29 |
+
"loss": 0.2584,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07643312101910828,
|
| 34 |
+
"grad_norm": 0.3145090623295482,
|
| 35 |
+
"learning_rate": 2.1621621621621623e-08,
|
| 36 |
+
"loss": 0.2562,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09554140127388536,
|
| 41 |
+
"grad_norm": 0.3250498078418424,
|
| 42 |
+
"learning_rate": 2.7027027027027028e-08,
|
| 43 |
+
"loss": 0.2557,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.11464968152866242,
|
| 48 |
+
"grad_norm": 0.3048291606019023,
|
| 49 |
+
"learning_rate": 3.2432432432432436e-08,
|
| 50 |
+
"loss": 0.2429,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.1337579617834395,
|
| 55 |
+
"grad_norm": 0.2813946488902139,
|
| 56 |
+
"learning_rate": 3.783783783783784e-08,
|
| 57 |
+
"loss": 0.244,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.15286624203821655,
|
| 62 |
+
"grad_norm": 0.2951556246869404,
|
| 63 |
+
"learning_rate": 4.3243243243243246e-08,
|
| 64 |
+
"loss": 0.2475,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.17197452229299362,
|
| 69 |
+
"grad_norm": 0.31375967109328035,
|
| 70 |
+
"learning_rate": 4.864864864864865e-08,
|
| 71 |
+
"loss": 0.2548,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1910828025477707,
|
| 76 |
+
"grad_norm": 0.320646972442393,
|
| 77 |
+
"learning_rate": 5.4054054054054056e-08,
|
| 78 |
+
"loss": 0.2593,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.21019108280254778,
|
| 83 |
+
"grad_norm": 0.3101235439496355,
|
| 84 |
+
"learning_rate": 5.945945945945946e-08,
|
| 85 |
+
"loss": 0.2606,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.22929936305732485,
|
| 90 |
+
"grad_norm": 0.29711763418174875,
|
| 91 |
+
"learning_rate": 6.486486486486487e-08,
|
| 92 |
+
"loss": 0.2517,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2484076433121019,
|
| 97 |
+
"grad_norm": 0.3071305484501633,
|
| 98 |
+
"learning_rate": 7.027027027027027e-08,
|
| 99 |
+
"loss": 0.2494,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.267515923566879,
|
| 104 |
+
"grad_norm": 0.2845310181092547,
|
| 105 |
+
"learning_rate": 7.567567567567568e-08,
|
| 106 |
+
"loss": 0.2491,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.28662420382165604,
|
| 111 |
+
"grad_norm": 0.3203987747997849,
|
| 112 |
+
"learning_rate": 8.108108108108108e-08,
|
| 113 |
+
"loss": 0.2543,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.3057324840764331,
|
| 118 |
+
"grad_norm": 0.28456107510027123,
|
| 119 |
+
"learning_rate": 8.648648648648649e-08,
|
| 120 |
+
"loss": 0.258,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3248407643312102,
|
| 125 |
+
"grad_norm": 0.27941243424990353,
|
| 126 |
+
"learning_rate": 9.189189189189189e-08,
|
| 127 |
+
"loss": 0.2574,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.34394904458598724,
|
| 132 |
+
"grad_norm": 0.2930471474625414,
|
| 133 |
+
"learning_rate": 9.72972972972973e-08,
|
| 134 |
+
"loss": 0.2743,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3630573248407643,
|
| 139 |
+
"grad_norm": 0.30494204356688037,
|
| 140 |
+
"learning_rate": 1.027027027027027e-07,
|
| 141 |
+
"loss": 0.269,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3821656050955414,
|
| 146 |
+
"grad_norm": 0.30231424553130715,
|
| 147 |
+
"learning_rate": 1.0810810810810811e-07,
|
| 148 |
+
"loss": 0.2557,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.4012738853503185,
|
| 153 |
+
"grad_norm": 0.3015266138690226,
|
| 154 |
+
"learning_rate": 1.135135135135135e-07,
|
| 155 |
+
"loss": 0.2523,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.42038216560509556,
|
| 160 |
+
"grad_norm": 0.3043409829723809,
|
| 161 |
+
"learning_rate": 1.1891891891891891e-07,
|
| 162 |
+
"loss": 0.2547,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.4394904458598726,
|
| 167 |
+
"grad_norm": 0.29444734831015446,
|
| 168 |
+
"learning_rate": 1.2432432432432432e-07,
|
| 169 |
+
"loss": 0.2517,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4585987261146497,
|
| 174 |
+
"grad_norm": 0.297937003697346,
|
| 175 |
+
"learning_rate": 1.2972972972972974e-07,
|
| 176 |
+
"loss": 0.2434,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.47770700636942676,
|
| 181 |
+
"grad_norm": 0.3027152740282303,
|
| 182 |
+
"learning_rate": 1.3513513513513512e-07,
|
| 183 |
+
"loss": 0.2538,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.4968152866242038,
|
| 188 |
+
"grad_norm": 0.3085528948055293,
|
| 189 |
+
"learning_rate": 1.4054054054054055e-07,
|
| 190 |
+
"loss": 0.2455,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.5159235668789809,
|
| 195 |
+
"grad_norm": 0.30288755432020104,
|
| 196 |
+
"learning_rate": 1.4594594594594595e-07,
|
| 197 |
+
"loss": 0.2642,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.535031847133758,
|
| 202 |
+
"grad_norm": 0.2888257868462652,
|
| 203 |
+
"learning_rate": 1.5135135135135135e-07,
|
| 204 |
+
"loss": 0.2535,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.554140127388535,
|
| 209 |
+
"grad_norm": 0.31240869601074167,
|
| 210 |
+
"learning_rate": 1.5675675675675675e-07,
|
| 211 |
+
"loss": 0.2574,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5732484076433121,
|
| 216 |
+
"grad_norm": 0.28779468390754065,
|
| 217 |
+
"learning_rate": 1.6216216216216215e-07,
|
| 218 |
+
"loss": 0.2562,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5923566878980892,
|
| 223 |
+
"grad_norm": 0.3026300082985163,
|
| 224 |
+
"learning_rate": 1.6756756756756755e-07,
|
| 225 |
+
"loss": 0.2489,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.6114649681528662,
|
| 230 |
+
"grad_norm": 0.29410215399776,
|
| 231 |
+
"learning_rate": 1.7297297297297298e-07,
|
| 232 |
+
"loss": 0.2513,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.6305732484076433,
|
| 237 |
+
"grad_norm": 0.2967099409221429,
|
| 238 |
+
"learning_rate": 1.7837837837837836e-07,
|
| 239 |
+
"loss": 0.2605,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6496815286624203,
|
| 244 |
+
"grad_norm": 0.28890784109752826,
|
| 245 |
+
"learning_rate": 1.8378378378378379e-07,
|
| 246 |
+
"loss": 0.2488,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6687898089171974,
|
| 251 |
+
"grad_norm": 0.3228375228965286,
|
| 252 |
+
"learning_rate": 1.891891891891892e-07,
|
| 253 |
+
"loss": 0.2629,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6878980891719745,
|
| 258 |
+
"grad_norm": 0.30208391730174905,
|
| 259 |
+
"learning_rate": 1.945945945945946e-07,
|
| 260 |
+
"loss": 0.2662,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.7070063694267515,
|
| 265 |
+
"grad_norm": 0.2895031799209941,
|
| 266 |
+
"learning_rate": 2e-07,
|
| 267 |
+
"loss": 0.2536,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.7261146496815286,
|
| 272 |
+
"grad_norm": 0.3036410296650111,
|
| 273 |
+
"learning_rate": 1.999953850085163e-07,
|
| 274 |
+
"loss": 0.2654,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.7452229299363057,
|
| 279 |
+
"grad_norm": 0.30128310490546484,
|
| 280 |
+
"learning_rate": 1.999815404600282e-07,
|
| 281 |
+
"loss": 0.2471,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7643312101910829,
|
| 286 |
+
"grad_norm": 0.3079903978462585,
|
| 287 |
+
"learning_rate": 1.999584676323851e-07,
|
| 288 |
+
"loss": 0.256,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7834394904458599,
|
| 293 |
+
"grad_norm": 0.2917481324310926,
|
| 294 |
+
"learning_rate": 1.9992616865520512e-07,
|
| 295 |
+
"loss": 0.2499,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.802547770700637,
|
| 300 |
+
"grad_norm": 0.31471272570274456,
|
| 301 |
+
"learning_rate": 1.998846465096783e-07,
|
| 302 |
+
"loss": 0.2641,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.821656050955414,
|
| 307 |
+
"grad_norm": 0.3152199748266918,
|
| 308 |
+
"learning_rate": 1.9983390502829166e-07,
|
| 309 |
+
"loss": 0.2593,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.8407643312101911,
|
| 314 |
+
"grad_norm": 0.29879027476626574,
|
| 315 |
+
"learning_rate": 1.9977394889447523e-07,
|
| 316 |
+
"loss": 0.2686,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8598726114649682,
|
| 321 |
+
"grad_norm": 0.3172098054667475,
|
| 322 |
+
"learning_rate": 1.9970478364216996e-07,
|
| 323 |
+
"loss": 0.2408,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8789808917197452,
|
| 328 |
+
"grad_norm": 0.29042679625683193,
|
| 329 |
+
"learning_rate": 1.996264156553169e-07,
|
| 330 |
+
"loss": 0.2664,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8980891719745223,
|
| 335 |
+
"grad_norm": 0.30019512901289386,
|
| 336 |
+
"learning_rate": 1.9953885216726785e-07,
|
| 337 |
+
"loss": 0.2508,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.9171974522292994,
|
| 342 |
+
"grad_norm": 0.28586435210675887,
|
| 343 |
+
"learning_rate": 1.9944210126011788e-07,
|
| 344 |
+
"loss": 0.2522,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.9363057324840764,
|
| 349 |
+
"grad_norm": 0.30773911686783834,
|
| 350 |
+
"learning_rate": 1.9933617186395914e-07,
|
| 351 |
+
"loss": 0.2502,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.9554140127388535,
|
| 356 |
+
"grad_norm": 0.31371477716041124,
|
| 357 |
+
"learning_rate": 1.9922107375605698e-07,
|
| 358 |
+
"loss": 0.2619,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9745222929936306,
|
| 363 |
+
"grad_norm": 0.2950309178219575,
|
| 364 |
+
"learning_rate": 1.990968175599471e-07,
|
| 365 |
+
"loss": 0.2613,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9936305732484076,
|
| 370 |
+
"grad_norm": 0.3320695491566013,
|
| 371 |
+
"learning_rate": 1.9896341474445524e-07,
|
| 372 |
+
"loss": 0.2541,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0127388535031847,
|
| 377 |
+
"grad_norm": 0.2902283899600383,
|
| 378 |
+
"learning_rate": 1.9882087762263852e-07,
|
| 379 |
+
"loss": 0.2485,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.0318471337579618,
|
| 384 |
+
"grad_norm": 0.30704788847785364,
|
| 385 |
+
"learning_rate": 1.9866921935064905e-07,
|
| 386 |
+
"loss": 0.2579,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.0509554140127388,
|
| 391 |
+
"grad_norm": 0.2786908392929887,
|
| 392 |
+
"learning_rate": 1.9850845392651947e-07,
|
| 393 |
+
"loss": 0.2638,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.070063694267516,
|
| 398 |
+
"grad_norm": 0.317768378195936,
|
| 399 |
+
"learning_rate": 1.983385961888711e-07,
|
| 400 |
+
"loss": 0.2483,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.089171974522293,
|
| 405 |
+
"grad_norm": 0.2999290392989265,
|
| 406 |
+
"learning_rate": 1.981596618155441e-07,
|
| 407 |
+
"loss": 0.2604,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.10828025477707,
|
| 412 |
+
"grad_norm": 0.31058999539375254,
|
| 413 |
+
"learning_rate": 1.9797166732215075e-07,
|
| 414 |
+
"loss": 0.2619,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.127388535031847,
|
| 419 |
+
"grad_norm": 0.3064782254040447,
|
| 420 |
+
"learning_rate": 1.977746300605507e-07,
|
| 421 |
+
"loss": 0.2607,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.1464968152866242,
|
| 426 |
+
"grad_norm": 0.3002412022740287,
|
| 427 |
+
"learning_rate": 1.9756856821724967e-07,
|
| 428 |
+
"loss": 0.2617,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.1656050955414012,
|
| 433 |
+
"grad_norm": 0.29772159876550963,
|
| 434 |
+
"learning_rate": 1.9735350081172067e-07,
|
| 435 |
+
"loss": 0.2481,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.1847133757961783,
|
| 440 |
+
"grad_norm": 0.31356783559096124,
|
| 441 |
+
"learning_rate": 1.9712944769464862e-07,
|
| 442 |
+
"loss": 0.2587,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.2038216560509554,
|
| 447 |
+
"grad_norm": 0.2888581149656851,
|
| 448 |
+
"learning_rate": 1.9689642954609806e-07,
|
| 449 |
+
"loss": 0.2538,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.2229299363057324,
|
| 454 |
+
"grad_norm": 0.30385671289298677,
|
| 455 |
+
"learning_rate": 1.966544678736044e-07,
|
| 456 |
+
"loss": 0.2527,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.2420382165605095,
|
| 461 |
+
"grad_norm": 0.2973647308882802,
|
| 462 |
+
"learning_rate": 1.9640358501018882e-07,
|
| 463 |
+
"loss": 0.2479,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.2611464968152866,
|
| 468 |
+
"grad_norm": 0.3095848016456744,
|
| 469 |
+
"learning_rate": 1.961438041122969e-07,
|
| 470 |
+
"loss": 0.2482,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.2802547770700636,
|
| 475 |
+
"grad_norm": 0.3515887000957586,
|
| 476 |
+
"learning_rate": 1.9587514915766122e-07,
|
| 477 |
+
"loss": 0.253,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.2993630573248407,
|
| 482 |
+
"grad_norm": 0.2931897201883793,
|
| 483 |
+
"learning_rate": 1.9559764494308834e-07,
|
| 484 |
+
"loss": 0.2457,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.3184713375796178,
|
| 489 |
+
"grad_norm": 0.31283649147238096,
|
| 490 |
+
"learning_rate": 1.9531131708217004e-07,
|
| 491 |
+
"loss": 0.2554,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.3375796178343948,
|
| 496 |
+
"grad_norm": 0.30295461967215936,
|
| 497 |
+
"learning_rate": 1.9501619200291905e-07,
|
| 498 |
+
"loss": 0.2561,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.356687898089172,
|
| 503 |
+
"grad_norm": 0.2910098967517895,
|
| 504 |
+
"learning_rate": 1.9471229694533e-07,
|
| 505 |
+
"loss": 0.264,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.3757961783439492,
|
| 510 |
+
"grad_norm": 0.29074377848027877,
|
| 511 |
+
"learning_rate": 1.9439965995886488e-07,
|
| 512 |
+
"loss": 0.2579,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.394904458598726,
|
| 517 |
+
"grad_norm": 0.2913116920333876,
|
| 518 |
+
"learning_rate": 1.9407830989986428e-07,
|
| 519 |
+
"loss": 0.2515,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.4140127388535033,
|
| 524 |
+
"grad_norm": 0.28025336367518355,
|
| 525 |
+
"learning_rate": 1.9374827642888395e-07,
|
| 526 |
+
"loss": 0.2592,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.4331210191082802,
|
| 531 |
+
"grad_norm": 0.308381777961692,
|
| 532 |
+
"learning_rate": 1.9340959000795706e-07,
|
| 533 |
+
"loss": 0.2559,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.4522292993630574,
|
| 538 |
+
"grad_norm": 0.29320496586078204,
|
| 539 |
+
"learning_rate": 1.9306228189778253e-07,
|
| 540 |
+
"loss": 0.2658,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.4713375796178343,
|
| 545 |
+
"grad_norm": 0.32221369887070983,
|
| 546 |
+
"learning_rate": 1.927063841548398e-07,
|
| 547 |
+
"loss": 0.2646,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.4904458598726116,
|
| 552 |
+
"grad_norm": 0.299728133006886,
|
| 553 |
+
"learning_rate": 1.923419296284299e-07,
|
| 554 |
+
"loss": 0.2581,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.5095541401273884,
|
| 559 |
+
"grad_norm": 0.3209108114734401,
|
| 560 |
+
"learning_rate": 1.919689519576436e-07,
|
| 561 |
+
"loss": 0.2567,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.5286624203821657,
|
| 566 |
+
"grad_norm": 0.2804129315244473,
|
| 567 |
+
"learning_rate": 1.9158748556825634e-07,
|
| 568 |
+
"loss": 0.253,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.5477707006369426,
|
| 573 |
+
"grad_norm": 0.3045816243631572,
|
| 574 |
+
"learning_rate": 1.911975656695509e-07,
|
| 575 |
+
"loss": 0.2584,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.5668789808917198,
|
| 580 |
+
"grad_norm": 0.3126990342175922,
|
| 581 |
+
"learning_rate": 1.907992282510675e-07,
|
| 582 |
+
"loss": 0.2605,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.5859872611464967,
|
| 587 |
+
"grad_norm": 0.30614700686477747,
|
| 588 |
+
"learning_rate": 1.90392510079282e-07,
|
| 589 |
+
"loss": 0.2548,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.605095541401274,
|
| 594 |
+
"grad_norm": 0.2971554730236426,
|
| 595 |
+
"learning_rate": 1.8997744869421245e-07,
|
| 596 |
+
"loss": 0.2354,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.6242038216560508,
|
| 601 |
+
"grad_norm": 0.2988208824377952,
|
| 602 |
+
"learning_rate": 1.8955408240595392e-07,
|
| 603 |
+
"loss": 0.2441,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.643312101910828,
|
| 608 |
+
"grad_norm": 0.2947992742232842,
|
| 609 |
+
"learning_rate": 1.8912245029114278e-07,
|
| 610 |
+
"loss": 0.2503,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.662420382165605,
|
| 615 |
+
"grad_norm": 0.3147211833492056,
|
| 616 |
+
"learning_rate": 1.8868259218934966e-07,
|
| 617 |
+
"loss": 0.2529,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.6815286624203822,
|
| 622 |
+
"grad_norm": 0.3150826072001003,
|
| 623 |
+
"learning_rate": 1.882345486994024e-07,
|
| 624 |
+
"loss": 0.2492,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.700636942675159,
|
| 629 |
+
"grad_norm": 0.29922927789546194,
|
| 630 |
+
"learning_rate": 1.877783611756389e-07,
|
| 631 |
+
"loss": 0.2657,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.7197452229299364,
|
| 636 |
+
"grad_norm": 0.2841068444708262,
|
| 637 |
+
"learning_rate": 1.8731407172408987e-07,
|
| 638 |
+
"loss": 0.2503,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.7388535031847132,
|
| 643 |
+
"grad_norm": 0.3000862442323796,
|
| 644 |
+
"learning_rate": 1.8684172319859257e-07,
|
| 645 |
+
"loss": 0.2553,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.7579617834394905,
|
| 650 |
+
"grad_norm": 0.3058443181160777,
|
| 651 |
+
"learning_rate": 1.863613591968355e-07,
|
| 652 |
+
"loss": 0.2572,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.7770700636942676,
|
| 657 |
+
"grad_norm": 0.31867483871696056,
|
| 658 |
+
"learning_rate": 1.8587302405633417e-07,
|
| 659 |
+
"loss": 0.2527,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.7961783439490446,
|
| 664 |
+
"grad_norm": 0.3025042298897706,
|
| 665 |
+
"learning_rate": 1.8537676285033885e-07,
|
| 666 |
+
"loss": 0.2595,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.8152866242038217,
|
| 671 |
+
"grad_norm": 0.2923220305017809,
|
| 672 |
+
"learning_rate": 1.848726213836744e-07,
|
| 673 |
+
"loss": 0.2577,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.8343949044585988,
|
| 678 |
+
"grad_norm": 0.3003765605795671,
|
| 679 |
+
"learning_rate": 1.8436064618851224e-07,
|
| 680 |
+
"loss": 0.2521,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.8535031847133758,
|
| 685 |
+
"grad_norm": 0.317610406351691,
|
| 686 |
+
"learning_rate": 1.8384088452007576e-07,
|
| 687 |
+
"loss": 0.2502,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.872611464968153,
|
| 692 |
+
"grad_norm": 0.2887579767892906,
|
| 693 |
+
"learning_rate": 1.8331338435227837e-07,
|
| 694 |
+
"loss": 0.2586,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.89171974522293,
|
| 699 |
+
"grad_norm": 0.3197071782148506,
|
| 700 |
+
"learning_rate": 1.8277819437329574e-07,
|
| 701 |
+
"loss": 0.267,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.910828025477707,
|
| 706 |
+
"grad_norm": 0.2849069609408447,
|
| 707 |
+
"learning_rate": 1.8223536398107174e-07,
|
| 708 |
+
"loss": 0.2485,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.929936305732484,
|
| 713 |
+
"grad_norm": 0.2887703170078543,
|
| 714 |
+
"learning_rate": 1.8168494327875916e-07,
|
| 715 |
+
"loss": 0.2499,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.9490445859872612,
|
| 720 |
+
"grad_norm": 0.3161969478865281,
|
| 721 |
+
"learning_rate": 1.8112698307009504e-07,
|
| 722 |
+
"loss": 0.2505,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.9681528662420382,
|
| 727 |
+
"grad_norm": 0.287901366472073,
|
| 728 |
+
"learning_rate": 1.8056153485471165e-07,
|
| 729 |
+
"loss": 0.2668,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.9872611464968153,
|
| 734 |
+
"grad_norm": 0.3012560634417445,
|
| 735 |
+
"learning_rate": 1.7998865082338287e-07,
|
| 736 |
+
"loss": 0.2565,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.0063694267515926,
|
| 741 |
+
"grad_norm": 0.29276615150168056,
|
| 742 |
+
"learning_rate": 1.7940838385320732e-07,
|
| 743 |
+
"loss": 0.2533,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.0254777070063694,
|
| 748 |
+
"grad_norm": 0.30402861564543393,
|
| 749 |
+
"learning_rate": 1.788207875027274e-07,
|
| 750 |
+
"loss": 0.2525,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.0445859872611467,
|
| 755 |
+
"grad_norm": 0.2967173967659541,
|
| 756 |
+
"learning_rate": 1.7822591600698629e-07,
|
| 757 |
+
"loss": 0.2532,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.0636942675159236,
|
| 762 |
+
"grad_norm": 0.30366484369127167,
|
| 763 |
+
"learning_rate": 1.7762382427252165e-07,
|
| 764 |
+
"loss": 0.2573,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.082802547770701,
|
| 769 |
+
"grad_norm": 0.2878188158407926,
|
| 770 |
+
"learning_rate": 1.7701456787229803e-07,
|
| 771 |
+
"loss": 0.2602,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.1019108280254777,
|
| 776 |
+
"grad_norm": 0.2743059722128589,
|
| 777 |
+
"learning_rate": 1.7639820304057742e-07,
|
| 778 |
+
"loss": 0.2554,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.121019108280255,
|
| 783 |
+
"grad_norm": 0.29808343098853324,
|
| 784 |
+
"learning_rate": 1.7577478666772882e-07,
|
| 785 |
+
"loss": 0.253,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.140127388535032,
|
| 790 |
+
"grad_norm": 0.2875587315242635,
|
| 791 |
+
"learning_rate": 1.7514437629497717e-07,
|
| 792 |
+
"loss": 0.2488,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.159235668789809,
|
| 797 |
+
"grad_norm": 0.30351361854827635,
|
| 798 |
+
"learning_rate": 1.7450703010909262e-07,
|
| 799 |
+
"loss": 0.2562,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.178343949044586,
|
| 804 |
+
"grad_norm": 0.30394129590401453,
|
| 805 |
+
"learning_rate": 1.738628069370195e-07,
|
| 806 |
+
"loss": 0.2607,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.1974522292993632,
|
| 811 |
+
"grad_norm": 0.32285950783403167,
|
| 812 |
+
"learning_rate": 1.7321176624044687e-07,
|
| 813 |
+
"loss": 0.2503,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.21656050955414,
|
| 818 |
+
"grad_norm": 0.3120864765033612,
|
| 819 |
+
"learning_rate": 1.7255396811032013e-07,
|
| 820 |
+
"loss": 0.2509,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.2356687898089174,
|
| 825 |
+
"grad_norm": 0.3236920273170069,
|
| 826 |
+
"learning_rate": 1.718894732612947e-07,
|
| 827 |
+
"loss": 0.2502,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.254777070063694,
|
| 832 |
+
"grad_norm": 0.29570882874175936,
|
| 833 |
+
"learning_rate": 1.7121834302613186e-07,
|
| 834 |
+
"loss": 0.2639,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.2738853503184715,
|
| 839 |
+
"grad_norm": 0.31678397348330944,
|
| 840 |
+
"learning_rate": 1.7054063935003812e-07,
|
| 841 |
+
"loss": 0.2496,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.2929936305732483,
|
| 846 |
+
"grad_norm": 0.2988770843936996,
|
| 847 |
+
"learning_rate": 1.6985642478494727e-07,
|
| 848 |
+
"loss": 0.2507,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.3121019108280256,
|
| 853 |
+
"grad_norm": 0.3109434916642981,
|
| 854 |
+
"learning_rate": 1.6916576248374716e-07,
|
| 855 |
+
"loss": 0.2616,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.3312101910828025,
|
| 860 |
+
"grad_norm": 0.3261007418875524,
|
| 861 |
+
"learning_rate": 1.684687161944506e-07,
|
| 862 |
+
"loss": 0.2554,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.3503184713375798,
|
| 867 |
+
"grad_norm": 0.31440782489796776,
|
| 868 |
+
"learning_rate": 1.6776535025431129e-07,
|
| 869 |
+
"loss": 0.2559,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.3694267515923566,
|
| 874 |
+
"grad_norm": 0.30410776957586444,
|
| 875 |
+
"learning_rate": 1.6705572958388573e-07,
|
| 876 |
+
"loss": 0.2603,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.388535031847134,
|
| 881 |
+
"grad_norm": 0.30666376843334026,
|
| 882 |
+
"learning_rate": 1.6633991968104092e-07,
|
| 883 |
+
"loss": 0.2439,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.4076433121019107,
|
| 888 |
+
"grad_norm": 0.3133700400691411,
|
| 889 |
+
"learning_rate": 1.6561798661490902e-07,
|
| 890 |
+
"loss": 0.2514,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.426751592356688,
|
| 895 |
+
"grad_norm": 0.30312163468190945,
|
| 896 |
+
"learning_rate": 1.6488999701978902e-07,
|
| 897 |
+
"loss": 0.2522,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.445859872611465,
|
| 902 |
+
"grad_norm": 0.2980517023756582,
|
| 903 |
+
"learning_rate": 1.6415601808899658e-07,
|
| 904 |
+
"loss": 0.2634,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.464968152866242,
|
| 909 |
+
"grad_norm": 0.29620642402954184,
|
| 910 |
+
"learning_rate": 1.63416117568662e-07,
|
| 911 |
+
"loss": 0.2523,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.484076433121019,
|
| 916 |
+
"grad_norm": 0.30847707783645134,
|
| 917 |
+
"learning_rate": 1.6267036375147723e-07,
|
| 918 |
+
"loss": 0.2519,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.5031847133757963,
|
| 923 |
+
"grad_norm": 0.3150233916320143,
|
| 924 |
+
"learning_rate": 1.6191882547039266e-07,
|
| 925 |
+
"loss": 0.257,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.522292993630573,
|
| 930 |
+
"grad_norm": 0.27358778066993983,
|
| 931 |
+
"learning_rate": 1.6116157209226352e-07,
|
| 932 |
+
"loss": 0.2647,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.5414012738853504,
|
| 937 |
+
"grad_norm": 0.3019435375253268,
|
| 938 |
+
"learning_rate": 1.6039867351144777e-07,
|
| 939 |
+
"loss": 0.2647,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.5605095541401273,
|
| 944 |
+
"grad_norm": 0.30895517117616256,
|
| 945 |
+
"learning_rate": 1.5963020014335436e-07,
|
| 946 |
+
"loss": 0.2559,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.5796178343949046,
|
| 951 |
+
"grad_norm": 0.31282726024916446,
|
| 952 |
+
"learning_rate": 1.5885622291794428e-07,
|
| 953 |
+
"loss": 0.2473,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.5987261146496814,
|
| 958 |
+
"grad_norm": 0.29386800602210383,
|
| 959 |
+
"learning_rate": 1.580768132731837e-07,
|
| 960 |
+
"loss": 0.2564,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.6178343949044587,
|
| 965 |
+
"grad_norm": 0.2994659441023476,
|
| 966 |
+
"learning_rate": 1.5729204314845e-07,
|
| 967 |
+
"loss": 0.2593,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.6369426751592355,
|
| 972 |
+
"grad_norm": 0.306461349101831,
|
| 973 |
+
"learning_rate": 1.56501984977892e-07,
|
| 974 |
+
"loss": 0.2618,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.656050955414013,
|
| 979 |
+
"grad_norm": 0.28736145386635525,
|
| 980 |
+
"learning_rate": 1.5570671168374436e-07,
|
| 981 |
+
"loss": 0.2702,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.6751592356687897,
|
| 986 |
+
"grad_norm": 0.2862972968768037,
|
| 987 |
+
"learning_rate": 1.5490629666959666e-07,
|
| 988 |
+
"loss": 0.2604,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.694267515923567,
|
| 993 |
+
"grad_norm": 0.3188610660653441,
|
| 994 |
+
"learning_rate": 1.5410081381361829e-07,
|
| 995 |
+
"loss": 0.2671,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.713375796178344,
|
| 1000 |
+
"grad_norm": 0.3094794846413303,
|
| 1001 |
+
"learning_rate": 1.5329033746173973e-07,
|
| 1002 |
+
"loss": 0.2535,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.732484076433121,
|
| 1007 |
+
"grad_norm": 0.29891576281966575,
|
| 1008 |
+
"learning_rate": 1.5247494242079021e-07,
|
| 1009 |
+
"loss": 0.252,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.7515923566878984,
|
| 1014 |
+
"grad_norm": 0.292028133356272,
|
| 1015 |
+
"learning_rate": 1.5165470395159313e-07,
|
| 1016 |
+
"loss": 0.2517,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.770700636942675,
|
| 1021 |
+
"grad_norm": 0.3015119959336549,
|
| 1022 |
+
"learning_rate": 1.5082969776201945e-07,
|
| 1023 |
+
"loss": 0.2485,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.789808917197452,
|
| 1028 |
+
"grad_norm": 0.30946055407183126,
|
| 1029 |
+
"learning_rate": 1.5e-07,
|
| 1030 |
+
"loss": 0.2549,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.8089171974522293,
|
| 1035 |
+
"grad_norm": 0.3011288987792503,
|
| 1036 |
+
"learning_rate": 1.4916568724649686e-07,
|
| 1037 |
+
"loss": 0.2526,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.8280254777070066,
|
| 1042 |
+
"grad_norm": 0.27935366900331016,
|
| 1043 |
+
"learning_rate": 1.4832683650843506e-07,
|
| 1044 |
+
"loss": 0.2569,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.8471337579617835,
|
| 1049 |
+
"grad_norm": 0.3053623378204874,
|
| 1050 |
+
"learning_rate": 1.4748352521159491e-07,
|
| 1051 |
+
"loss": 0.2543,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.8662420382165603,
|
| 1056 |
+
"grad_norm": 0.30986894267799603,
|
| 1057 |
+
"learning_rate": 1.4663583119346538e-07,
|
| 1058 |
+
"loss": 0.2414,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.8853503184713376,
|
| 1063 |
+
"grad_norm": 0.3029701849877518,
|
| 1064 |
+
"learning_rate": 1.4578383269606002e-07,
|
| 1065 |
+
"loss": 0.2645,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.904458598726115,
|
| 1070 |
+
"grad_norm": 0.3007935845890958,
|
| 1071 |
+
"learning_rate": 1.4492760835869502e-07,
|
| 1072 |
+
"loss": 0.2524,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.9235668789808917,
|
| 1077 |
+
"grad_norm": 0.3027359482846415,
|
| 1078 |
+
"learning_rate": 1.4406723721073087e-07,
|
| 1079 |
+
"loss": 0.2399,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.9426751592356686,
|
| 1084 |
+
"grad_norm": 0.286593283803398,
|
| 1085 |
+
"learning_rate": 1.4320279866427796e-07,
|
| 1086 |
+
"loss": 0.2491,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.961783439490446,
|
| 1091 |
+
"grad_norm": 0.2975044719847865,
|
| 1092 |
+
"learning_rate": 1.4233437250686693e-07,
|
| 1093 |
+
"loss": 0.2556,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.980891719745223,
|
| 1098 |
+
"grad_norm": 0.29788132002693174,
|
| 1099 |
+
"learning_rate": 1.4146203889408418e-07,
|
| 1100 |
+
"loss": 0.242,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 3.0,
|
| 1105 |
+
"grad_norm": 0.2947537039751411,
|
| 1106 |
+
"learning_rate": 1.4058587834217354e-07,
|
| 1107 |
+
"loss": 0.2584,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 3.0191082802547773,
|
| 1112 |
+
"grad_norm": 0.29945926345973045,
|
| 1113 |
+
"learning_rate": 1.397059717206048e-07,
|
| 1114 |
+
"loss": 0.2591,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 3.038216560509554,
|
| 1119 |
+
"grad_norm": 0.3078621269226234,
|
| 1120 |
+
"learning_rate": 1.3882240024460924e-07,
|
| 1121 |
+
"loss": 0.2587,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 3.0573248407643314,
|
| 1126 |
+
"grad_norm": 0.30621687808211867,
|
| 1127 |
+
"learning_rate": 1.3793524546768356e-07,
|
| 1128 |
+
"loss": 0.2603,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 3.0764331210191083,
|
| 1133 |
+
"grad_norm": 0.30958764885598344,
|
| 1134 |
+
"learning_rate": 1.370445892740626e-07,
|
| 1135 |
+
"loss": 0.2594,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 3.0955414012738856,
|
| 1140 |
+
"grad_norm": 0.3117099403903537,
|
| 1141 |
+
"learning_rate": 1.361505138711613e-07,
|
| 1142 |
+
"loss": 0.2538,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.1146496815286624,
|
| 1147 |
+
"grad_norm": 0.31084130617285904,
|
| 1148 |
+
"learning_rate": 1.3525310178198706e-07,
|
| 1149 |
+
"loss": 0.2658,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 3.1337579617834397,
|
| 1154 |
+
"grad_norm": 0.30690130397546955,
|
| 1155 |
+
"learning_rate": 1.343524358375229e-07,
|
| 1156 |
+
"loss": 0.2495,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 3.1528662420382165,
|
| 1161 |
+
"grad_norm": 0.290775795962111,
|
| 1162 |
+
"learning_rate": 1.3344859916908204e-07,
|
| 1163 |
+
"loss": 0.2574,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 3.171974522292994,
|
| 1168 |
+
"grad_norm": 0.31192756387768816,
|
| 1169 |
+
"learning_rate": 1.325416752006351e-07,
|
| 1170 |
+
"loss": 0.2548,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 3.1910828025477707,
|
| 1175 |
+
"grad_norm": 0.2970090514175917,
|
| 1176 |
+
"learning_rate": 1.3163174764110982e-07,
|
| 1177 |
+
"loss": 0.248,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 3.210191082802548,
|
| 1182 |
+
"grad_norm": 0.2904565168466135,
|
| 1183 |
+
"learning_rate": 1.3071890047666496e-07,
|
| 1184 |
+
"loss": 0.2469,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 3.229299363057325,
|
| 1189 |
+
"grad_norm": 0.30572059607999896,
|
| 1190 |
+
"learning_rate": 1.2980321796293835e-07,
|
| 1191 |
+
"loss": 0.2433,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 3.248407643312102,
|
| 1196 |
+
"grad_norm": 0.3080588666919199,
|
| 1197 |
+
"learning_rate": 1.288847846172701e-07,
|
| 1198 |
+
"loss": 0.2455,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 3.267515923566879,
|
| 1203 |
+
"grad_norm": 0.33087482625160974,
|
| 1204 |
+
"learning_rate": 1.2796368521090143e-07,
|
| 1205 |
+
"loss": 0.2563,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.286624203821656,
|
| 1210 |
+
"grad_norm": 0.29813533166624945,
|
| 1211 |
+
"learning_rate": 1.270400047611508e-07,
|
| 1212 |
+
"loss": 0.2486,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 3.305732484076433,
|
| 1217 |
+
"grad_norm": 0.3086312021030327,
|
| 1218 |
+
"learning_rate": 1.261138285235663e-07,
|
| 1219 |
+
"loss": 0.2458,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.3248407643312103,
|
| 1224 |
+
"grad_norm": 0.29331773923085075,
|
| 1225 |
+
"learning_rate": 1.2518524198405698e-07,
|
| 1226 |
+
"loss": 0.2709,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 3.343949044585987,
|
| 1231 |
+
"grad_norm": 0.3389129034608298,
|
| 1232 |
+
"learning_rate": 1.2425433085100222e-07,
|
| 1233 |
+
"loss": 0.249,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 3.3630573248407645,
|
| 1238 |
+
"grad_norm": 0.27583808400445303,
|
| 1239 |
+
"learning_rate": 1.2332118104734109e-07,
|
| 1240 |
+
"loss": 0.2593,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 3.3821656050955413,
|
| 1245 |
+
"grad_norm": 0.30026419416491845,
|
| 1246 |
+
"learning_rate": 1.223858787026415e-07,
|
| 1247 |
+
"loss": 0.2571,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 3.4012738853503186,
|
| 1252 |
+
"grad_norm": 0.2870756139104833,
|
| 1253 |
+
"learning_rate": 1.2144851014515054e-07,
|
| 1254 |
+
"loss": 0.2433,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 3.4203821656050954,
|
| 1259 |
+
"grad_norm": 0.2872438134366015,
|
| 1260 |
+
"learning_rate": 1.2050916189382645e-07,
|
| 1261 |
+
"loss": 0.2612,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 3.4394904458598727,
|
| 1266 |
+
"grad_norm": 0.3163865718075819,
|
| 1267 |
+
"learning_rate": 1.195679206503528e-07,
|
| 1268 |
+
"loss": 0.2549,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 3.4585987261146496,
|
| 1273 |
+
"grad_norm": 0.2978921734506274,
|
| 1274 |
+
"learning_rate": 1.1862487329113604e-07,
|
| 1275 |
+
"loss": 0.2622,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 3.477707006369427,
|
| 1280 |
+
"grad_norm": 0.2948938802787196,
|
| 1281 |
+
"learning_rate": 1.1768010685928685e-07,
|
| 1282 |
+
"loss": 0.2556,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 3.4968152866242037,
|
| 1287 |
+
"grad_norm": 0.31770785254670114,
|
| 1288 |
+
"learning_rate": 1.1673370855658591e-07,
|
| 1289 |
+
"loss": 0.2564,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 3.515923566878981,
|
| 1294 |
+
"grad_norm": 0.3077187664088862,
|
| 1295 |
+
"learning_rate": 1.1578576573543539e-07,
|
| 1296 |
+
"loss": 0.2603,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 3.535031847133758,
|
| 1301 |
+
"grad_norm": 0.27940138197404885,
|
| 1302 |
+
"learning_rate": 1.1483636589079626e-07,
|
| 1303 |
+
"loss": 0.2537,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 3.554140127388535,
|
| 1308 |
+
"grad_norm": 0.30920602027740407,
|
| 1309 |
+
"learning_rate": 1.138855966521124e-07,
|
| 1310 |
+
"loss": 0.2605,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 3.573248407643312,
|
| 1315 |
+
"grad_norm": 0.2878030111034016,
|
| 1316 |
+
"learning_rate": 1.1293354577522263e-07,
|
| 1317 |
+
"loss": 0.2642,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 3.5923566878980893,
|
| 1322 |
+
"grad_norm": 0.31242615729718354,
|
| 1323 |
+
"learning_rate": 1.1198030113426074e-07,
|
| 1324 |
+
"loss": 0.2689,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 3.611464968152866,
|
| 1329 |
+
"grad_norm": 0.2948055240186696,
|
| 1330 |
+
"learning_rate": 1.110259507135447e-07,
|
| 1331 |
+
"loss": 0.268,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 3.6305732484076434,
|
| 1336 |
+
"grad_norm": 0.2983537474983395,
|
| 1337 |
+
"learning_rate": 1.1007058259945583e-07,
|
| 1338 |
+
"loss": 0.25,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 3.6496815286624202,
|
| 1343 |
+
"grad_norm": 0.30237646104024896,
|
| 1344 |
+
"learning_rate": 1.0911428497230832e-07,
|
| 1345 |
+
"loss": 0.2398,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 3.6687898089171975,
|
| 1350 |
+
"grad_norm": 0.29700896373492647,
|
| 1351 |
+
"learning_rate": 1.0815714609821025e-07,
|
| 1352 |
+
"loss": 0.2568,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 3.6878980891719744,
|
| 1357 |
+
"grad_norm": 0.30679225984995734,
|
| 1358 |
+
"learning_rate": 1.071992543209167e-07,
|
| 1359 |
+
"loss": 0.2401,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 3.7070063694267517,
|
| 1364 |
+
"grad_norm": 0.3007632644983776,
|
| 1365 |
+
"learning_rate": 1.0624069805367557e-07,
|
| 1366 |
+
"loss": 0.2477,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 3.7261146496815285,
|
| 1371 |
+
"grad_norm": 0.3112671485408903,
|
| 1372 |
+
"learning_rate": 1.0528156577106702e-07,
|
| 1373 |
+
"loss": 0.2463,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.745222929936306,
|
| 1378 |
+
"grad_norm": 0.30401064479289264,
|
| 1379 |
+
"learning_rate": 1.0432194600083739e-07,
|
| 1380 |
+
"loss": 0.2574,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 3.7643312101910826,
|
| 1385 |
+
"grad_norm": 0.2851374502628386,
|
| 1386 |
+
"learning_rate": 1.0336192731572803e-07,
|
| 1387 |
+
"loss": 0.2582,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 3.78343949044586,
|
| 1392 |
+
"grad_norm": 0.2914379090084522,
|
| 1393 |
+
"learning_rate": 1.0240159832530007e-07,
|
| 1394 |
+
"loss": 0.2555,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 3.802547770700637,
|
| 1399 |
+
"grad_norm": 0.33178330655356186,
|
| 1400 |
+
"learning_rate": 1.0144104766775572e-07,
|
| 1401 |
+
"loss": 0.2614,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 3.821656050955414,
|
| 1406 |
+
"grad_norm": 0.31640502862571734,
|
| 1407 |
+
"learning_rate": 1.0048036400175708e-07,
|
| 1408 |
+
"loss": 0.235,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 3.840764331210191,
|
| 1413 |
+
"grad_norm": 0.3161044897697338,
|
| 1414 |
+
"learning_rate": 9.951963599824293e-08,
|
| 1415 |
+
"loss": 0.2433,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 3.859872611464968,
|
| 1420 |
+
"grad_norm": 0.2960452451709703,
|
| 1421 |
+
"learning_rate": 9.855895233224429e-08,
|
| 1422 |
+
"loss": 0.2589,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 3.8789808917197455,
|
| 1427 |
+
"grad_norm": 0.2881678617442765,
|
| 1428 |
+
"learning_rate": 9.759840167469994e-08,
|
| 1429 |
+
"loss": 0.257,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 3.8980891719745223,
|
| 1434 |
+
"grad_norm": 0.3020128060320192,
|
| 1435 |
+
"learning_rate": 9.663807268427197e-08,
|
| 1436 |
+
"loss": 0.2489,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 3.917197452229299,
|
| 1441 |
+
"grad_norm": 0.303673255967667,
|
| 1442 |
+
"learning_rate": 9.567805399916259e-08,
|
| 1443 |
+
"loss": 0.2552,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 3.9363057324840764,
|
| 1448 |
+
"grad_norm": 0.3029517107073107,
|
| 1449 |
+
"learning_rate": 9.471843422893297e-08,
|
| 1450 |
+
"loss": 0.2545,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 3.9554140127388537,
|
| 1455 |
+
"grad_norm": 0.29991781394420536,
|
| 1456 |
+
"learning_rate": 9.375930194632446e-08,
|
| 1457 |
+
"loss": 0.2541,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 3.9745222929936306,
|
| 1462 |
+
"grad_norm": 0.29801939082138607,
|
| 1463 |
+
"learning_rate": 9.28007456790833e-08,
|
| 1464 |
+
"loss": 0.2651,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 3.9936305732484074,
|
| 1469 |
+
"grad_norm": 0.3107836552290268,
|
| 1470 |
+
"learning_rate": 9.184285390178977e-08,
|
| 1471 |
+
"loss": 0.245,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 4.012738853503185,
|
| 1476 |
+
"grad_norm": 0.31402034579974897,
|
| 1477 |
+
"learning_rate": 9.088571502769167e-08,
|
| 1478 |
+
"loss": 0.261,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 4.031847133757962,
|
| 1483 |
+
"grad_norm": 0.29505379895390366,
|
| 1484 |
+
"learning_rate": 8.992941740054417e-08,
|
| 1485 |
+
"loss": 0.2435,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 4.050955414012739,
|
| 1490 |
+
"grad_norm": 0.29509128808113044,
|
| 1491 |
+
"learning_rate": 8.897404928645527e-08,
|
| 1492 |
+
"loss": 0.2445,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 4.070063694267516,
|
| 1497 |
+
"grad_norm": 0.29570344259163295,
|
| 1498 |
+
"learning_rate": 8.801969886573929e-08,
|
| 1499 |
+
"loss": 0.2698,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 4.089171974522293,
|
| 1504 |
+
"grad_norm": 0.29346662152654746,
|
| 1505 |
+
"learning_rate": 8.706645422477737e-08,
|
| 1506 |
+
"loss": 0.2597,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 4.10828025477707,
|
| 1511 |
+
"grad_norm": 0.2805185175645672,
|
| 1512 |
+
"learning_rate": 8.611440334788762e-08,
|
| 1513 |
+
"loss": 0.2586,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 4.127388535031847,
|
| 1518 |
+
"grad_norm": 0.30742002754181397,
|
| 1519 |
+
"learning_rate": 8.516363410920375e-08,
|
| 1520 |
+
"loss": 0.2478,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 4.146496815286624,
|
| 1525 |
+
"grad_norm": 0.2850884458824681,
|
| 1526 |
+
"learning_rate": 8.42142342645646e-08,
|
| 1527 |
+
"loss": 0.2602,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 4.165605095541402,
|
| 1532 |
+
"grad_norm": 0.306269262929328,
|
| 1533 |
+
"learning_rate": 8.326629144341405e-08,
|
| 1534 |
+
"loss": 0.2515,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 4.1847133757961785,
|
| 1539 |
+
"grad_norm": 0.3018034076236653,
|
| 1540 |
+
"learning_rate": 8.231989314071316e-08,
|
| 1541 |
+
"loss": 0.2471,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 4.203821656050955,
|
| 1546 |
+
"grad_norm": 0.3185658018423822,
|
| 1547 |
+
"learning_rate": 8.137512670886396e-08,
|
| 1548 |
+
"loss": 0.2615,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 4.222929936305732,
|
| 1553 |
+
"grad_norm": 0.29628833966395685,
|
| 1554 |
+
"learning_rate": 8.04320793496472e-08,
|
| 1555 |
+
"loss": 0.2577,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 4.24203821656051,
|
| 1560 |
+
"grad_norm": 0.2922501668352887,
|
| 1561 |
+
"learning_rate": 7.949083810617357e-08,
|
| 1562 |
+
"loss": 0.2705,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 4.261146496815287,
|
| 1567 |
+
"grad_norm": 0.2961392897242003,
|
| 1568 |
+
"learning_rate": 7.855148985484945e-08,
|
| 1569 |
+
"loss": 0.2499,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 4.280254777070064,
|
| 1574 |
+
"grad_norm": 0.2953084567404899,
|
| 1575 |
+
"learning_rate": 7.761412129735851e-08,
|
| 1576 |
+
"loss": 0.2529,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 4.2993630573248405,
|
| 1581 |
+
"grad_norm": 0.31348462098681923,
|
| 1582 |
+
"learning_rate": 7.667881895265893e-08,
|
| 1583 |
+
"loss": 0.2458,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 4.318471337579618,
|
| 1588 |
+
"grad_norm": 0.30412264596940947,
|
| 1589 |
+
"learning_rate": 7.574566914899778e-08,
|
| 1590 |
+
"loss": 0.2548,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 4.337579617834395,
|
| 1595 |
+
"grad_norm": 0.2847001566971413,
|
| 1596 |
+
"learning_rate": 7.481475801594301e-08,
|
| 1597 |
+
"loss": 0.2534,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 4.356687898089172,
|
| 1602 |
+
"grad_norm": 0.2853350717625523,
|
| 1603 |
+
"learning_rate": 7.38861714764337e-08,
|
| 1604 |
+
"loss": 0.2422,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 4.375796178343949,
|
| 1609 |
+
"grad_norm": 0.29979993619083456,
|
| 1610 |
+
"learning_rate": 7.29599952388492e-08,
|
| 1611 |
+
"loss": 0.2592,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 4.3949044585987265,
|
| 1616 |
+
"grad_norm": 0.3007849327371836,
|
| 1617 |
+
"learning_rate": 7.203631478909857e-08,
|
| 1618 |
+
"loss": 0.2487,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 4.414012738853503,
|
| 1623 |
+
"grad_norm": 0.2964965887396194,
|
| 1624 |
+
"learning_rate": 7.111521538272996e-08,
|
| 1625 |
+
"loss": 0.2591,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 4.43312101910828,
|
| 1630 |
+
"grad_norm": 0.29153090184611197,
|
| 1631 |
+
"learning_rate": 7.019678203706163e-08,
|
| 1632 |
+
"loss": 0.2506,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 4.452229299363057,
|
| 1637 |
+
"grad_norm": 0.31293925272597667,
|
| 1638 |
+
"learning_rate": 6.928109952333506e-08,
|
| 1639 |
+
"loss": 0.2545,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 4.471337579617835,
|
| 1644 |
+
"grad_norm": 0.31681568074468586,
|
| 1645 |
+
"learning_rate": 6.836825235889018e-08,
|
| 1646 |
+
"loss": 0.2566,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 4.490445859872612,
|
| 1651 |
+
"grad_norm": 0.3303594370705204,
|
| 1652 |
+
"learning_rate": 6.74583247993649e-08,
|
| 1653 |
+
"loss": 0.2528,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 4.509554140127388,
|
| 1658 |
+
"grad_norm": 0.3117652488003975,
|
| 1659 |
+
"learning_rate": 6.655140083091793e-08,
|
| 1660 |
+
"loss": 0.2467,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 4.528662420382165,
|
| 1665 |
+
"grad_norm": 0.3169598110655772,
|
| 1666 |
+
"learning_rate": 6.56475641624771e-08,
|
| 1667 |
+
"loss": 0.2639,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 4.547770700636943,
|
| 1672 |
+
"grad_norm": 0.3243563781200814,
|
| 1673 |
+
"learning_rate": 6.474689821801294e-08,
|
| 1674 |
+
"loss": 0.2687,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 4.56687898089172,
|
| 1679 |
+
"grad_norm": 0.289124780632532,
|
| 1680 |
+
"learning_rate": 6.384948612883871e-08,
|
| 1681 |
+
"loss": 0.2603,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.585987261146497,
|
| 1686 |
+
"grad_norm": 0.29440728893533646,
|
| 1687 |
+
"learning_rate": 6.29554107259374e-08,
|
| 1688 |
+
"loss": 0.2477,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 4.6050955414012735,
|
| 1693 |
+
"grad_norm": 0.28967915725187654,
|
| 1694 |
+
"learning_rate": 6.206475453231643e-08,
|
| 1695 |
+
"loss": 0.2498,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 4.624203821656051,
|
| 1700 |
+
"grad_norm": 0.3078511207653716,
|
| 1701 |
+
"learning_rate": 6.117759975539074e-08,
|
| 1702 |
+
"loss": 0.2536,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 4.643312101910828,
|
| 1707 |
+
"grad_norm": 0.3087724727800373,
|
| 1708 |
+
"learning_rate": 6.029402827939519e-08,
|
| 1709 |
+
"loss": 0.2475,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 4.662420382165605,
|
| 1714 |
+
"grad_norm": 0.30886308506307386,
|
| 1715 |
+
"learning_rate": 5.941412165782644e-08,
|
| 1716 |
+
"loss": 0.2646,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 4.681528662420382,
|
| 1721 |
+
"grad_norm": 0.30050126959049084,
|
| 1722 |
+
"learning_rate": 5.853796110591582e-08,
|
| 1723 |
+
"loss": 0.2516,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 4.7006369426751595,
|
| 1728 |
+
"grad_norm": 0.2863534353646146,
|
| 1729 |
+
"learning_rate": 5.7665627493133084e-08,
|
| 1730 |
+
"loss": 0.254,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 4.719745222929936,
|
| 1735 |
+
"grad_norm": 0.31186833842781353,
|
| 1736 |
+
"learning_rate": 5.6797201335722055e-08,
|
| 1737 |
+
"loss": 0.2636,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 4.738853503184713,
|
| 1742 |
+
"grad_norm": 0.33081843113519477,
|
| 1743 |
+
"learning_rate": 5.593276278926912e-08,
|
| 1744 |
+
"loss": 0.2439,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 4.757961783439491,
|
| 1749 |
+
"grad_norm": 0.3034226509534758,
|
| 1750 |
+
"learning_rate": 5.5072391641305003e-08,
|
| 1751 |
+
"loss": 0.2547,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 4.777070063694268,
|
| 1756 |
+
"grad_norm": 0.29722750131690423,
|
| 1757 |
+
"learning_rate": 5.4216167303939996e-08,
|
| 1758 |
+
"loss": 0.2526,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 4.796178343949045,
|
| 1763 |
+
"grad_norm": 0.2988145357986916,
|
| 1764 |
+
"learning_rate": 5.33641688065346e-08,
|
| 1765 |
+
"loss": 0.2547,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 4.8152866242038215,
|
| 1770 |
+
"grad_norm": 0.30646135021295096,
|
| 1771 |
+
"learning_rate": 5.251647478840511e-08,
|
| 1772 |
+
"loss": 0.2484,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 4.834394904458598,
|
| 1777 |
+
"grad_norm": 0.3157144375006897,
|
| 1778 |
+
"learning_rate": 5.167316349156494e-08,
|
| 1779 |
+
"loss": 0.2419,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 4.853503184713376,
|
| 1784 |
+
"grad_norm": 0.2983724975662055,
|
| 1785 |
+
"learning_rate": 5.0834312753503117e-08,
|
| 1786 |
+
"loss": 0.2589,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 4.872611464968153,
|
| 1791 |
+
"grad_norm": 0.3102633335125285,
|
| 1792 |
+
"learning_rate": 5.000000000000002e-08,
|
| 1793 |
+
"loss": 0.2599,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 4.89171974522293,
|
| 1798 |
+
"grad_norm": 0.29723998911569716,
|
| 1799 |
+
"learning_rate": 4.9170302237980564e-08,
|
| 1800 |
+
"loss": 0.2457,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 4.9108280254777075,
|
| 1805 |
+
"grad_norm": 0.3054372275805486,
|
| 1806 |
+
"learning_rate": 4.8345296048406856e-08,
|
| 1807 |
+
"loss": 0.2538,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 4.929936305732484,
|
| 1812 |
+
"grad_norm": 0.30363830684403537,
|
| 1813 |
+
"learning_rate": 4.752505757920977e-08,
|
| 1814 |
+
"loss": 0.2486,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 4.949044585987261,
|
| 1819 |
+
"grad_norm": 0.2898474470428452,
|
| 1820 |
+
"learning_rate": 4.6709662538260266e-08,
|
| 1821 |
+
"loss": 0.2581,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 4.968152866242038,
|
| 1826 |
+
"grad_norm": 0.2910012884854503,
|
| 1827 |
+
"learning_rate": 4.5899186186381725e-08,
|
| 1828 |
+
"loss": 0.2537,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 4.987261146496815,
|
| 1833 |
+
"grad_norm": 0.3149542929423315,
|
| 1834 |
+
"learning_rate": 4.5093703330403374e-08,
|
| 1835 |
+
"loss": 0.2535,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 5.006369426751593,
|
| 1840 |
+
"grad_norm": 0.3175825995451886,
|
| 1841 |
+
"learning_rate": 4.429328831625565e-08,
|
| 1842 |
+
"loss": 0.2585,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 5.025477707006369,
|
| 1847 |
+
"grad_norm": 0.3236452474738274,
|
| 1848 |
+
"learning_rate": 4.3498015022108e-08,
|
| 1849 |
+
"loss": 0.2653,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 5.044585987261146,
|
| 1854 |
+
"grad_norm": 0.29626391836670807,
|
| 1855 |
+
"learning_rate": 4.270795685155001e-08,
|
| 1856 |
+
"loss": 0.257,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 5.063694267515924,
|
| 1861 |
+
"grad_norm": 0.3026250583365394,
|
| 1862 |
+
"learning_rate": 4.1923186726816305e-08,
|
| 1863 |
+
"loss": 0.2559,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 5.082802547770701,
|
| 1868 |
+
"grad_norm": 0.2832298751120326,
|
| 1869 |
+
"learning_rate": 4.114377708205571e-08,
|
| 1870 |
+
"loss": 0.2627,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 5.101910828025478,
|
| 1875 |
+
"grad_norm": 0.29799965486813934,
|
| 1876 |
+
"learning_rate": 4.036979985664566e-08,
|
| 1877 |
+
"loss": 0.2494,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 5.1210191082802545,
|
| 1882 |
+
"grad_norm": 0.30594749582433756,
|
| 1883 |
+
"learning_rate": 3.9601326488552255e-08,
|
| 1884 |
+
"loss": 0.258,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 5.140127388535032,
|
| 1889 |
+
"grad_norm": 0.2762446500911052,
|
| 1890 |
+
"learning_rate": 3.883842790773647e-08,
|
| 1891 |
+
"loss": 0.2427,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 5.159235668789809,
|
| 1896 |
+
"grad_norm": 0.30019745369817896,
|
| 1897 |
+
"learning_rate": 3.808117452960734e-08,
|
| 1898 |
+
"loss": 0.2547,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 5.178343949044586,
|
| 1903 |
+
"grad_norm": 0.29753102050585134,
|
| 1904 |
+
"learning_rate": 3.732963624852274e-08,
|
| 1905 |
+
"loss": 0.2535,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 5.197452229299363,
|
| 1910 |
+
"grad_norm": 0.2935625407698436,
|
| 1911 |
+
"learning_rate": 3.658388243133804e-08,
|
| 1912 |
+
"loss": 0.2587,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 5.2165605095541405,
|
| 1917 |
+
"grad_norm": 0.3209655263508771,
|
| 1918 |
+
"learning_rate": 3.584398191100341e-08,
|
| 1919 |
+
"loss": 0.2452,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 5.235668789808917,
|
| 1924 |
+
"grad_norm": 0.31820460210115087,
|
| 1925 |
+
"learning_rate": 3.5110002980210973e-08,
|
| 1926 |
+
"loss": 0.2432,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 5.254777070063694,
|
| 1931 |
+
"grad_norm": 0.2992144995803717,
|
| 1932 |
+
"learning_rate": 3.438201338509098e-08,
|
| 1933 |
+
"loss": 0.2431,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 5.273885350318471,
|
| 1938 |
+
"grad_norm": 0.31045639532084096,
|
| 1939 |
+
"learning_rate": 3.366008031895904e-08,
|
| 1940 |
+
"loss": 0.2545,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 5.292993630573249,
|
| 1945 |
+
"grad_norm": 0.29354822656777496,
|
| 1946 |
+
"learning_rate": 3.294427041611425e-08,
|
| 1947 |
+
"loss": 0.2396,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 5.312101910828026,
|
| 1952 |
+
"grad_norm": 0.30124303336940966,
|
| 1953 |
+
"learning_rate": 3.223464974568874e-08,
|
| 1954 |
+
"loss": 0.2477,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 5.3312101910828025,
|
| 1959 |
+
"grad_norm": 0.30605544638967863,
|
| 1960 |
+
"learning_rate": 3.15312838055494e-08,
|
| 1961 |
+
"loss": 0.2591,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 5.350318471337579,
|
| 1966 |
+
"grad_norm": 0.31776982294402606,
|
| 1967 |
+
"learning_rate": 3.083423751625281e-08,
|
| 1968 |
+
"loss": 0.2515,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 5.369426751592357,
|
| 1973 |
+
"grad_norm": 0.30233438292813775,
|
| 1974 |
+
"learning_rate": 3.014357521505273e-08,
|
| 1975 |
+
"loss": 0.2609,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 5.388535031847134,
|
| 1980 |
+
"grad_norm": 0.29933329063857167,
|
| 1981 |
+
"learning_rate": 2.9459360649961896e-08,
|
| 1982 |
+
"loss": 0.2378,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 5.407643312101911,
|
| 1987 |
+
"grad_norm": 0.3093582978729572,
|
| 1988 |
+
"learning_rate": 2.878165697386812e-08,
|
| 1989 |
+
"loss": 0.2542,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 5.426751592356688,
|
| 1994 |
+
"grad_norm": 0.2992731993684574,
|
| 1995 |
+
"learning_rate": 2.811052673870534e-08,
|
| 1996 |
+
"loss": 0.2411,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 5.445859872611465,
|
| 2001 |
+
"grad_norm": 0.3240374627908033,
|
| 2002 |
+
"learning_rate": 2.7446031889679888e-08,
|
| 2003 |
+
"loss": 0.2483,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 5.464968152866242,
|
| 2008 |
+
"grad_norm": 0.3012591950583537,
|
| 2009 |
+
"learning_rate": 2.6788233759553138e-08,
|
| 2010 |
+
"loss": 0.2594,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 5.484076433121019,
|
| 2015 |
+
"grad_norm": 0.3232535802137296,
|
| 2016 |
+
"learning_rate": 2.61371930629805e-08,
|
| 2017 |
+
"loss": 0.2424,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 5.503184713375796,
|
| 2022 |
+
"grad_norm": 0.3024568075598484,
|
| 2023 |
+
"learning_rate": 2.549296989090738e-08,
|
| 2024 |
+
"loss": 0.2637,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 5.522292993630574,
|
| 2029 |
+
"grad_norm": 0.2862988052808537,
|
| 2030 |
+
"learning_rate": 2.4855623705022788e-08,
|
| 2031 |
+
"loss": 0.2656,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 5.54140127388535,
|
| 2036 |
+
"grad_norm": 0.30274335948775416,
|
| 2037 |
+
"learning_rate": 2.4225213332271198e-08,
|
| 2038 |
+
"loss": 0.2625,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 5.560509554140127,
|
| 2043 |
+
"grad_norm": 0.30372044645832685,
|
| 2044 |
+
"learning_rate": 2.3601796959422582e-08,
|
| 2045 |
+
"loss": 0.2534,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 5.579617834394904,
|
| 2050 |
+
"grad_norm": 0.2932251760911936,
|
| 2051 |
+
"learning_rate": 2.2985432127701942e-08,
|
| 2052 |
+
"loss": 0.2609,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 5.598726114649682,
|
| 2057 |
+
"grad_norm": 0.3003519183393173,
|
| 2058 |
+
"learning_rate": 2.237617572747834e-08,
|
| 2059 |
+
"loss": 0.2586,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 5.617834394904459,
|
| 2064 |
+
"grad_norm": 0.29295961875583704,
|
| 2065 |
+
"learning_rate": 2.1774083993013716e-08,
|
| 2066 |
+
"loss": 0.2514,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 5.6369426751592355,
|
| 2071 |
+
"grad_norm": 0.30284063679621004,
|
| 2072 |
+
"learning_rate": 2.117921249727258e-08,
|
| 2073 |
+
"loss": 0.2517,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 5.656050955414012,
|
| 2078 |
+
"grad_norm": 0.2944923570429743,
|
| 2079 |
+
"learning_rate": 2.0591616146792702e-08,
|
| 2080 |
+
"loss": 0.2571,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 5.67515923566879,
|
| 2085 |
+
"grad_norm": 0.2926540183721099,
|
| 2086 |
+
"learning_rate": 2.001134917661713e-08,
|
| 2087 |
+
"loss": 0.2699,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 5.694267515923567,
|
| 2092 |
+
"grad_norm": 0.28358364929517027,
|
| 2093 |
+
"learning_rate": 1.9438465145288373e-08,
|
| 2094 |
+
"loss": 0.2607,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 5.713375796178344,
|
| 2099 |
+
"grad_norm": 0.3058440366761672,
|
| 2100 |
+
"learning_rate": 1.8873016929904938e-08,
|
| 2101 |
+
"loss": 0.2545,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 5.732484076433121,
|
| 2106 |
+
"grad_norm": 0.2956033323033416,
|
| 2107 |
+
"learning_rate": 1.831505672124083e-08,
|
| 2108 |
+
"loss": 0.2441,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 5.751592356687898,
|
| 2113 |
+
"grad_norm": 0.3260543878256,
|
| 2114 |
+
"learning_rate": 1.776463601892825e-08,
|
| 2115 |
+
"loss": 0.2498,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 5.770700636942675,
|
| 2120 |
+
"grad_norm": 0.2815433176272766,
|
| 2121 |
+
"learning_rate": 1.7221805626704277e-08,
|
| 2122 |
+
"loss": 0.2561,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 5.789808917197452,
|
| 2127 |
+
"grad_norm": 0.32147374978023563,
|
| 2128 |
+
"learning_rate": 1.6686615647721637e-08,
|
| 2129 |
+
"loss": 0.2507,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 5.80891719745223,
|
| 2134 |
+
"grad_norm": 0.3136030754385911,
|
| 2135 |
+
"learning_rate": 1.615911547992426e-08,
|
| 2136 |
+
"loss": 0.2591,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 5.828025477707007,
|
| 2141 |
+
"grad_norm": 0.2925892696690556,
|
| 2142 |
+
"learning_rate": 1.5639353811487744e-08,
|
| 2143 |
+
"loss": 0.2487,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 5.8471337579617835,
|
| 2148 |
+
"grad_norm": 0.30292693895644507,
|
| 2149 |
+
"learning_rate": 1.5127378616325602e-08,
|
| 2150 |
+
"loss": 0.2514,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 5.86624203821656,
|
| 2155 |
+
"grad_norm": 0.31710406869483737,
|
| 2156 |
+
"learning_rate": 1.4623237149661139e-08,
|
| 2157 |
+
"loss": 0.2629,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 5.885350318471337,
|
| 2162 |
+
"grad_norm": 0.30143729522096213,
|
| 2163 |
+
"learning_rate": 1.4126975943665842e-08,
|
| 2164 |
+
"loss": 0.2388,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 5.904458598726115,
|
| 2169 |
+
"grad_norm": 0.2928856492144717,
|
| 2170 |
+
"learning_rate": 1.3638640803164514e-08,
|
| 2171 |
+
"loss": 0.2591,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 5.923566878980892,
|
| 2176 |
+
"grad_norm": 0.30225338228365756,
|
| 2177 |
+
"learning_rate": 1.3158276801407431e-08,
|
| 2178 |
+
"loss": 0.2549,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 5.942675159235669,
|
| 2183 |
+
"grad_norm": 0.31999882843186456,
|
| 2184 |
+
"learning_rate": 1.268592827591014e-08,
|
| 2185 |
+
"loss": 0.2552,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 5.961783439490446,
|
| 2190 |
+
"grad_norm": 0.3116709540538741,
|
| 2191 |
+
"learning_rate": 1.2221638824361069e-08,
|
| 2192 |
+
"loss": 0.2561,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 5.980891719745223,
|
| 2197 |
+
"grad_norm": 0.2852514751876125,
|
| 2198 |
+
"learning_rate": 1.1765451300597573e-08,
|
| 2199 |
+
"loss": 0.2494,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 6.0,
|
| 2204 |
+
"grad_norm": 0.30127697140383475,
|
| 2205 |
+
"learning_rate": 1.131740781065037e-08,
|
| 2206 |
+
"loss": 0.2677,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 6.019108280254777,
|
| 2211 |
+
"grad_norm": 0.299470849511753,
|
| 2212 |
+
"learning_rate": 1.0877549708857225e-08,
|
| 2213 |
+
"loss": 0.2492,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 6.038216560509555,
|
| 2218 |
+
"grad_norm": 0.29525030906520566,
|
| 2219 |
+
"learning_rate": 1.0445917594046071e-08,
|
| 2220 |
+
"loss": 0.2573,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 6.057324840764331,
|
| 2225 |
+
"grad_norm": 0.309895318768612,
|
| 2226 |
+
"learning_rate": 1.0022551305787563e-08,
|
| 2227 |
+
"loss": 0.2478,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 6.076433121019108,
|
| 2232 |
+
"grad_norm": 0.3116428679714083,
|
| 2233 |
+
"learning_rate": 9.607489920717981e-09,
|
| 2234 |
+
"loss": 0.2616,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 6.095541401273885,
|
| 2239 |
+
"grad_norm": 0.28485740043622854,
|
| 2240 |
+
"learning_rate": 9.200771748932512e-09,
|
| 2241 |
+
"loss": 0.23,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 6.114649681528663,
|
| 2246 |
+
"grad_norm": 0.30867096345949346,
|
| 2247 |
+
"learning_rate": 8.802434330449127e-09,
|
| 2248 |
+
"loss": 0.2423,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 6.13375796178344,
|
| 2253 |
+
"grad_norm": 0.29735720879032396,
|
| 2254 |
+
"learning_rate": 8.412514431743656e-09,
|
| 2255 |
+
"loss": 0.2506,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 6.1528662420382165,
|
| 2260 |
+
"grad_norm": 0.296939152603047,
|
| 2261 |
+
"learning_rate": 8.031048042356392e-09,
|
| 2262 |
+
"loss": 0.2518,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 6.171974522292993,
|
| 2267 |
+
"grad_norm": 0.3047288976089965,
|
| 2268 |
+
"learning_rate": 7.65807037157007e-09,
|
| 2269 |
+
"loss": 0.2571,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 6.191082802547771,
|
| 2274 |
+
"grad_norm": 0.30499824231597183,
|
| 2275 |
+
"learning_rate": 7.293615845160195e-09,
|
| 2276 |
+
"loss": 0.2492,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 6.210191082802548,
|
| 2281 |
+
"grad_norm": 0.28149771596925177,
|
| 2282 |
+
"learning_rate": 6.9377181022174604e-09,
|
| 2283 |
+
"loss": 0.2486,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 6.229299363057325,
|
| 2288 |
+
"grad_norm": 0.29778166813478346,
|
| 2289 |
+
"learning_rate": 6.590409992042956e-09,
|
| 2290 |
+
"loss": 0.253,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 6.248407643312102,
|
| 2295 |
+
"grad_norm": 0.3094630597200761,
|
| 2296 |
+
"learning_rate": 6.25172357111603e-09,
|
| 2297 |
+
"loss": 0.2552,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 6.267515923566879,
|
| 2302 |
+
"grad_norm": 0.3190922374282261,
|
| 2303 |
+
"learning_rate": 5.921690100135712e-09,
|
| 2304 |
+
"loss": 0.2585,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 6.286624203821656,
|
| 2309 |
+
"grad_norm": 0.3193448427368392,
|
| 2310 |
+
"learning_rate": 5.600340041135132e-09,
|
| 2311 |
+
"loss": 0.2566,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 6.305732484076433,
|
| 2316 |
+
"grad_norm": 0.30874010954068026,
|
| 2317 |
+
"learning_rate": 5.2877030546700115e-09,
|
| 2318 |
+
"loss": 0.2476,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 6.32484076433121,
|
| 2323 |
+
"grad_norm": 0.29530151136694954,
|
| 2324 |
+
"learning_rate": 4.9838079970809245e-09,
|
| 2325 |
+
"loss": 0.2526,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 6.343949044585988,
|
| 2330 |
+
"grad_norm": 0.29201059542728436,
|
| 2331 |
+
"learning_rate": 4.688682917829967e-09,
|
| 2332 |
+
"loss": 0.2743,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 6.3630573248407645,
|
| 2337 |
+
"grad_norm": 0.29014727881359437,
|
| 2338 |
+
"learning_rate": 4.402355056911655e-09,
|
| 2339 |
+
"loss": 0.2506,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 6.382165605095541,
|
| 2344 |
+
"grad_norm": 0.30021554157449915,
|
| 2345 |
+
"learning_rate": 4.124850842338778e-09,
|
| 2346 |
+
"loss": 0.2658,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 6.401273885350318,
|
| 2351 |
+
"grad_norm": 0.29191739918759346,
|
| 2352 |
+
"learning_rate": 3.856195887703095e-09,
|
| 2353 |
+
"loss": 0.2526,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 6.420382165605096,
|
| 2358 |
+
"grad_norm": 0.2970091109257102,
|
| 2359 |
+
"learning_rate": 3.5964149898111585e-09,
|
| 2360 |
+
"loss": 0.2515,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 6.439490445859873,
|
| 2365 |
+
"grad_norm": 0.301615837755627,
|
| 2366 |
+
"learning_rate": 3.345532126395578e-09,
|
| 2367 |
+
"loss": 0.2525,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 6.45859872611465,
|
| 2372 |
+
"grad_norm": 0.29483374535498386,
|
| 2373 |
+
"learning_rate": 3.103570453901938e-09,
|
| 2374 |
+
"loss": 0.2518,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 6.477707006369426,
|
| 2379 |
+
"grad_norm": 0.31193863122862164,
|
| 2380 |
+
"learning_rate": 2.8705523053513814e-09,
|
| 2381 |
+
"loss": 0.2581,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 6.496815286624204,
|
| 2386 |
+
"grad_norm": 0.3102645257081597,
|
| 2387 |
+
"learning_rate": 2.6464991882793277e-09,
|
| 2388 |
+
"loss": 0.2596,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 6.515923566878981,
|
| 2393 |
+
"grad_norm": 0.29720330855238497,
|
| 2394 |
+
"learning_rate": 2.4314317827503373e-09,
|
| 2395 |
+
"loss": 0.2482,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 6.535031847133758,
|
| 2400 |
+
"grad_norm": 0.30554779244411534,
|
| 2401 |
+
"learning_rate": 2.2253699394493065e-09,
|
| 2402 |
+
"loss": 0.2495,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 6.554140127388535,
|
| 2407 |
+
"grad_norm": 0.3132780643873276,
|
| 2408 |
+
"learning_rate": 2.0283326778492536e-09,
|
| 2409 |
+
"loss": 0.2445,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 6.573248407643312,
|
| 2414 |
+
"grad_norm": 0.2930656726881392,
|
| 2415 |
+
"learning_rate": 1.8403381844558808e-09,
|
| 2416 |
+
"loss": 0.2538,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 6.592356687898089,
|
| 2421 |
+
"grad_norm": 0.29674947300302884,
|
| 2422 |
+
"learning_rate": 1.661403811128903e-09,
|
| 2423 |
+
"loss": 0.2668,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 6.611464968152866,
|
| 2428 |
+
"grad_norm": 0.3075402793564451,
|
| 2429 |
+
"learning_rate": 1.4915460734805096e-09,
|
| 2430 |
+
"loss": 0.269,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 6.630573248407643,
|
| 2435 |
+
"grad_norm": 0.28966161314222383,
|
| 2436 |
+
"learning_rate": 1.3307806493509377e-09,
|
| 2437 |
+
"loss": 0.247,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 6.649681528662421,
|
| 2442 |
+
"grad_norm": 0.31994849636539213,
|
| 2443 |
+
"learning_rate": 1.1791223773614634e-09,
|
| 2444 |
+
"loss": 0.2594,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 6.6687898089171975,
|
| 2449 |
+
"grad_norm": 0.3026013256083666,
|
| 2450 |
+
"learning_rate": 1.036585255544764e-09,
|
| 2451 |
+
"loss": 0.2638,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 6.687898089171974,
|
| 2456 |
+
"grad_norm": 0.3132269351064342,
|
| 2457 |
+
"learning_rate": 9.031824400528854e-10,
|
| 2458 |
+
"loss": 0.2528,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 6.707006369426751,
|
| 2463 |
+
"grad_norm": 0.2882624273113626,
|
| 2464 |
+
"learning_rate": 7.789262439430012e-10,
|
| 2465 |
+
"loss": 0.2469,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 6.726114649681529,
|
| 2470 |
+
"grad_norm": 0.3064618613259807,
|
| 2471 |
+
"learning_rate": 6.638281360408338e-10,
|
| 2472 |
+
"loss": 0.2574,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 6.745222929936306,
|
| 2477 |
+
"grad_norm": 0.30382975584991506,
|
| 2478 |
+
"learning_rate": 5.578987398821344e-10,
|
| 2479 |
+
"loss": 0.2493,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 6.764331210191083,
|
| 2484 |
+
"grad_norm": 0.2902011016409033,
|
| 2485 |
+
"learning_rate": 4.611478327321339e-10,
|
| 2486 |
+
"loss": 0.2605,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 6.7834394904458595,
|
| 2491 |
+
"grad_norm": 0.29282370941155056,
|
| 2492 |
+
"learning_rate": 3.735843446830866e-10,
|
| 2493 |
+
"loss": 0.2531,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 6.802547770700637,
|
| 2498 |
+
"grad_norm": 0.3078439684683759,
|
| 2499 |
+
"learning_rate": 2.952163578300193e-10,
|
| 2500 |
+
"loss": 0.2473,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 6.821656050955414,
|
| 2505 |
+
"grad_norm": 0.3087832606093336,
|
| 2506 |
+
"learning_rate": 2.2605110552477157e-10,
|
| 2507 |
+
"loss": 0.2672,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 6.840764331210191,
|
| 2512 |
+
"grad_norm": 0.3098671142360526,
|
| 2513 |
+
"learning_rate": 1.6609497170834154e-10,
|
| 2514 |
+
"loss": 0.2569,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 6.859872611464969,
|
| 2519 |
+
"grad_norm": 0.29429296975358027,
|
| 2520 |
+
"learning_rate": 1.1535349032167907e-10,
|
| 2521 |
+
"loss": 0.2546,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 6.8789808917197455,
|
| 2526 |
+
"grad_norm": 0.31345754430512796,
|
| 2527 |
+
"learning_rate": 7.38313447948724e-11,
|
| 2528 |
+
"loss": 0.2639,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 6.898089171974522,
|
| 2533 |
+
"grad_norm": 0.2921055018298526,
|
| 2534 |
+
"learning_rate": 4.153236761488266e-11,
|
| 2535 |
+
"loss": 0.2612,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 6.917197452229299,
|
| 2540 |
+
"grad_norm": 0.28576822025063914,
|
| 2541 |
+
"learning_rate": 1.8459539971804605e-11,
|
| 2542 |
+
"loss": 0.2473,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 6.936305732484076,
|
| 2547 |
+
"grad_norm": 0.2826191180491914,
|
| 2548 |
+
"learning_rate": 4.614991483686825e-12,
|
| 2549 |
+
"loss": 0.2395,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 6.955414012738854,
|
| 2554 |
+
"grad_norm": 0.33090612760402355,
|
| 2555 |
+
"learning_rate": 0.0,
|
| 2556 |
+
"loss": 0.2531,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 6.955414012738854,
|
| 2561 |
+
"step": 364,
|
| 2562 |
+
"total_flos": 3.947982283988664e+17,
|
| 2563 |
+
"train_loss": 0.25456836733680505,
|
| 2564 |
+
"train_runtime": 5658.0504,
|
| 2565 |
+
"train_samples_per_second": 6.186,
|
| 2566 |
+
"train_steps_per_second": 0.064
|
| 2567 |
+
}
|
| 2568 |
+
],
|
| 2569 |
+
"logging_steps": 1,
|
| 2570 |
+
"max_steps": 364,
|
| 2571 |
+
"num_input_tokens_seen": 0,
|
| 2572 |
+
"num_train_epochs": 7,
|
| 2573 |
+
"save_steps": 500,
|
| 2574 |
+
"stateful_callbacks": {
|
| 2575 |
+
"TrainerControl": {
|
| 2576 |
+
"args": {
|
| 2577 |
+
"should_epoch_stop": false,
|
| 2578 |
+
"should_evaluate": false,
|
| 2579 |
+
"should_log": false,
|
| 2580 |
+
"should_save": true,
|
| 2581 |
+
"should_training_stop": true
|
| 2582 |
+
},
|
| 2583 |
+
"attributes": {}
|
| 2584 |
+
}
|
| 2585 |
+
},
|
| 2586 |
+
"total_flos": 3.947982283988664e+17,
|
| 2587 |
+
"train_batch_size": 1,
|
| 2588 |
+
"trial_name": null,
|
| 2589 |
+
"trial_params": null
|
| 2590 |
+
}
|
training_loss.png
ADDED
|