End of training
Browse files- README.md +3 -2
- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2581 -0
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: qwen2.5-7b-idea-review-mvp-1
|
|
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# qwen2.5-7b-idea-review-mvp-1
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 1.
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- lora
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: qwen2.5-7b-idea-review-mvp-1
|
|
|
|
| 16 |
|
| 17 |
# qwen2.5-7b-idea-review-mvp-1
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B](https://huggingface.co/Qwen/Qwen2.5-7B) on the openreview_idea_review dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 1.2680
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 1.2680052518844604,
|
| 4 |
+
"eval_runtime": 37.3047,
|
| 5 |
+
"eval_samples_per_second": 10.32,
|
| 6 |
+
"eval_steps_per_second": 2.6,
|
| 7 |
+
"perplexity": 3.553756638060757,
|
| 8 |
+
"total_flos": 234145141751808.0,
|
| 9 |
+
"train_loss": 1.2264163225196905,
|
| 10 |
+
"train_runtime": 4742.9935,
|
| 11 |
+
"train_samples_per_second": 24.104,
|
| 12 |
+
"train_steps_per_second": 0.753
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_loss": 1.2680052518844604,
|
| 4 |
+
"eval_runtime": 37.3047,
|
| 5 |
+
"eval_samples_per_second": 10.32,
|
| 6 |
+
"eval_steps_per_second": 2.6,
|
| 7 |
+
"perplexity": 3.553756638060757
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 234145141751808.0,
|
| 4 |
+
"train_loss": 1.2264163225196905,
|
| 5 |
+
"train_runtime": 4742.9935,
|
| 6 |
+
"train_samples_per_second": 24.104,
|
| 7 |
+
"train_steps_per_second": 0.753
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2581 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 600,
|
| 6 |
+
"global_step": 3573,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008396305625524769,
|
| 13 |
+
"grad_norm": 3.039360795711161,
|
| 14 |
+
"learning_rate": 1.3966480446927373e-06,
|
| 15 |
+
"loss": 1.7981,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.016792611251049538,
|
| 20 |
+
"grad_norm": 2.1081820956343322,
|
| 21 |
+
"learning_rate": 2.7932960893854746e-06,
|
| 22 |
+
"loss": 1.7816,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02518891687657431,
|
| 27 |
+
"grad_norm": 0.563623056034138,
|
| 28 |
+
"learning_rate": 4.189944134078212e-06,
|
| 29 |
+
"loss": 1.6797,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.033585222502099076,
|
| 34 |
+
"grad_norm": 0.4076099849250059,
|
| 35 |
+
"learning_rate": 5.586592178770949e-06,
|
| 36 |
+
"loss": 1.5657,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.041981528127623846,
|
| 41 |
+
"grad_norm": 0.30151480402917696,
|
| 42 |
+
"learning_rate": 6.983240223463687e-06,
|
| 43 |
+
"loss": 1.4759,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05037783375314862,
|
| 48 |
+
"grad_norm": 0.23775034251274657,
|
| 49 |
+
"learning_rate": 8.379888268156424e-06,
|
| 50 |
+
"loss": 1.4312,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05877413937867338,
|
| 55 |
+
"grad_norm": 0.2421182919342462,
|
| 56 |
+
"learning_rate": 9.776536312849161e-06,
|
| 57 |
+
"loss": 1.4082,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.06717044500419815,
|
| 62 |
+
"grad_norm": 0.23011928549656843,
|
| 63 |
+
"learning_rate": 1.1173184357541899e-05,
|
| 64 |
+
"loss": 1.4263,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.07556675062972293,
|
| 69 |
+
"grad_norm": 0.21851210361234266,
|
| 70 |
+
"learning_rate": 1.2569832402234637e-05,
|
| 71 |
+
"loss": 1.4216,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.08396305625524769,
|
| 76 |
+
"grad_norm": 0.23106315932058966,
|
| 77 |
+
"learning_rate": 1.3966480446927374e-05,
|
| 78 |
+
"loss": 1.4207,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.09235936188077246,
|
| 83 |
+
"grad_norm": 0.23062228974277152,
|
| 84 |
+
"learning_rate": 1.5363128491620113e-05,
|
| 85 |
+
"loss": 1.399,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.10075566750629723,
|
| 90 |
+
"grad_norm": 0.23855661202824083,
|
| 91 |
+
"learning_rate": 1.675977653631285e-05,
|
| 92 |
+
"loss": 1.404,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.109151973131822,
|
| 97 |
+
"grad_norm": 0.2302715158226374,
|
| 98 |
+
"learning_rate": 1.8156424581005588e-05,
|
| 99 |
+
"loss": 1.4057,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.11754827875734676,
|
| 104 |
+
"grad_norm": 0.23018949408838,
|
| 105 |
+
"learning_rate": 1.9553072625698323e-05,
|
| 106 |
+
"loss": 1.3956,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.12594458438287154,
|
| 111 |
+
"grad_norm": 0.23103802546960453,
|
| 112 |
+
"learning_rate": 2.0949720670391062e-05,
|
| 113 |
+
"loss": 1.4179,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.1343408900083963,
|
| 118 |
+
"grad_norm": 0.2287991193827312,
|
| 119 |
+
"learning_rate": 2.2346368715083797e-05,
|
| 120 |
+
"loss": 1.412,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.14273719563392107,
|
| 125 |
+
"grad_norm": 0.22988402418477955,
|
| 126 |
+
"learning_rate": 2.3743016759776536e-05,
|
| 127 |
+
"loss": 1.39,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.15113350125944586,
|
| 132 |
+
"grad_norm": 0.21974152422479476,
|
| 133 |
+
"learning_rate": 2.5139664804469275e-05,
|
| 134 |
+
"loss": 1.4004,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.15952980688497062,
|
| 139 |
+
"grad_norm": 0.23284763880742212,
|
| 140 |
+
"learning_rate": 2.6536312849162014e-05,
|
| 141 |
+
"loss": 1.386,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.16792611251049538,
|
| 146 |
+
"grad_norm": 0.21486774869633107,
|
| 147 |
+
"learning_rate": 2.793296089385475e-05,
|
| 148 |
+
"loss": 1.4012,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.17632241813602015,
|
| 153 |
+
"grad_norm": 0.22783180948472934,
|
| 154 |
+
"learning_rate": 2.9329608938547488e-05,
|
| 155 |
+
"loss": 1.3707,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1847187237615449,
|
| 160 |
+
"grad_norm": 0.22399716105168718,
|
| 161 |
+
"learning_rate": 3.0726256983240227e-05,
|
| 162 |
+
"loss": 1.4125,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.19311502938706968,
|
| 167 |
+
"grad_norm": 0.2147254679707046,
|
| 168 |
+
"learning_rate": 3.212290502793296e-05,
|
| 169 |
+
"loss": 1.3699,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.20151133501259447,
|
| 174 |
+
"grad_norm": 0.21426390076476165,
|
| 175 |
+
"learning_rate": 3.35195530726257e-05,
|
| 176 |
+
"loss": 1.374,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.20990764063811923,
|
| 181 |
+
"grad_norm": 0.22424846384922523,
|
| 182 |
+
"learning_rate": 3.491620111731844e-05,
|
| 183 |
+
"loss": 1.3788,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.218303946263644,
|
| 188 |
+
"grad_norm": 0.21440605091273993,
|
| 189 |
+
"learning_rate": 3.6312849162011175e-05,
|
| 190 |
+
"loss": 1.3565,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.22670025188916876,
|
| 195 |
+
"grad_norm": 0.21539799426915837,
|
| 196 |
+
"learning_rate": 3.770949720670392e-05,
|
| 197 |
+
"loss": 1.3521,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.23509655751469352,
|
| 202 |
+
"grad_norm": 0.22056667191105353,
|
| 203 |
+
"learning_rate": 3.9106145251396646e-05,
|
| 204 |
+
"loss": 1.3613,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2434928631402183,
|
| 209 |
+
"grad_norm": 0.2041245512715553,
|
| 210 |
+
"learning_rate": 4.050279329608939e-05,
|
| 211 |
+
"loss": 1.36,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2518891687657431,
|
| 216 |
+
"grad_norm": 0.2053261976415489,
|
| 217 |
+
"learning_rate": 4.1899441340782123e-05,
|
| 218 |
+
"loss": 1.3799,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.26028547439126787,
|
| 223 |
+
"grad_norm": 0.20324705222690662,
|
| 224 |
+
"learning_rate": 4.3296089385474866e-05,
|
| 225 |
+
"loss": 1.3797,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.2686817800167926,
|
| 230 |
+
"grad_norm": 0.20526982003231917,
|
| 231 |
+
"learning_rate": 4.4692737430167594e-05,
|
| 232 |
+
"loss": 1.3675,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.2770780856423174,
|
| 237 |
+
"grad_norm": 0.20273779659012153,
|
| 238 |
+
"learning_rate": 4.6089385474860336e-05,
|
| 239 |
+
"loss": 1.3888,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.28547439126784213,
|
| 244 |
+
"grad_norm": 0.2061146702772721,
|
| 245 |
+
"learning_rate": 4.748603351955307e-05,
|
| 246 |
+
"loss": 1.3626,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.2938706968933669,
|
| 251 |
+
"grad_norm": 0.18820362655678458,
|
| 252 |
+
"learning_rate": 4.8882681564245814e-05,
|
| 253 |
+
"loss": 1.3598,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.3022670025188917,
|
| 258 |
+
"grad_norm": 0.21352845647640323,
|
| 259 |
+
"learning_rate": 4.9999952257225154e-05,
|
| 260 |
+
"loss": 1.3933,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.31066330814441645,
|
| 265 |
+
"grad_norm": 0.21205929000762902,
|
| 266 |
+
"learning_rate": 4.9998281279252004e-05,
|
| 267 |
+
"loss": 1.3723,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.31905961376994124,
|
| 272 |
+
"grad_norm": 0.20026396935477256,
|
| 273 |
+
"learning_rate": 4.9994223344882976e-05,
|
| 274 |
+
"loss": 1.3889,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.327455919395466,
|
| 279 |
+
"grad_norm": 0.2027355098630582,
|
| 280 |
+
"learning_rate": 4.9987778841589184e-05,
|
| 281 |
+
"loss": 1.3757,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.33585222502099077,
|
| 286 |
+
"grad_norm": 0.1917389523789995,
|
| 287 |
+
"learning_rate": 4.997894838472288e-05,
|
| 288 |
+
"loss": 1.3602,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.34424853064651556,
|
| 293 |
+
"grad_norm": 0.19386030500153506,
|
| 294 |
+
"learning_rate": 4.996773281745864e-05,
|
| 295 |
+
"loss": 1.3483,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.3526448362720403,
|
| 300 |
+
"grad_norm": 0.18924586602111995,
|
| 301 |
+
"learning_rate": 4.9954133210712904e-05,
|
| 302 |
+
"loss": 1.3773,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.3610411418975651,
|
| 307 |
+
"grad_norm": 0.19404566415989755,
|
| 308 |
+
"learning_rate": 4.9938150863041666e-05,
|
| 309 |
+
"loss": 1.3817,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3694374475230898,
|
| 314 |
+
"grad_norm": 0.1864716184637937,
|
| 315 |
+
"learning_rate": 4.991978730051653e-05,
|
| 316 |
+
"loss": 1.3436,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.3778337531486146,
|
| 321 |
+
"grad_norm": 0.18756275917822954,
|
| 322 |
+
"learning_rate": 4.9899044276578956e-05,
|
| 323 |
+
"loss": 1.376,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.38623005877413935,
|
| 328 |
+
"grad_norm": 0.19103603911697656,
|
| 329 |
+
"learning_rate": 4.987592377187286e-05,
|
| 330 |
+
"loss": 1.3699,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.39462636439966414,
|
| 335 |
+
"grad_norm": 0.19933360659538543,
|
| 336 |
+
"learning_rate": 4.985042799405547e-05,
|
| 337 |
+
"loss": 1.3831,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.40302267002518893,
|
| 342 |
+
"grad_norm": 0.19498666555973088,
|
| 343 |
+
"learning_rate": 4.982255937758657e-05,
|
| 344 |
+
"loss": 1.3579,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.41141897565071367,
|
| 349 |
+
"grad_norm": 0.19300141279658406,
|
| 350 |
+
"learning_rate": 4.9792320583495975e-05,
|
| 351 |
+
"loss": 1.348,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.41981528127623846,
|
| 356 |
+
"grad_norm": 0.19742724416821206,
|
| 357 |
+
"learning_rate": 4.975971449912952e-05,
|
| 358 |
+
"loss": 1.3474,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.4282115869017632,
|
| 363 |
+
"grad_norm": 0.19421992100107466,
|
| 364 |
+
"learning_rate": 4.972474423787331e-05,
|
| 365 |
+
"loss": 1.3609,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.436607892527288,
|
| 370 |
+
"grad_norm": 0.2078937193948646,
|
| 371 |
+
"learning_rate": 4.9687413138856446e-05,
|
| 372 |
+
"loss": 1.3829,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.4450041981528128,
|
| 377 |
+
"grad_norm": 0.19272346953581754,
|
| 378 |
+
"learning_rate": 4.964772476663222e-05,
|
| 379 |
+
"loss": 1.3554,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.4534005037783375,
|
| 384 |
+
"grad_norm": 0.19281909541060518,
|
| 385 |
+
"learning_rate": 4.960568291083772e-05,
|
| 386 |
+
"loss": 1.3515,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.4617968094038623,
|
| 391 |
+
"grad_norm": 0.19781631242122302,
|
| 392 |
+
"learning_rate": 4.956129158583199e-05,
|
| 393 |
+
"loss": 1.3547,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.47019311502938704,
|
| 398 |
+
"grad_norm": 0.21165698099082017,
|
| 399 |
+
"learning_rate": 4.951455503031271e-05,
|
| 400 |
+
"loss": 1.329,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.47858942065491183,
|
| 405 |
+
"grad_norm": 0.1954676247733346,
|
| 406 |
+
"learning_rate": 4.946547770691152e-05,
|
| 407 |
+
"loss": 1.3547,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.4869857262804366,
|
| 412 |
+
"grad_norm": 0.1934746445894875,
|
| 413 |
+
"learning_rate": 4.9414064301767786e-05,
|
| 414 |
+
"loss": 1.3421,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.49538203190596136,
|
| 419 |
+
"grad_norm": 0.19263774524822871,
|
| 420 |
+
"learning_rate": 4.9360319724081285e-05,
|
| 421 |
+
"loss": 1.3576,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5037783375314862,
|
| 426 |
+
"grad_norm": 0.18876853207737465,
|
| 427 |
+
"learning_rate": 4.930424910564334e-05,
|
| 428 |
+
"loss": 1.3667,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.5037783375314862,
|
| 433 |
+
"eval_loss": 1.3807679414749146,
|
| 434 |
+
"eval_runtime": 37.4693,
|
| 435 |
+
"eval_samples_per_second": 10.275,
|
| 436 |
+
"eval_steps_per_second": 2.589,
|
| 437 |
+
"step": 600
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5121746431570109,
|
| 441 |
+
"grad_norm": 0.19209309775474845,
|
| 442 |
+
"learning_rate": 4.92458578003469e-05,
|
| 443 |
+
"loss": 1.3504,
|
| 444 |
+
"step": 610
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5205709487825357,
|
| 448 |
+
"grad_norm": 0.19079678945155537,
|
| 449 |
+
"learning_rate": 4.9185151383675246e-05,
|
| 450 |
+
"loss": 1.339,
|
| 451 |
+
"step": 620
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5289672544080605,
|
| 455 |
+
"grad_norm": 0.194534067082253,
|
| 456 |
+
"learning_rate": 4.912213565216966e-05,
|
| 457 |
+
"loss": 1.3358,
|
| 458 |
+
"step": 630
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5373635600335852,
|
| 462 |
+
"grad_norm": 0.19307901988822676,
|
| 463 |
+
"learning_rate": 4.905681662287595e-05,
|
| 464 |
+
"loss": 1.3708,
|
| 465 |
+
"step": 640
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5457598656591099,
|
| 469 |
+
"grad_norm": 0.1990057756809498,
|
| 470 |
+
"learning_rate": 4.898920053276989e-05,
|
| 471 |
+
"loss": 1.363,
|
| 472 |
+
"step": 650
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.5541561712846348,
|
| 476 |
+
"grad_norm": 0.2010697996628507,
|
| 477 |
+
"learning_rate": 4.891929383816169e-05,
|
| 478 |
+
"loss": 1.3993,
|
| 479 |
+
"step": 660
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.5625524769101595,
|
| 483 |
+
"grad_norm": 0.20209675418432702,
|
| 484 |
+
"learning_rate": 4.8847103214079524e-05,
|
| 485 |
+
"loss": 1.351,
|
| 486 |
+
"step": 670
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.5709487825356843,
|
| 490 |
+
"grad_norm": 0.20138078865908093,
|
| 491 |
+
"learning_rate": 4.877263555363216e-05,
|
| 492 |
+
"loss": 1.3494,
|
| 493 |
+
"step": 680
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.5793450881612091,
|
| 497 |
+
"grad_norm": 0.20340329720637837,
|
| 498 |
+
"learning_rate": 4.869589796735078e-05,
|
| 499 |
+
"loss": 1.3699,
|
| 500 |
+
"step": 690
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5877413937867338,
|
| 504 |
+
"grad_norm": 0.18554034872359323,
|
| 505 |
+
"learning_rate": 4.8616897782510006e-05,
|
| 506 |
+
"loss": 1.3358,
|
| 507 |
+
"step": 700
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.5961376994122586,
|
| 511 |
+
"grad_norm": 0.20470702291536869,
|
| 512 |
+
"learning_rate": 4.853564254242831e-05,
|
| 513 |
+
"loss": 1.3417,
|
| 514 |
+
"step": 710
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6045340050377834,
|
| 518 |
+
"grad_norm": 0.20233392948438,
|
| 519 |
+
"learning_rate": 4.845214000574768e-05,
|
| 520 |
+
"loss": 1.3633,
|
| 521 |
+
"step": 720
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6129303106633082,
|
| 525 |
+
"grad_norm": 0.19552940545300115,
|
| 526 |
+
"learning_rate": 4.8366398145692824e-05,
|
| 527 |
+
"loss": 1.3582,
|
| 528 |
+
"step": 730
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6213266162888329,
|
| 532 |
+
"grad_norm": 0.20658182549440357,
|
| 533 |
+
"learning_rate": 4.827842514930985e-05,
|
| 534 |
+
"loss": 1.3702,
|
| 535 |
+
"step": 740
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.6297229219143576,
|
| 539 |
+
"grad_norm": 0.2029780054195454,
|
| 540 |
+
"learning_rate": 4.81882294166845e-05,
|
| 541 |
+
"loss": 1.3807,
|
| 542 |
+
"step": 750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6381192275398825,
|
| 546 |
+
"grad_norm": 0.19667484329069757,
|
| 547 |
+
"learning_rate": 4.8095819560140096e-05,
|
| 548 |
+
"loss": 1.3593,
|
| 549 |
+
"step": 760
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.6465155331654072,
|
| 553 |
+
"grad_norm": 0.2048531959544165,
|
| 554 |
+
"learning_rate": 4.8001204403415204e-05,
|
| 555 |
+
"loss": 1.3583,
|
| 556 |
+
"step": 770
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.654911838790932,
|
| 560 |
+
"grad_norm": 0.2036838290154731,
|
| 561 |
+
"learning_rate": 4.7904392980821035e-05,
|
| 562 |
+
"loss": 1.3478,
|
| 563 |
+
"step": 780
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.6633081444164568,
|
| 567 |
+
"grad_norm": 0.20036896998141893,
|
| 568 |
+
"learning_rate": 4.78053945363789e-05,
|
| 569 |
+
"loss": 1.361,
|
| 570 |
+
"step": 790
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.6717044500419815,
|
| 574 |
+
"grad_norm": 0.19874326596454528,
|
| 575 |
+
"learning_rate": 4.770421852293746e-05,
|
| 576 |
+
"loss": 1.3445,
|
| 577 |
+
"step": 800
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.6801007556675063,
|
| 581 |
+
"grad_norm": 0.20224592915430473,
|
| 582 |
+
"learning_rate": 4.760087460127019e-05,
|
| 583 |
+
"loss": 1.3328,
|
| 584 |
+
"step": 810
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.6884970612930311,
|
| 588 |
+
"grad_norm": 0.20341065735706346,
|
| 589 |
+
"learning_rate": 4.749537263915288e-05,
|
| 590 |
+
"loss": 1.327,
|
| 591 |
+
"step": 820
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.6968933669185559,
|
| 595 |
+
"grad_norm": 0.19878763475266986,
|
| 596 |
+
"learning_rate": 4.7387722710421414e-05,
|
| 597 |
+
"loss": 1.3508,
|
| 598 |
+
"step": 830
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.7052896725440806,
|
| 602 |
+
"grad_norm": 0.1901873610814321,
|
| 603 |
+
"learning_rate": 4.7277935094009896e-05,
|
| 604 |
+
"loss": 1.331,
|
| 605 |
+
"step": 840
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.7136859781696053,
|
| 609 |
+
"grad_norm": 0.20858107457668973,
|
| 610 |
+
"learning_rate": 4.716602027296916e-05,
|
| 611 |
+
"loss": 1.3398,
|
| 612 |
+
"step": 850
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7220822837951302,
|
| 616 |
+
"grad_norm": 0.2127448740445764,
|
| 617 |
+
"learning_rate": 4.705198893346578e-05,
|
| 618 |
+
"loss": 1.3557,
|
| 619 |
+
"step": 860
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.7304785894206549,
|
| 623 |
+
"grad_norm": 0.2052323089476111,
|
| 624 |
+
"learning_rate": 4.693585196376171e-05,
|
| 625 |
+
"loss": 1.3422,
|
| 626 |
+
"step": 870
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.7388748950461796,
|
| 630 |
+
"grad_norm": 0.2034886804011326,
|
| 631 |
+
"learning_rate": 4.681762045317464e-05,
|
| 632 |
+
"loss": 1.3767,
|
| 633 |
+
"step": 880
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.7472712006717045,
|
| 637 |
+
"grad_norm": 0.20205914104891914,
|
| 638 |
+
"learning_rate": 4.6697305691019105e-05,
|
| 639 |
+
"loss": 1.3459,
|
| 640 |
+
"step": 890
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.7556675062972292,
|
| 644 |
+
"grad_norm": 0.1947344074048,
|
| 645 |
+
"learning_rate": 4.6574919165528567e-05,
|
| 646 |
+
"loss": 1.3253,
|
| 647 |
+
"step": 900
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.764063811922754,
|
| 651 |
+
"grad_norm": 0.20617146845645057,
|
| 652 |
+
"learning_rate": 4.645047256275839e-05,
|
| 653 |
+
"loss": 1.3419,
|
| 654 |
+
"step": 910
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.7724601175482787,
|
| 658 |
+
"grad_norm": 0.19999709652925082,
|
| 659 |
+
"learning_rate": 4.632397776547011e-05,
|
| 660 |
+
"loss": 1.3478,
|
| 661 |
+
"step": 920
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.7808564231738035,
|
| 665 |
+
"grad_norm": 0.20682503408598918,
|
| 666 |
+
"learning_rate": 4.619544685199669e-05,
|
| 667 |
+
"loss": 1.3391,
|
| 668 |
+
"step": 930
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.7892527287993283,
|
| 672 |
+
"grad_norm": 0.2111768615345554,
|
| 673 |
+
"learning_rate": 4.6064892095089326e-05,
|
| 674 |
+
"loss": 1.3445,
|
| 675 |
+
"step": 940
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.797649034424853,
|
| 679 |
+
"grad_norm": 0.2081975290000968,
|
| 680 |
+
"learning_rate": 4.59323259607455e-05,
|
| 681 |
+
"loss": 1.3167,
|
| 682 |
+
"step": 950
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.8060453400503779,
|
| 686 |
+
"grad_norm": 0.20101213815546085,
|
| 687 |
+
"learning_rate": 4.5797761107018754e-05,
|
| 688 |
+
"loss": 1.3383,
|
| 689 |
+
"step": 960
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8144416456759026,
|
| 693 |
+
"grad_norm": 0.2174273103650515,
|
| 694 |
+
"learning_rate": 4.566121038280995e-05,
|
| 695 |
+
"loss": 1.3404,
|
| 696 |
+
"step": 970
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.8228379513014273,
|
| 700 |
+
"grad_norm": 0.19965611041565887,
|
| 701 |
+
"learning_rate": 4.552268682664045e-05,
|
| 702 |
+
"loss": 1.3333,
|
| 703 |
+
"step": 980
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.8312342569269522,
|
| 707 |
+
"grad_norm": 0.20691842606557614,
|
| 708 |
+
"learning_rate": 4.538220366540713e-05,
|
| 709 |
+
"loss": 1.3182,
|
| 710 |
+
"step": 990
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.8396305625524769,
|
| 714 |
+
"grad_norm": 0.21181131160704766,
|
| 715 |
+
"learning_rate": 4.52397743131194e-05,
|
| 716 |
+
"loss": 1.3494,
|
| 717 |
+
"step": 1000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.8480268681780017,
|
| 721 |
+
"grad_norm": 0.20088379875239834,
|
| 722 |
+
"learning_rate": 4.509541236961835e-05,
|
| 723 |
+
"loss": 1.3372,
|
| 724 |
+
"step": 1010
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.8564231738035264,
|
| 728 |
+
"grad_norm": 0.21111113264702316,
|
| 729 |
+
"learning_rate": 4.4949131619278233e-05,
|
| 730 |
+
"loss": 1.3459,
|
| 731 |
+
"step": 1020
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.8648194794290512,
|
| 735 |
+
"grad_norm": 0.2099829717112053,
|
| 736 |
+
"learning_rate": 4.4800946029690205e-05,
|
| 737 |
+
"loss": 1.3288,
|
| 738 |
+
"step": 1030
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.873215785054576,
|
| 742 |
+
"grad_norm": 0.217914005065172,
|
| 743 |
+
"learning_rate": 4.465086975032866e-05,
|
| 744 |
+
"loss": 1.3524,
|
| 745 |
+
"step": 1040
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.8816120906801007,
|
| 749 |
+
"grad_norm": 0.2063304535624245,
|
| 750 |
+
"learning_rate": 4.4498917111200154e-05,
|
| 751 |
+
"loss": 1.3233,
|
| 752 |
+
"step": 1050
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.8900083963056256,
|
| 756 |
+
"grad_norm": 0.21350579375508957,
|
| 757 |
+
"learning_rate": 4.434510262147514e-05,
|
| 758 |
+
"loss": 1.3486,
|
| 759 |
+
"step": 1060
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.8984047019311503,
|
| 763 |
+
"grad_norm": 0.2091334109956843,
|
| 764 |
+
"learning_rate": 4.418944096810253e-05,
|
| 765 |
+
"loss": 1.3375,
|
| 766 |
+
"step": 1070
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.906801007556675,
|
| 770 |
+
"grad_norm": 0.20954696844936122,
|
| 771 |
+
"learning_rate": 4.403194701440733e-05,
|
| 772 |
+
"loss": 1.3428,
|
| 773 |
+
"step": 1080
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.9151973131821999,
|
| 777 |
+
"grad_norm": 0.21790164835706194,
|
| 778 |
+
"learning_rate": 4.3872635798671384e-05,
|
| 779 |
+
"loss": 1.3312,
|
| 780 |
+
"step": 1090
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.9235936188077246,
|
| 784 |
+
"grad_norm": 0.21625055602980792,
|
| 785 |
+
"learning_rate": 4.3711522532697526e-05,
|
| 786 |
+
"loss": 1.3268,
|
| 787 |
+
"step": 1100
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.9319899244332494,
|
| 791 |
+
"grad_norm": 0.21992048192275737,
|
| 792 |
+
"learning_rate": 4.354862260035699e-05,
|
| 793 |
+
"loss": 1.2904,
|
| 794 |
+
"step": 1110
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.9403862300587741,
|
| 798 |
+
"grad_norm": 0.22805590389300254,
|
| 799 |
+
"learning_rate": 4.338395155612055e-05,
|
| 800 |
+
"loss": 1.3352,
|
| 801 |
+
"step": 1120
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.9487825356842989,
|
| 805 |
+
"grad_norm": 0.2300433128490517,
|
| 806 |
+
"learning_rate": 4.321752512357328e-05,
|
| 807 |
+
"loss": 1.3327,
|
| 808 |
+
"step": 1130
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.9571788413098237,
|
| 812 |
+
"grad_norm": 0.2564818367135206,
|
| 813 |
+
"learning_rate": 4.304935919391317e-05,
|
| 814 |
+
"loss": 1.3293,
|
| 815 |
+
"step": 1140
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.9655751469353484,
|
| 819 |
+
"grad_norm": 0.21597640959608433,
|
| 820 |
+
"learning_rate": 4.2879469824433814e-05,
|
| 821 |
+
"loss": 1.3171,
|
| 822 |
+
"step": 1150
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.9739714525608733,
|
| 826 |
+
"grad_norm": 0.21055331712423173,
|
| 827 |
+
"learning_rate": 4.2707873236991116e-05,
|
| 828 |
+
"loss": 1.3291,
|
| 829 |
+
"step": 1160
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.982367758186398,
|
| 833 |
+
"grad_norm": 0.21558556200945594,
|
| 834 |
+
"learning_rate": 4.2534585816454384e-05,
|
| 835 |
+
"loss": 1.2906,
|
| 836 |
+
"step": 1170
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.9907640638119227,
|
| 840 |
+
"grad_norm": 0.2134813755097652,
|
| 841 |
+
"learning_rate": 4.235962410914185e-05,
|
| 842 |
+
"loss": 1.3516,
|
| 843 |
+
"step": 1180
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.9991603694374476,
|
| 847 |
+
"grad_norm": 0.22460106184588927,
|
| 848 |
+
"learning_rate": 4.218300482124069e-05,
|
| 849 |
+
"loss": 1.3249,
|
| 850 |
+
"step": 1190
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.0075566750629723,
|
| 854 |
+
"grad_norm": 0.2510896281590031,
|
| 855 |
+
"learning_rate": 4.2004744817211866e-05,
|
| 856 |
+
"loss": 1.2766,
|
| 857 |
+
"step": 1200
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.0075566750629723,
|
| 861 |
+
"eval_loss": 1.3506666421890259,
|
| 862 |
+
"eval_runtime": 37.3255,
|
| 863 |
+
"eval_samples_per_second": 10.315,
|
| 864 |
+
"eval_steps_per_second": 2.599,
|
| 865 |
+
"step": 1200
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.015952980688497,
|
| 869 |
+
"grad_norm": 0.2491416849165635,
|
| 870 |
+
"learning_rate": 4.182486111817983e-05,
|
| 871 |
+
"loss": 1.2816,
|
| 872 |
+
"step": 1210
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.0243492863140218,
|
| 876 |
+
"grad_norm": 0.2711674700370134,
|
| 877 |
+
"learning_rate": 4.164337090030727e-05,
|
| 878 |
+
"loss": 1.2815,
|
| 879 |
+
"step": 1220
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.0327455919395465,
|
| 883 |
+
"grad_norm": 0.25237777386220234,
|
| 884 |
+
"learning_rate": 4.146029149315502e-05,
|
| 885 |
+
"loss": 1.2692,
|
| 886 |
+
"step": 1230
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.0411418975650715,
|
| 890 |
+
"grad_norm": 0.24173423889336904,
|
| 891 |
+
"learning_rate": 4.1275640378027367e-05,
|
| 892 |
+
"loss": 1.2572,
|
| 893 |
+
"step": 1240
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.0495382031905962,
|
| 897 |
+
"grad_norm": 0.2516517653129535,
|
| 898 |
+
"learning_rate": 4.108943518630287e-05,
|
| 899 |
+
"loss": 1.2636,
|
| 900 |
+
"step": 1250
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.057934508816121,
|
| 904 |
+
"grad_norm": 0.2663809997677264,
|
| 905 |
+
"learning_rate": 4.09016936977508e-05,
|
| 906 |
+
"loss": 1.246,
|
| 907 |
+
"step": 1260
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.0663308144416457,
|
| 911 |
+
"grad_norm": 0.2666405948242722,
|
| 912 |
+
"learning_rate": 4.071243383883344e-05,
|
| 913 |
+
"loss": 1.2416,
|
| 914 |
+
"step": 1270
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.0747271200671704,
|
| 918 |
+
"grad_norm": 0.27372912036143243,
|
| 919 |
+
"learning_rate": 4.052167368099443e-05,
|
| 920 |
+
"loss": 1.2684,
|
| 921 |
+
"step": 1280
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.0831234256926952,
|
| 925 |
+
"grad_norm": 0.2822619978126531,
|
| 926 |
+
"learning_rate": 4.032943143893315e-05,
|
| 927 |
+
"loss": 1.2679,
|
| 928 |
+
"step": 1290
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.0915197313182199,
|
| 932 |
+
"grad_norm": 0.2630554056666267,
|
| 933 |
+
"learning_rate": 4.0135725468865545e-05,
|
| 934 |
+
"loss": 1.2801,
|
| 935 |
+
"step": 1300
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.0999160369437448,
|
| 939 |
+
"grad_norm": 0.27334249585434633,
|
| 940 |
+
"learning_rate": 3.994057426677135e-05,
|
| 941 |
+
"loss": 1.2552,
|
| 942 |
+
"step": 1310
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.1083123425692696,
|
| 946 |
+
"grad_norm": 0.2751762295182287,
|
| 947 |
+
"learning_rate": 3.974399646662804e-05,
|
| 948 |
+
"loss": 1.2467,
|
| 949 |
+
"step": 1320
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.1167086481947943,
|
| 953 |
+
"grad_norm": 0.27811851573000057,
|
| 954 |
+
"learning_rate": 3.9546010838631523e-05,
|
| 955 |
+
"loss": 1.2393,
|
| 956 |
+
"step": 1330
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.125104953820319,
|
| 960 |
+
"grad_norm": 0.2786179703795066,
|
| 961 |
+
"learning_rate": 3.9346636287403936e-05,
|
| 962 |
+
"loss": 1.2413,
|
| 963 |
+
"step": 1340
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.1335012594458438,
|
| 967 |
+
"grad_norm": 0.2757598420534043,
|
| 968 |
+
"learning_rate": 3.9145891850188446e-05,
|
| 969 |
+
"loss": 1.2344,
|
| 970 |
+
"step": 1350
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.1418975650713685,
|
| 974 |
+
"grad_norm": 0.2826938023138817,
|
| 975 |
+
"learning_rate": 3.894379669503159e-05,
|
| 976 |
+
"loss": 1.2723,
|
| 977 |
+
"step": 1360
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.1502938706968933,
|
| 981 |
+
"grad_norm": 0.28640980146174666,
|
| 982 |
+
"learning_rate": 3.874037011895294e-05,
|
| 983 |
+
"loss": 1.2441,
|
| 984 |
+
"step": 1370
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.1586901763224182,
|
| 988 |
+
"grad_norm": 0.2806223555234711,
|
| 989 |
+
"learning_rate": 3.853563154610255e-05,
|
| 990 |
+
"loss": 1.2599,
|
| 991 |
+
"step": 1380
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.167086481947943,
|
| 995 |
+
"grad_norm": 0.280705015660275,
|
| 996 |
+
"learning_rate": 3.832960052590626e-05,
|
| 997 |
+
"loss": 1.2596,
|
| 998 |
+
"step": 1390
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.1754827875734677,
|
| 1002 |
+
"grad_norm": 0.2794281887739961,
|
| 1003 |
+
"learning_rate": 3.812229673119902e-05,
|
| 1004 |
+
"loss": 1.2679,
|
| 1005 |
+
"step": 1400
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.1838790931989924,
|
| 1009 |
+
"grad_norm": 0.29244474520092206,
|
| 1010 |
+
"learning_rate": 3.791373995634641e-05,
|
| 1011 |
+
"loss": 1.2893,
|
| 1012 |
+
"step": 1410
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.1922753988245172,
|
| 1016 |
+
"grad_norm": 0.27087843377656023,
|
| 1017 |
+
"learning_rate": 3.770395011535459e-05,
|
| 1018 |
+
"loss": 1.2858,
|
| 1019 |
+
"step": 1420
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.200671704450042,
|
| 1023 |
+
"grad_norm": 0.3084320815357459,
|
| 1024 |
+
"learning_rate": 3.749294723996884e-05,
|
| 1025 |
+
"loss": 1.2544,
|
| 1026 |
+
"step": 1430
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.2090680100755669,
|
| 1030 |
+
"grad_norm": 0.2970014625754613,
|
| 1031 |
+
"learning_rate": 3.72807514777608e-05,
|
| 1032 |
+
"loss": 1.2406,
|
| 1033 |
+
"step": 1440
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.2174643157010916,
|
| 1037 |
+
"grad_norm": 0.2858465781764202,
|
| 1038 |
+
"learning_rate": 3.7067383090204696e-05,
|
| 1039 |
+
"loss": 1.2687,
|
| 1040 |
+
"step": 1450
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.2258606213266163,
|
| 1044 |
+
"grad_norm": 0.29654341377963134,
|
| 1045 |
+
"learning_rate": 3.6852862450742664e-05,
|
| 1046 |
+
"loss": 1.2529,
|
| 1047 |
+
"step": 1460
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.234256926952141,
|
| 1051 |
+
"grad_norm": 0.2999209714333275,
|
| 1052 |
+
"learning_rate": 3.663721004283942e-05,
|
| 1053 |
+
"loss": 1.2607,
|
| 1054 |
+
"step": 1470
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.2426532325776658,
|
| 1058 |
+
"grad_norm": 0.32376267785640894,
|
| 1059 |
+
"learning_rate": 3.642044645802638e-05,
|
| 1060 |
+
"loss": 1.2633,
|
| 1061 |
+
"step": 1480
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.2510495382031905,
|
| 1065 |
+
"grad_norm": 0.2946271678215341,
|
| 1066 |
+
"learning_rate": 3.620259239393549e-05,
|
| 1067 |
+
"loss": 1.2631,
|
| 1068 |
+
"step": 1490
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.2594458438287153,
|
| 1072 |
+
"grad_norm": 0.29934101463933066,
|
| 1073 |
+
"learning_rate": 3.5983668652322914e-05,
|
| 1074 |
+
"loss": 1.2323,
|
| 1075 |
+
"step": 1500
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.26784214945424,
|
| 1079 |
+
"grad_norm": 0.3058737589423296,
|
| 1080 |
+
"learning_rate": 3.576369613708278e-05,
|
| 1081 |
+
"loss": 1.2499,
|
| 1082 |
+
"step": 1510
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.276238455079765,
|
| 1086 |
+
"grad_norm": 0.294073209044752,
|
| 1087 |
+
"learning_rate": 3.554269585225117e-05,
|
| 1088 |
+
"loss": 1.2623,
|
| 1089 |
+
"step": 1520
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.2846347607052897,
|
| 1093 |
+
"grad_norm": 0.3133601837566205,
|
| 1094 |
+
"learning_rate": 3.532068890000059e-05,
|
| 1095 |
+
"loss": 1.25,
|
| 1096 |
+
"step": 1530
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 1.2930310663308144,
|
| 1100 |
+
"grad_norm": 0.3030161218843826,
|
| 1101 |
+
"learning_rate": 3.5097696478624956e-05,
|
| 1102 |
+
"loss": 1.2615,
|
| 1103 |
+
"step": 1540
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.3014273719563392,
|
| 1107 |
+
"grad_norm": 0.2944854213261804,
|
| 1108 |
+
"learning_rate": 3.487373988051556e-05,
|
| 1109 |
+
"loss": 1.2577,
|
| 1110 |
+
"step": 1550
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.309823677581864,
|
| 1114 |
+
"grad_norm": 0.3198932031443526,
|
| 1115 |
+
"learning_rate": 3.464884049012794e-05,
|
| 1116 |
+
"loss": 1.2252,
|
| 1117 |
+
"step": 1560
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 1.3182199832073889,
|
| 1121 |
+
"grad_norm": 0.3075614601508588,
|
| 1122 |
+
"learning_rate": 3.442301978193996e-05,
|
| 1123 |
+
"loss": 1.2333,
|
| 1124 |
+
"step": 1570
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.3266162888329136,
|
| 1128 |
+
"grad_norm": 0.3025833538356648,
|
| 1129 |
+
"learning_rate": 3.419629931840137e-05,
|
| 1130 |
+
"loss": 1.2309,
|
| 1131 |
+
"step": 1580
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.3350125944584383,
|
| 1135 |
+
"grad_norm": 0.33238464076222285,
|
| 1136 |
+
"learning_rate": 3.396870074787489e-05,
|
| 1137 |
+
"loss": 1.2347,
|
| 1138 |
+
"step": 1590
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 1.343408900083963,
|
| 1142 |
+
"grad_norm": 0.33142866431526025,
|
| 1143 |
+
"learning_rate": 3.374024580256913e-05,
|
| 1144 |
+
"loss": 1.2278,
|
| 1145 |
+
"step": 1600
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.3518052057094878,
|
| 1149 |
+
"grad_norm": 0.3068832931071064,
|
| 1150 |
+
"learning_rate": 3.351095629646348e-05,
|
| 1151 |
+
"loss": 1.2459,
|
| 1152 |
+
"step": 1610
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 1.3602015113350125,
|
| 1156 |
+
"grad_norm": 0.32787598875401036,
|
| 1157 |
+
"learning_rate": 3.3280854123225245e-05,
|
| 1158 |
+
"loss": 1.2567,
|
| 1159 |
+
"step": 1620
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 1.3685978169605373,
|
| 1163 |
+
"grad_norm": 0.334986160782939,
|
| 1164 |
+
"learning_rate": 3.3049961254119077e-05,
|
| 1165 |
+
"loss": 1.2427,
|
| 1166 |
+
"step": 1630
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.376994122586062,
|
| 1170 |
+
"grad_norm": 0.3257919179674348,
|
| 1171 |
+
"learning_rate": 3.281829973590909e-05,
|
| 1172 |
+
"loss": 1.2292,
|
| 1173 |
+
"step": 1640
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 1.385390428211587,
|
| 1177 |
+
"grad_norm": 0.33488335118724716,
|
| 1178 |
+
"learning_rate": 3.258589168875373e-05,
|
| 1179 |
+
"loss": 1.2231,
|
| 1180 |
+
"step": 1650
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.3937867338371117,
|
| 1184 |
+
"grad_norm": 0.29956276796558484,
|
| 1185 |
+
"learning_rate": 3.2352759304093624e-05,
|
| 1186 |
+
"loss": 1.2399,
|
| 1187 |
+
"step": 1660
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 1.4021830394626364,
|
| 1191 |
+
"grad_norm": 0.32793658183091984,
|
| 1192 |
+
"learning_rate": 3.211892484253261e-05,
|
| 1193 |
+
"loss": 1.2786,
|
| 1194 |
+
"step": 1670
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.4105793450881612,
|
| 1198 |
+
"grad_norm": 0.3437947332874714,
|
| 1199 |
+
"learning_rate": 3.1884410631712235e-05,
|
| 1200 |
+
"loss": 1.2423,
|
| 1201 |
+
"step": 1680
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 1.418975650713686,
|
| 1205 |
+
"grad_norm": 0.32417782627379454,
|
| 1206 |
+
"learning_rate": 3.164923906417979e-05,
|
| 1207 |
+
"loss": 1.2399,
|
| 1208 |
+
"step": 1690
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.4273719563392109,
|
| 1212 |
+
"grad_norm": 0.32303675587973024,
|
| 1213 |
+
"learning_rate": 3.1413432595250134e-05,
|
| 1214 |
+
"loss": 1.2575,
|
| 1215 |
+
"step": 1700
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.4357682619647356,
|
| 1219 |
+
"grad_norm": 0.3107110639387478,
|
| 1220 |
+
"learning_rate": 3.117701374086158e-05,
|
| 1221 |
+
"loss": 1.2155,
|
| 1222 |
+
"step": 1710
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 1.4441645675902604,
|
| 1226 |
+
"grad_norm": 0.31980715839644586,
|
| 1227 |
+
"learning_rate": 3.094000507542595e-05,
|
| 1228 |
+
"loss": 1.2561,
|
| 1229 |
+
"step": 1720
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 1.452560873215785,
|
| 1233 |
+
"grad_norm": 0.3452092427446392,
|
| 1234 |
+
"learning_rate": 3.07024292296731e-05,
|
| 1235 |
+
"loss": 1.2389,
|
| 1236 |
+
"step": 1730
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 1.4609571788413098,
|
| 1240 |
+
"grad_norm": 0.34168169619885197,
|
| 1241 |
+
"learning_rate": 3.0464308888489936e-05,
|
| 1242 |
+
"loss": 1.2323,
|
| 1243 |
+
"step": 1740
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 1.4693534844668346,
|
| 1247 |
+
"grad_norm": 0.33064208693690544,
|
| 1248 |
+
"learning_rate": 3.022566678875446e-05,
|
| 1249 |
+
"loss": 1.228,
|
| 1250 |
+
"step": 1750
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.4777497900923593,
|
| 1254 |
+
"grad_norm": 0.3362430973154644,
|
| 1255 |
+
"learning_rate": 2.9986525717164694e-05,
|
| 1256 |
+
"loss": 1.2025,
|
| 1257 |
+
"step": 1760
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1.486146095717884,
|
| 1261 |
+
"grad_norm": 0.34796677921939057,
|
| 1262 |
+
"learning_rate": 2.974690850806287e-05,
|
| 1263 |
+
"loss": 1.2585,
|
| 1264 |
+
"step": 1770
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 1.4945424013434088,
|
| 1268 |
+
"grad_norm": 0.31369265907711963,
|
| 1269 |
+
"learning_rate": 2.9506838041255147e-05,
|
| 1270 |
+
"loss": 1.2438,
|
| 1271 |
+
"step": 1780
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.5029387069689337,
|
| 1275 |
+
"grad_norm": 0.35233288464429297,
|
| 1276 |
+
"learning_rate": 2.926633723982692e-05,
|
| 1277 |
+
"loss": 1.23,
|
| 1278 |
+
"step": 1790
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 1.5113350125944585,
|
| 1282 |
+
"grad_norm": 0.33824484157682727,
|
| 1283 |
+
"learning_rate": 2.9025429067953975e-05,
|
| 1284 |
+
"loss": 1.2089,
|
| 1285 |
+
"step": 1800
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1.5113350125944585,
|
| 1289 |
+
"eval_loss": 1.3161499500274658,
|
| 1290 |
+
"eval_runtime": 37.5007,
|
| 1291 |
+
"eval_samples_per_second": 10.266,
|
| 1292 |
+
"eval_steps_per_second": 2.587,
|
| 1293 |
+
"step": 1800
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 1.5197313182199832,
|
| 1297 |
+
"grad_norm": 0.3450077958737747,
|
| 1298 |
+
"learning_rate": 2.878413652870983e-05,
|
| 1299 |
+
"loss": 1.2452,
|
| 1300 |
+
"step": 1810
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 1.528127623845508,
|
| 1304 |
+
"grad_norm": 0.3401369524359448,
|
| 1305 |
+
"learning_rate": 2.8542482661869228e-05,
|
| 1306 |
+
"loss": 1.2335,
|
| 1307 |
+
"step": 1820
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 1.536523929471033,
|
| 1311 |
+
"grad_norm": 0.35772471665666244,
|
| 1312 |
+
"learning_rate": 2.8300490541708226e-05,
|
| 1313 |
+
"loss": 1.2107,
|
| 1314 |
+
"step": 1830
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 1.5449202350965576,
|
| 1318 |
+
"grad_norm": 0.36182719622922654,
|
| 1319 |
+
"learning_rate": 2.8058183274800933e-05,
|
| 1320 |
+
"loss": 1.2323,
|
| 1321 |
+
"step": 1840
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 1.5533165407220824,
|
| 1325 |
+
"grad_norm": 0.3427498080631671,
|
| 1326 |
+
"learning_rate": 2.7815583997813176e-05,
|
| 1327 |
+
"loss": 1.2208,
|
| 1328 |
+
"step": 1850
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 1.561712846347607,
|
| 1332 |
+
"grad_norm": 0.3416312895125341,
|
| 1333 |
+
"learning_rate": 2.7572715875293336e-05,
|
| 1334 |
+
"loss": 1.2519,
|
| 1335 |
+
"step": 1860
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 1.5701091519731318,
|
| 1339 |
+
"grad_norm": 0.34134497126888136,
|
| 1340 |
+
"learning_rate": 2.7329602097460453e-05,
|
| 1341 |
+
"loss": 1.2072,
|
| 1342 |
+
"step": 1870
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 1.5785054575986566,
|
| 1346 |
+
"grad_norm": 0.3414019695200471,
|
| 1347 |
+
"learning_rate": 2.7086265877989907e-05,
|
| 1348 |
+
"loss": 1.2171,
|
| 1349 |
+
"step": 1880
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 1.5869017632241813,
|
| 1353 |
+
"grad_norm": 0.3493625784848411,
|
| 1354 |
+
"learning_rate": 2.68427304517969e-05,
|
| 1355 |
+
"loss": 1.2032,
|
| 1356 |
+
"step": 1890
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 1.595298068849706,
|
| 1360 |
+
"grad_norm": 0.3432387874546048,
|
| 1361 |
+
"learning_rate": 2.6599019072817823e-05,
|
| 1362 |
+
"loss": 1.2233,
|
| 1363 |
+
"step": 1900
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 1.6036943744752308,
|
| 1367 |
+
"grad_norm": 0.34931851248995316,
|
| 1368 |
+
"learning_rate": 2.6355155011789918e-05,
|
| 1369 |
+
"loss": 1.2302,
|
| 1370 |
+
"step": 1910
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 1.6120906801007555,
|
| 1374 |
+
"grad_norm": 0.3423165132865183,
|
| 1375 |
+
"learning_rate": 2.6111161554029216e-05,
|
| 1376 |
+
"loss": 1.2237,
|
| 1377 |
+
"step": 1920
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 1.6204869857262805,
|
| 1381 |
+
"grad_norm": 0.3286103317410491,
|
| 1382 |
+
"learning_rate": 2.5867061997207208e-05,
|
| 1383 |
+
"loss": 1.2373,
|
| 1384 |
+
"step": 1930
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 1.6288832913518052,
|
| 1388 |
+
"grad_norm": 0.34439194836734,
|
| 1389 |
+
"learning_rate": 2.5622879649126204e-05,
|
| 1390 |
+
"loss": 1.2366,
|
| 1391 |
+
"step": 1940
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 1.63727959697733,
|
| 1395 |
+
"grad_norm": 0.37147052522944746,
|
| 1396 |
+
"learning_rate": 2.5378637825493827e-05,
|
| 1397 |
+
"loss": 1.2175,
|
| 1398 |
+
"step": 1950
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 1.645675902602855,
|
| 1402 |
+
"grad_norm": 0.3720397203000226,
|
| 1403 |
+
"learning_rate": 2.5134359847696725e-05,
|
| 1404 |
+
"loss": 1.2226,
|
| 1405 |
+
"step": 1960
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 1.6540722082283796,
|
| 1409 |
+
"grad_norm": 0.372249827112124,
|
| 1410 |
+
"learning_rate": 2.48900690405737e-05,
|
| 1411 |
+
"loss": 1.2354,
|
| 1412 |
+
"step": 1970
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 1.6624685138539044,
|
| 1416 |
+
"grad_norm": 0.32759893432286225,
|
| 1417 |
+
"learning_rate": 2.4645788730188595e-05,
|
| 1418 |
+
"loss": 1.2346,
|
| 1419 |
+
"step": 1980
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 1.670864819479429,
|
| 1423 |
+
"grad_norm": 0.36258855898143105,
|
| 1424 |
+
"learning_rate": 2.4401542241602936e-05,
|
| 1425 |
+
"loss": 1.2266,
|
| 1426 |
+
"step": 1990
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 1.6792611251049538,
|
| 1430 |
+
"grad_norm": 0.35828351901418637,
|
| 1431 |
+
"learning_rate": 2.4157352896648785e-05,
|
| 1432 |
+
"loss": 1.2489,
|
| 1433 |
+
"step": 2000
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 1.6876574307304786,
|
| 1437 |
+
"grad_norm": 0.3547226628271256,
|
| 1438 |
+
"learning_rate": 2.391324401170189e-05,
|
| 1439 |
+
"loss": 1.2321,
|
| 1440 |
+
"step": 2010
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 1.6960537363560033,
|
| 1444 |
+
"grad_norm": 0.37844822229404196,
|
| 1445 |
+
"learning_rate": 2.3669238895455257e-05,
|
| 1446 |
+
"loss": 1.2133,
|
| 1447 |
+
"step": 2020
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 1.704450041981528,
|
| 1451 |
+
"grad_norm": 0.36535971407503814,
|
| 1452 |
+
"learning_rate": 2.3425360846693546e-05,
|
| 1453 |
+
"loss": 1.1832,
|
| 1454 |
+
"step": 2030
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 1.7128463476070528,
|
| 1458 |
+
"grad_norm": 0.351966888983441,
|
| 1459 |
+
"learning_rate": 2.3181633152068444e-05,
|
| 1460 |
+
"loss": 1.1978,
|
| 1461 |
+
"step": 2040
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 1.7212426532325775,
|
| 1465 |
+
"grad_norm": 0.35857960344757944,
|
| 1466 |
+
"learning_rate": 2.293807908387507e-05,
|
| 1467 |
+
"loss": 1.2274,
|
| 1468 |
+
"step": 2050
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 1.7296389588581025,
|
| 1472 |
+
"grad_norm": 0.36067730619373684,
|
| 1473 |
+
"learning_rate": 2.269472189782984e-05,
|
| 1474 |
+
"loss": 1.2107,
|
| 1475 |
+
"step": 2060
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 1.7380352644836272,
|
| 1479 |
+
"grad_norm": 0.3686118668283621,
|
| 1480 |
+
"learning_rate": 2.2451584830849963e-05,
|
| 1481 |
+
"loss": 1.2453,
|
| 1482 |
+
"step": 2070
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 1.746431570109152,
|
| 1486 |
+
"grad_norm": 0.3574767970991608,
|
| 1487 |
+
"learning_rate": 2.2208691098834574e-05,
|
| 1488 |
+
"loss": 1.2112,
|
| 1489 |
+
"step": 2080
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 1.7548278757346767,
|
| 1493 |
+
"grad_norm": 0.3649596547435675,
|
| 1494 |
+
"learning_rate": 2.196606389444802e-05,
|
| 1495 |
+
"loss": 1.2301,
|
| 1496 |
+
"step": 2090
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 1.7632241813602016,
|
| 1500 |
+
"grad_norm": 0.35608820401696456,
|
| 1501 |
+
"learning_rate": 2.1723726384905332e-05,
|
| 1502 |
+
"loss": 1.2254,
|
| 1503 |
+
"step": 2100
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 1.7716204869857264,
|
| 1507 |
+
"grad_norm": 0.3903682280760152,
|
| 1508 |
+
"learning_rate": 2.1481701709760054e-05,
|
| 1509 |
+
"loss": 1.2191,
|
| 1510 |
+
"step": 2110
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 1.7800167926112511,
|
| 1514 |
+
"grad_norm": 0.3700744583799887,
|
| 1515 |
+
"learning_rate": 2.1240012978694786e-05,
|
| 1516 |
+
"loss": 1.1897,
|
| 1517 |
+
"step": 2120
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 1.7884130982367759,
|
| 1521 |
+
"grad_norm": 0.3678524367281197,
|
| 1522 |
+
"learning_rate": 2.0998683269314613e-05,
|
| 1523 |
+
"loss": 1.1936,
|
| 1524 |
+
"step": 2130
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 1.7968094038623006,
|
| 1528 |
+
"grad_norm": 0.39001676927124185,
|
| 1529 |
+
"learning_rate": 2.0757735624943448e-05,
|
| 1530 |
+
"loss": 1.2248,
|
| 1531 |
+
"step": 2140
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 1.8052057094878253,
|
| 1535 |
+
"grad_norm": 0.39780376699395825,
|
| 1536 |
+
"learning_rate": 2.0517193052423823e-05,
|
| 1537 |
+
"loss": 1.2022,
|
| 1538 |
+
"step": 2150
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 1.81360201511335,
|
| 1542 |
+
"grad_norm": 0.380325076212259,
|
| 1543 |
+
"learning_rate": 2.0277078519920044e-05,
|
| 1544 |
+
"loss": 1.1938,
|
| 1545 |
+
"step": 2160
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 1.8219983207388748,
|
| 1549 |
+
"grad_norm": 0.3871471596296449,
|
| 1550 |
+
"learning_rate": 2.0037414954725117e-05,
|
| 1551 |
+
"loss": 1.1825,
|
| 1552 |
+
"step": 2170
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 1.8303946263643995,
|
| 1556 |
+
"grad_norm": 0.37964701587107547,
|
| 1557 |
+
"learning_rate": 1.979822524107148e-05,
|
| 1558 |
+
"loss": 1.2092,
|
| 1559 |
+
"step": 2180
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 1.8387909319899243,
|
| 1563 |
+
"grad_norm": 0.36807392164929836,
|
| 1564 |
+
"learning_rate": 1.9559532217945974e-05,
|
| 1565 |
+
"loss": 1.2378,
|
| 1566 |
+
"step": 2190
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 1.8471872376154492,
|
| 1570 |
+
"grad_norm": 0.3558532380492764,
|
| 1571 |
+
"learning_rate": 1.932135867690901e-05,
|
| 1572 |
+
"loss": 1.2142,
|
| 1573 |
+
"step": 2200
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 1.855583543240974,
|
| 1577 |
+
"grad_norm": 0.3795291868557196,
|
| 1578 |
+
"learning_rate": 1.908372735991836e-05,
|
| 1579 |
+
"loss": 1.2003,
|
| 1580 |
+
"step": 2210
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 1.8639798488664987,
|
| 1584 |
+
"grad_norm": 0.3698629794511824,
|
| 1585 |
+
"learning_rate": 1.8846660957157626e-05,
|
| 1586 |
+
"loss": 1.2279,
|
| 1587 |
+
"step": 2220
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 1.8723761544920237,
|
| 1591 |
+
"grad_norm": 0.37197609159817596,
|
| 1592 |
+
"learning_rate": 1.8610182104869676e-05,
|
| 1593 |
+
"loss": 1.2175,
|
| 1594 |
+
"step": 2230
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 1.8807724601175484,
|
| 1598 |
+
"grad_norm": 0.39691535067925626,
|
| 1599 |
+
"learning_rate": 1.837431338319523e-05,
|
| 1600 |
+
"loss": 1.1965,
|
| 1601 |
+
"step": 2240
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 1.8891687657430731,
|
| 1605 |
+
"grad_norm": 0.39830491258185996,
|
| 1606 |
+
"learning_rate": 1.813907731401677e-05,
|
| 1607 |
+
"loss": 1.2082,
|
| 1608 |
+
"step": 2250
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 1.8975650713685979,
|
| 1612 |
+
"grad_norm": 0.40042636908106866,
|
| 1613 |
+
"learning_rate": 1.790449635880813e-05,
|
| 1614 |
+
"loss": 1.1969,
|
| 1615 |
+
"step": 2260
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 1.9059613769941226,
|
| 1619 |
+
"grad_norm": 0.3631756871613192,
|
| 1620 |
+
"learning_rate": 1.7670592916489637e-05,
|
| 1621 |
+
"loss": 1.2094,
|
| 1622 |
+
"step": 2270
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 1.9143576826196473,
|
| 1626 |
+
"grad_norm": 0.41767311488963316,
|
| 1627 |
+
"learning_rate": 1.743738932128945e-05,
|
| 1628 |
+
"loss": 1.2171,
|
| 1629 |
+
"step": 2280
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 1.922753988245172,
|
| 1633 |
+
"grad_norm": 0.42201095235456104,
|
| 1634 |
+
"learning_rate": 1.7204907840610968e-05,
|
| 1635 |
+
"loss": 1.1902,
|
| 1636 |
+
"step": 2290
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 1.9311502938706968,
|
| 1640 |
+
"grad_norm": 0.38604694393746347,
|
| 1641 |
+
"learning_rate": 1.6973170672906592e-05,
|
| 1642 |
+
"loss": 1.1915,
|
| 1643 |
+
"step": 2300
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 1.9395465994962215,
|
| 1647 |
+
"grad_norm": 0.3699669985546787,
|
| 1648 |
+
"learning_rate": 1.6742199945558116e-05,
|
| 1649 |
+
"loss": 1.2018,
|
| 1650 |
+
"step": 2310
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 1.9479429051217463,
|
| 1654 |
+
"grad_norm": 0.3862222953863642,
|
| 1655 |
+
"learning_rate": 1.651201771276397e-05,
|
| 1656 |
+
"loss": 1.1742,
|
| 1657 |
+
"step": 2320
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 1.9563392107472712,
|
| 1661 |
+
"grad_norm": 0.4294759850725183,
|
| 1662 |
+
"learning_rate": 1.6282645953433278e-05,
|
| 1663 |
+
"loss": 1.2024,
|
| 1664 |
+
"step": 2330
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 1.964735516372796,
|
| 1668 |
+
"grad_norm": 0.4231278271726052,
|
| 1669 |
+
"learning_rate": 1.6054106569087247e-05,
|
| 1670 |
+
"loss": 1.2072,
|
| 1671 |
+
"step": 2340
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 1.9731318219983207,
|
| 1675 |
+
"grad_norm": 0.40525859849828033,
|
| 1676 |
+
"learning_rate": 1.5826421381767943e-05,
|
| 1677 |
+
"loss": 1.1996,
|
| 1678 |
+
"step": 2350
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 1.9815281276238457,
|
| 1682 |
+
"grad_norm": 0.41477520137980894,
|
| 1683 |
+
"learning_rate": 1.5599612131954562e-05,
|
| 1684 |
+
"loss": 1.1963,
|
| 1685 |
+
"step": 2360
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 1.9899244332493704,
|
| 1689 |
+
"grad_norm": 0.4201296814849193,
|
| 1690 |
+
"learning_rate": 1.537370047648755e-05,
|
| 1691 |
+
"loss": 1.1909,
|
| 1692 |
+
"step": 2370
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 1.9983207388748951,
|
| 1696 |
+
"grad_norm": 0.39991078592707974,
|
| 1697 |
+
"learning_rate": 1.5148707986500754e-05,
|
| 1698 |
+
"loss": 1.1934,
|
| 1699 |
+
"step": 2380
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 2.00671704450042,
|
| 1703 |
+
"grad_norm": 0.542630060134636,
|
| 1704 |
+
"learning_rate": 1.4924656145361642e-05,
|
| 1705 |
+
"loss": 1.1015,
|
| 1706 |
+
"step": 2390
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 2.0151133501259446,
|
| 1710 |
+
"grad_norm": 0.44916796266698206,
|
| 1711 |
+
"learning_rate": 1.4701566346619994e-05,
|
| 1712 |
+
"loss": 1.0911,
|
| 1713 |
+
"step": 2400
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 2.0151133501259446,
|
| 1717 |
+
"eval_loss": 1.288699746131897,
|
| 1718 |
+
"eval_runtime": 37.7556,
|
| 1719 |
+
"eval_samples_per_second": 10.197,
|
| 1720 |
+
"eval_steps_per_second": 2.569,
|
| 1721 |
+
"step": 2400
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 2.0235096557514693,
|
| 1725 |
+
"grad_norm": 0.44991891228670405,
|
| 1726 |
+
"learning_rate": 1.447945989196518e-05,
|
| 1727 |
+
"loss": 1.0922,
|
| 1728 |
+
"step": 2410
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 2.031905961376994,
|
| 1732 |
+
"grad_norm": 0.4773364949208615,
|
| 1733 |
+
"learning_rate": 1.4258357989192089e-05,
|
| 1734 |
+
"loss": 1.0882,
|
| 1735 |
+
"step": 2420
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 2.040302267002519,
|
| 1739 |
+
"grad_norm": 0.47709647551922374,
|
| 1740 |
+
"learning_rate": 1.403828175017618e-05,
|
| 1741 |
+
"loss": 1.0976,
|
| 1742 |
+
"step": 2430
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 2.0486985726280436,
|
| 1746 |
+
"grad_norm": 0.5391514428897679,
|
| 1747 |
+
"learning_rate": 1.381925218885759e-05,
|
| 1748 |
+
"loss": 1.0999,
|
| 1749 |
+
"step": 2440
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 2.0570948782535683,
|
| 1753 |
+
"grad_norm": 0.482275832558956,
|
| 1754 |
+
"learning_rate": 1.3601290219234642e-05,
|
| 1755 |
+
"loss": 1.0848,
|
| 1756 |
+
"step": 2450
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 2.065491183879093,
|
| 1760 |
+
"grad_norm": 0.4879837972652172,
|
| 1761 |
+
"learning_rate": 1.3384416653366796e-05,
|
| 1762 |
+
"loss": 1.1053,
|
| 1763 |
+
"step": 2460
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 2.0738874895046178,
|
| 1767 |
+
"grad_norm": 0.5108185234761726,
|
| 1768 |
+
"learning_rate": 1.3168652199387565e-05,
|
| 1769 |
+
"loss": 1.0565,
|
| 1770 |
+
"step": 2470
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 2.082283795130143,
|
| 1774 |
+
"grad_norm": 0.47147504315736866,
|
| 1775 |
+
"learning_rate": 1.2954017459527037e-05,
|
| 1776 |
+
"loss": 1.1135,
|
| 1777 |
+
"step": 2480
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 2.0906801007556677,
|
| 1781 |
+
"grad_norm": 0.4683105843680792,
|
| 1782 |
+
"learning_rate": 1.2740532928144785e-05,
|
| 1783 |
+
"loss": 1.0978,
|
| 1784 |
+
"step": 2490
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 2.0990764063811924,
|
| 1788 |
+
"grad_norm": 0.46420490471813974,
|
| 1789 |
+
"learning_rate": 1.2528218989772928e-05,
|
| 1790 |
+
"loss": 1.0924,
|
| 1791 |
+
"step": 2500
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 2.107472712006717,
|
| 1795 |
+
"grad_norm": 0.5116237328178771,
|
| 1796 |
+
"learning_rate": 1.2317095917169729e-05,
|
| 1797 |
+
"loss": 1.1122,
|
| 1798 |
+
"step": 2510
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 2.115869017632242,
|
| 1802 |
+
"grad_norm": 0.48364433468846557,
|
| 1803 |
+
"learning_rate": 1.2107183869383817e-05,
|
| 1804 |
+
"loss": 1.0654,
|
| 1805 |
+
"step": 2520
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 2.1242653232577666,
|
| 1809 |
+
"grad_norm": 0.47577790302982037,
|
| 1810 |
+
"learning_rate": 1.1898502889829367e-05,
|
| 1811 |
+
"loss": 1.0885,
|
| 1812 |
+
"step": 2530
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 2.1326616288832914,
|
| 1816 |
+
"grad_norm": 0.5105053373884131,
|
| 1817 |
+
"learning_rate": 1.1691072904372211e-05,
|
| 1818 |
+
"loss": 1.0814,
|
| 1819 |
+
"step": 2540
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 2.141057934508816,
|
| 1823 |
+
"grad_norm": 0.45277641048792205,
|
| 1824 |
+
"learning_rate": 1.1484913719427245e-05,
|
| 1825 |
+
"loss": 1.1074,
|
| 1826 |
+
"step": 2550
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 2.149454240134341,
|
| 1830 |
+
"grad_norm": 0.4721310040821966,
|
| 1831 |
+
"learning_rate": 1.1280045020067173e-05,
|
| 1832 |
+
"loss": 1.0624,
|
| 1833 |
+
"step": 2560
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 2.1578505457598656,
|
| 1837 |
+
"grad_norm": 0.5264060503334844,
|
| 1838 |
+
"learning_rate": 1.1076486368142974e-05,
|
| 1839 |
+
"loss": 1.075,
|
| 1840 |
+
"step": 2570
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 2.1662468513853903,
|
| 1844 |
+
"grad_norm": 0.49378445456124825,
|
| 1845 |
+
"learning_rate": 1.0874257200415921e-05,
|
| 1846 |
+
"loss": 1.1239,
|
| 1847 |
+
"step": 2580
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 2.174643157010915,
|
| 1851 |
+
"grad_norm": 0.4696082013605863,
|
| 1852 |
+
"learning_rate": 1.0673376826701764e-05,
|
| 1853 |
+
"loss": 1.0855,
|
| 1854 |
+
"step": 2590
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 2.1830394626364398,
|
| 1858 |
+
"grad_norm": 0.520016314444527,
|
| 1859 |
+
"learning_rate": 1.0473864428026903e-05,
|
| 1860 |
+
"loss": 1.0613,
|
| 1861 |
+
"step": 2600
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 2.1914357682619645,
|
| 1865 |
+
"grad_norm": 0.5085845086562594,
|
| 1866 |
+
"learning_rate": 1.0275739054796849e-05,
|
| 1867 |
+
"loss": 1.0874,
|
| 1868 |
+
"step": 2610
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 2.1998320738874897,
|
| 1872 |
+
"grad_norm": 0.5229182760518745,
|
| 1873 |
+
"learning_rate": 1.0079019624977277e-05,
|
| 1874 |
+
"loss": 1.093,
|
| 1875 |
+
"step": 2620
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 2.2082283795130144,
|
| 1879 |
+
"grad_norm": 0.4847844823377718,
|
| 1880 |
+
"learning_rate": 9.883724922287593e-06,
|
| 1881 |
+
"loss": 1.0792,
|
| 1882 |
+
"step": 2630
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 2.216624685138539,
|
| 1886 |
+
"grad_norm": 0.4731502709572998,
|
| 1887 |
+
"learning_rate": 9.689873594407398e-06,
|
| 1888 |
+
"loss": 1.0513,
|
| 1889 |
+
"step": 2640
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 2.225020990764064,
|
| 1893 |
+
"grad_norm": 0.5138138380876534,
|
| 1894 |
+
"learning_rate": 9.497484151195874e-06,
|
| 1895 |
+
"loss": 1.0663,
|
| 1896 |
+
"step": 2650
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 2.2334172963895886,
|
| 1900 |
+
"grad_norm": 0.5026329202465635,
|
| 1901 |
+
"learning_rate": 9.30657496292447e-06,
|
| 1902 |
+
"loss": 1.0787,
|
| 1903 |
+
"step": 2660
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 2.2418136020151134,
|
| 1907 |
+
"grad_norm": 0.44925382614215525,
|
| 1908 |
+
"learning_rate": 9.117164258522695e-06,
|
| 1909 |
+
"loss": 1.0579,
|
| 1910 |
+
"step": 2670
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 2.250209907640638,
|
| 1914 |
+
"grad_norm": 0.5293508173928859,
|
| 1915 |
+
"learning_rate": 8.929270123837632e-06,
|
| 1916 |
+
"loss": 1.0512,
|
| 1917 |
+
"step": 2680
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 2.258606213266163,
|
| 1921 |
+
"grad_norm": 0.5247265106521677,
|
| 1922 |
+
"learning_rate": 8.742910499906973e-06,
|
| 1923 |
+
"loss": 1.0658,
|
| 1924 |
+
"step": 2690
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 2.2670025188916876,
|
| 1928 |
+
"grad_norm": 0.5190390778509362,
|
| 1929 |
+
"learning_rate": 8.558103181245921e-06,
|
| 1930 |
+
"loss": 1.0696,
|
| 1931 |
+
"step": 2700
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 2.2753988245172123,
|
| 1935 |
+
"grad_norm": 0.55132785975463,
|
| 1936 |
+
"learning_rate": 8.374865814148073e-06,
|
| 1937 |
+
"loss": 1.092,
|
| 1938 |
+
"step": 2710
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 2.283795130142737,
|
| 1942 |
+
"grad_norm": 0.5213613994205901,
|
| 1943 |
+
"learning_rate": 8.193215895000526e-06,
|
| 1944 |
+
"loss": 1.0822,
|
| 1945 |
+
"step": 2720
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 2.292191435768262,
|
| 1949 |
+
"grad_norm": 0.4796879456796719,
|
| 1950 |
+
"learning_rate": 8.013170768613146e-06,
|
| 1951 |
+
"loss": 1.0568,
|
| 1952 |
+
"step": 2730
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 2.3005877413937865,
|
| 1956 |
+
"grad_norm": 0.5385546333675868,
|
| 1957 |
+
"learning_rate": 7.834747626562484e-06,
|
| 1958 |
+
"loss": 1.0731,
|
| 1959 |
+
"step": 2740
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 2.3089840470193117,
|
| 1963 |
+
"grad_norm": 0.5377987038023434,
|
| 1964 |
+
"learning_rate": 7.657963505550189e-06,
|
| 1965 |
+
"loss": 1.075,
|
| 1966 |
+
"step": 2750
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 2.3173803526448364,
|
| 1970 |
+
"grad_norm": 0.562428992862054,
|
| 1971 |
+
"learning_rate": 7.482835285776305e-06,
|
| 1972 |
+
"loss": 1.0854,
|
| 1973 |
+
"step": 2760
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 2.325776658270361,
|
| 1977 |
+
"grad_norm": 0.4988726750554791,
|
| 1978 |
+
"learning_rate": 7.309379689327409e-06,
|
| 1979 |
+
"loss": 1.093,
|
| 1980 |
+
"step": 2770
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 2.334172963895886,
|
| 1984 |
+
"grad_norm": 0.5433657502932824,
|
| 1985 |
+
"learning_rate": 7.137613278579964e-06,
|
| 1986 |
+
"loss": 1.0926,
|
| 1987 |
+
"step": 2780
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 2.3425692695214106,
|
| 1991 |
+
"grad_norm": 0.5087897129635799,
|
| 1992 |
+
"learning_rate": 6.967552454618839e-06,
|
| 1993 |
+
"loss": 1.1078,
|
| 1994 |
+
"step": 2790
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 2.3509655751469354,
|
| 1998 |
+
"grad_norm": 0.5170416004589418,
|
| 1999 |
+
"learning_rate": 6.799213455671255e-06,
|
| 2000 |
+
"loss": 1.0781,
|
| 2001 |
+
"step": 2800
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 2.35936188077246,
|
| 2005 |
+
"grad_norm": 0.5063141748122139,
|
| 2006 |
+
"learning_rate": 6.632612355556256e-06,
|
| 2007 |
+
"loss": 1.0824,
|
| 2008 |
+
"step": 2810
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 2.367758186397985,
|
| 2012 |
+
"grad_norm": 0.5169250792091412,
|
| 2013 |
+
"learning_rate": 6.467765062149977e-06,
|
| 2014 |
+
"loss": 1.0686,
|
| 2015 |
+
"step": 2820
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 2.3761544920235096,
|
| 2019 |
+
"grad_norm": 0.47167776818464224,
|
| 2020 |
+
"learning_rate": 6.304687315866589e-06,
|
| 2021 |
+
"loss": 1.0902,
|
| 2022 |
+
"step": 2830
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 2.3845507976490343,
|
| 2026 |
+
"grad_norm": 0.5230768936155751,
|
| 2027 |
+
"learning_rate": 6.143394688155396e-06,
|
| 2028 |
+
"loss": 1.0704,
|
| 2029 |
+
"step": 2840
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 2.392947103274559,
|
| 2033 |
+
"grad_norm": 0.5113178637899721,
|
| 2034 |
+
"learning_rate": 5.98390258001397e-06,
|
| 2035 |
+
"loss": 1.0824,
|
| 2036 |
+
"step": 2850
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 2.401343408900084,
|
| 2040 |
+
"grad_norm": 0.5008499019390629,
|
| 2041 |
+
"learning_rate": 5.826226220517606e-06,
|
| 2042 |
+
"loss": 1.0549,
|
| 2043 |
+
"step": 2860
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 2.4097397145256085,
|
| 2047 |
+
"grad_norm": 0.5530580136554508,
|
| 2048 |
+
"learning_rate": 5.670380665365144e-06,
|
| 2049 |
+
"loss": 1.0999,
|
| 2050 |
+
"step": 2870
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 2.4181360201511337,
|
| 2054 |
+
"grad_norm": 0.48265348835919253,
|
| 2055 |
+
"learning_rate": 5.516380795441417e-06,
|
| 2056 |
+
"loss": 1.0929,
|
| 2057 |
+
"step": 2880
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 2.4265323257766584,
|
| 2061 |
+
"grad_norm": 0.5159406165392976,
|
| 2062 |
+
"learning_rate": 5.364241315396335e-06,
|
| 2063 |
+
"loss": 1.0857,
|
| 2064 |
+
"step": 2890
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 2.434928631402183,
|
| 2068 |
+
"grad_norm": 0.5456400716277449,
|
| 2069 |
+
"learning_rate": 5.21397675224081e-06,
|
| 2070 |
+
"loss": 1.0763,
|
| 2071 |
+
"step": 2900
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 2.443324937027708,
|
| 2075 |
+
"grad_norm": 0.5275724452771665,
|
| 2076 |
+
"learning_rate": 5.065601453959659e-06,
|
| 2077 |
+
"loss": 1.0837,
|
| 2078 |
+
"step": 2910
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 2.4517212426532327,
|
| 2082 |
+
"grad_norm": 0.5054255164716437,
|
| 2083 |
+
"learning_rate": 4.919129588141563e-06,
|
| 2084 |
+
"loss": 1.0655,
|
| 2085 |
+
"step": 2920
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 2.4601175482787574,
|
| 2089 |
+
"grad_norm": 0.5226226880297187,
|
| 2090 |
+
"learning_rate": 4.7745751406263165e-06,
|
| 2091 |
+
"loss": 1.0604,
|
| 2092 |
+
"step": 2930
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 2.468513853904282,
|
| 2096 |
+
"grad_norm": 0.4917792079813032,
|
| 2097 |
+
"learning_rate": 4.631951914169363e-06,
|
| 2098 |
+
"loss": 1.0635,
|
| 2099 |
+
"step": 2940
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 2.476910159529807,
|
| 2103 |
+
"grad_norm": 0.5656339286373594,
|
| 2104 |
+
"learning_rate": 4.491273527123866e-06,
|
| 2105 |
+
"loss": 1.0838,
|
| 2106 |
+
"step": 2950
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 2.4853064651553316,
|
| 2110 |
+
"grad_norm": 0.5253948788481366,
|
| 2111 |
+
"learning_rate": 4.352553412140303e-06,
|
| 2112 |
+
"loss": 1.06,
|
| 2113 |
+
"step": 2960
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 2.4937027707808563,
|
| 2117 |
+
"grad_norm": 0.525259011295812,
|
| 2118 |
+
"learning_rate": 4.215804814883959e-06,
|
| 2119 |
+
"loss": 1.0728,
|
| 2120 |
+
"step": 2970
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 2.502099076406381,
|
| 2124 |
+
"grad_norm": 0.4945489147573741,
|
| 2125 |
+
"learning_rate": 4.08104079277005e-06,
|
| 2126 |
+
"loss": 1.0567,
|
| 2127 |
+
"step": 2980
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 2.510495382031906,
|
| 2131 |
+
"grad_norm": 0.5289037096962053,
|
| 2132 |
+
"learning_rate": 3.948274213717015e-06,
|
| 2133 |
+
"loss": 1.0833,
|
| 2134 |
+
"step": 2990
|
| 2135 |
+
},
|
| 2136 |
+
{
|
| 2137 |
+
"epoch": 2.5188916876574305,
|
| 2138 |
+
"grad_norm": 0.520353171293509,
|
| 2139 |
+
"learning_rate": 3.817517754917802e-06,
|
| 2140 |
+
"loss": 1.0794,
|
| 2141 |
+
"step": 3000
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 2.5188916876574305,
|
| 2145 |
+
"eval_loss": 1.271108627319336,
|
| 2146 |
+
"eval_runtime": 37.5109,
|
| 2147 |
+
"eval_samples_per_second": 10.264,
|
| 2148 |
+
"eval_steps_per_second": 2.586,
|
| 2149 |
+
"step": 3000
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 2.5272879932829557,
|
| 2153 |
+
"grad_norm": 0.47216191409024166,
|
| 2154 |
+
"learning_rate": 3.6887839016293807e-06,
|
| 2155 |
+
"loss": 1.0797,
|
| 2156 |
+
"step": 3010
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 2.53568429890848,
|
| 2160 |
+
"grad_norm": 0.5168618762259352,
|
| 2161 |
+
"learning_rate": 3.5620849459805983e-06,
|
| 2162 |
+
"loss": 1.0809,
|
| 2163 |
+
"step": 3020
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 2.544080604534005,
|
| 2167 |
+
"grad_norm": 0.5702868726978964,
|
| 2168 |
+
"learning_rate": 3.4374329857984637e-06,
|
| 2169 |
+
"loss": 1.0841,
|
| 2170 |
+
"step": 3030
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 2.55247691015953,
|
| 2174 |
+
"grad_norm": 0.5324695724343947,
|
| 2175 |
+
"learning_rate": 3.314839923453003e-06,
|
| 2176 |
+
"loss": 1.0783,
|
| 2177 |
+
"step": 3040
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 2.5608732157850547,
|
| 2181 |
+
"grad_norm": 0.5856245582119982,
|
| 2182 |
+
"learning_rate": 3.1943174647207474e-06,
|
| 2183 |
+
"loss": 1.0657,
|
| 2184 |
+
"step": 3050
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 2.5692695214105794,
|
| 2188 |
+
"grad_norm": 0.5622043278789876,
|
| 2189 |
+
"learning_rate": 3.0758771176669948e-06,
|
| 2190 |
+
"loss": 1.0811,
|
| 2191 |
+
"step": 3060
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 2.577665827036104,
|
| 2195 |
+
"grad_norm": 0.5164314316915084,
|
| 2196 |
+
"learning_rate": 2.9595301915470246e-06,
|
| 2197 |
+
"loss": 1.085,
|
| 2198 |
+
"step": 3070
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 2.586062132661629,
|
| 2202 |
+
"grad_norm": 0.5302956251674786,
|
| 2203 |
+
"learning_rate": 2.8452877957261537e-06,
|
| 2204 |
+
"loss": 1.0682,
|
| 2205 |
+
"step": 3080
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 2.5944584382871536,
|
| 2209 |
+
"grad_norm": 0.5749370462779954,
|
| 2210 |
+
"learning_rate": 2.73316083861902e-06,
|
| 2211 |
+
"loss": 1.0705,
|
| 2212 |
+
"step": 3090
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 2.6028547439126783,
|
| 2216 |
+
"grad_norm": 0.528328794476431,
|
| 2217 |
+
"learning_rate": 2.623160026647978e-06,
|
| 2218 |
+
"loss": 1.08,
|
| 2219 |
+
"step": 3100
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 2.611251049538203,
|
| 2223 |
+
"grad_norm": 0.5235304102916779,
|
| 2224 |
+
"learning_rate": 2.515295863220796e-06,
|
| 2225 |
+
"loss": 1.055,
|
| 2226 |
+
"step": 3110
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 2.619647355163728,
|
| 2230 |
+
"grad_norm": 0.5312138570667937,
|
| 2231 |
+
"learning_rate": 2.4095786477277253e-06,
|
| 2232 |
+
"loss": 1.0541,
|
| 2233 |
+
"step": 3120
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 2.6280436607892526,
|
| 2237 |
+
"grad_norm": 0.5443569327871725,
|
| 2238 |
+
"learning_rate": 2.3060184745580972e-06,
|
| 2239 |
+
"loss": 1.0712,
|
| 2240 |
+
"step": 3130
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 2.6364399664147777,
|
| 2244 |
+
"grad_norm": 0.48867610044695425,
|
| 2245 |
+
"learning_rate": 2.2046252321364368e-06,
|
| 2246 |
+
"loss": 1.0785,
|
| 2247 |
+
"step": 3140
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 2.644836272040302,
|
| 2251 |
+
"grad_norm": 0.5523392458602866,
|
| 2252 |
+
"learning_rate": 2.1054086019782785e-06,
|
| 2253 |
+
"loss": 1.0623,
|
| 2254 |
+
"step": 3150
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 2.653232577665827,
|
| 2258 |
+
"grad_norm": 0.5415782176601202,
|
| 2259 |
+
"learning_rate": 2.008378057765725e-06,
|
| 2260 |
+
"loss": 1.0629,
|
| 2261 |
+
"step": 3160
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 2.661628883291352,
|
| 2265 |
+
"grad_norm": 0.5486934335306471,
|
| 2266 |
+
"learning_rate": 1.913542864442849e-06,
|
| 2267 |
+
"loss": 1.0682,
|
| 2268 |
+
"step": 3170
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 2.6700251889168767,
|
| 2272 |
+
"grad_norm": 0.5176003080136743,
|
| 2273 |
+
"learning_rate": 1.8209120773310468e-06,
|
| 2274 |
+
"loss": 1.0352,
|
| 2275 |
+
"step": 3180
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 2.6784214945424014,
|
| 2279 |
+
"grad_norm": 0.5415079933956461,
|
| 2280 |
+
"learning_rate": 1.7304945412643825e-06,
|
| 2281 |
+
"loss": 1.0617,
|
| 2282 |
+
"step": 3190
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 2.686817800167926,
|
| 2286 |
+
"grad_norm": 0.5461504669544123,
|
| 2287 |
+
"learning_rate": 1.6422988897450392e-06,
|
| 2288 |
+
"loss": 1.0712,
|
| 2289 |
+
"step": 3200
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 2.695214105793451,
|
| 2293 |
+
"grad_norm": 0.5356674558567608,
|
| 2294 |
+
"learning_rate": 1.5563335441189458e-06,
|
| 2295 |
+
"loss": 1.0735,
|
| 2296 |
+
"step": 3210
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 2.7036104114189756,
|
| 2300 |
+
"grad_norm": 0.5177763070870024,
|
| 2301 |
+
"learning_rate": 1.4726067127716958e-06,
|
| 2302 |
+
"loss": 1.0621,
|
| 2303 |
+
"step": 3220
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 2.7120067170445004,
|
| 2307 |
+
"grad_norm": 0.5696590300719524,
|
| 2308 |
+
"learning_rate": 1.3911263903447241e-06,
|
| 2309 |
+
"loss": 1.0652,
|
| 2310 |
+
"step": 3230
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 2.720403022670025,
|
| 2314 |
+
"grad_norm": 0.4910661226799717,
|
| 2315 |
+
"learning_rate": 1.3119003569719789e-06,
|
| 2316 |
+
"loss": 1.0622,
|
| 2317 |
+
"step": 3240
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 2.72879932829555,
|
| 2321 |
+
"grad_norm": 0.5036708642245188,
|
| 2322 |
+
"learning_rate": 1.2349361775370245e-06,
|
| 2323 |
+
"loss": 1.063,
|
| 2324 |
+
"step": 3250
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 2.7371956339210746,
|
| 2328 |
+
"grad_norm": 0.5183995701558842,
|
| 2329 |
+
"learning_rate": 1.1602412009507014e-06,
|
| 2330 |
+
"loss": 1.0598,
|
| 2331 |
+
"step": 3260
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 2.7455919395465997,
|
| 2335 |
+
"grad_norm": 0.4923573412537131,
|
| 2336 |
+
"learning_rate": 1.087822559449425e-06,
|
| 2337 |
+
"loss": 1.037,
|
| 2338 |
+
"step": 3270
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 2.753988245172124,
|
| 2342 |
+
"grad_norm": 0.5351167051477966,
|
| 2343 |
+
"learning_rate": 1.0176871679141664e-06,
|
| 2344 |
+
"loss": 1.0721,
|
| 2345 |
+
"step": 3280
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 2.762384550797649,
|
| 2349 |
+
"grad_norm": 0.5138535715173458,
|
| 2350 |
+
"learning_rate": 9.49841723210182e-07,
|
| 2351 |
+
"loss": 1.0624,
|
| 2352 |
+
"step": 3290
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 2.770780856423174,
|
| 2356 |
+
"grad_norm": 0.5562194428595657,
|
| 2357 |
+
"learning_rate": 8.842927035475763e-07,
|
| 2358 |
+
"loss": 1.0868,
|
| 2359 |
+
"step": 3300
|
| 2360 |
+
},
|
| 2361 |
+
{
|
| 2362 |
+
"epoch": 2.7791771620486987,
|
| 2363 |
+
"grad_norm": 0.5352125908767389,
|
| 2364 |
+
"learning_rate": 8.210463678626989e-07,
|
| 2365 |
+
"loss": 1.0509,
|
| 2366 |
+
"step": 3310
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 2.7875734676742234,
|
| 2370 |
+
"grad_norm": 0.5617426830025886,
|
| 2371 |
+
"learning_rate": 7.601087552205621e-07,
|
| 2372 |
+
"loss": 1.0568,
|
| 2373 |
+
"step": 3320
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 2.795969773299748,
|
| 2377 |
+
"grad_norm": 0.5445075989319467,
|
| 2378 |
+
"learning_rate": 7.014856842381484e-07,
|
| 2379 |
+
"loss": 1.0745,
|
| 2380 |
+
"step": 3330
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 2.804366078925273,
|
| 2384 |
+
"grad_norm": 0.5294691240524095,
|
| 2385 |
+
"learning_rate": 6.451827525288612e-07,
|
| 2386 |
+
"loss": 1.0708,
|
| 2387 |
+
"step": 3340
|
| 2388 |
+
},
|
| 2389 |
+
{
|
| 2390 |
+
"epoch": 2.8127623845507976,
|
| 2391 |
+
"grad_norm": 0.5469429580225932,
|
| 2392 |
+
"learning_rate": 5.912053361680159e-07,
|
| 2393 |
+
"loss": 1.0772,
|
| 2394 |
+
"step": 3350
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 2.8211586901763224,
|
| 2398 |
+
"grad_norm": 0.5197486795769499,
|
| 2399 |
+
"learning_rate": 5.395585891795174e-07,
|
| 2400 |
+
"loss": 1.0437,
|
| 2401 |
+
"step": 3360
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 2.829554995801847,
|
| 2405 |
+
"grad_norm": 0.5804047203332435,
|
| 2406 |
+
"learning_rate": 4.90247443043712e-07,
|
| 2407 |
+
"loss": 1.064,
|
| 2408 |
+
"step": 3370
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 2.837951301427372,
|
| 2412 |
+
"grad_norm": 0.5645169239872654,
|
| 2413 |
+
"learning_rate": 4.43276606226542e-07,
|
| 2414 |
+
"loss": 1.0702,
|
| 2415 |
+
"step": 3380
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 2.8463476070528966,
|
| 2419 |
+
"grad_norm": 0.5350844658781305,
|
| 2420 |
+
"learning_rate": 3.9865056372990763e-07,
|
| 2421 |
+
"loss": 1.0731,
|
| 2422 |
+
"step": 3390
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 2.8547439126784218,
|
| 2426 |
+
"grad_norm": 0.5252887341555534,
|
| 2427 |
+
"learning_rate": 3.563735766634629e-07,
|
| 2428 |
+
"loss": 1.0576,
|
| 2429 |
+
"step": 3400
|
| 2430 |
+
},
|
| 2431 |
+
{
|
| 2432 |
+
"epoch": 2.863140218303946,
|
| 2433 |
+
"grad_norm": 0.4893728491690957,
|
| 2434 |
+
"learning_rate": 3.164496818377188e-07,
|
| 2435 |
+
"loss": 1.0526,
|
| 2436 |
+
"step": 3410
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 2.8715365239294712,
|
| 2440 |
+
"grad_norm": 0.5238206701307841,
|
| 2441 |
+
"learning_rate": 2.7888269137860413e-07,
|
| 2442 |
+
"loss": 1.0631,
|
| 2443 |
+
"step": 3420
|
| 2444 |
+
},
|
| 2445 |
+
{
|
| 2446 |
+
"epoch": 2.879932829554996,
|
| 2447 |
+
"grad_norm": 0.5616265204939842,
|
| 2448 |
+
"learning_rate": 2.436761923634456e-07,
|
| 2449 |
+
"loss": 1.1013,
|
| 2450 |
+
"step": 3430
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"epoch": 2.8883291351805207,
|
| 2454 |
+
"grad_norm": 0.52918003457263,
|
| 2455 |
+
"learning_rate": 2.108335464784811e-07,
|
| 2456 |
+
"loss": 1.0737,
|
| 2457 |
+
"step": 3440
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 2.8967254408060454,
|
| 2461 |
+
"grad_norm": 0.5686960611410121,
|
| 2462 |
+
"learning_rate": 1.8035788969784638e-07,
|
| 2463 |
+
"loss": 1.0823,
|
| 2464 |
+
"step": 3450
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 2.90512174643157,
|
| 2468 |
+
"grad_norm": 0.4994659787416042,
|
| 2469 |
+
"learning_rate": 1.522521319841541e-07,
|
| 2470 |
+
"loss": 1.0692,
|
| 2471 |
+
"step": 3460
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 2.913518052057095,
|
| 2475 |
+
"grad_norm": 0.5629206398185236,
|
| 2476 |
+
"learning_rate": 1.2651895701063233e-07,
|
| 2477 |
+
"loss": 1.0698,
|
| 2478 |
+
"step": 3470
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 2.9219143576826196,
|
| 2482 |
+
"grad_norm": 0.5675179585041099,
|
| 2483 |
+
"learning_rate": 1.0316082190486298e-07,
|
| 2484 |
+
"loss": 1.0746,
|
| 2485 |
+
"step": 3480
|
| 2486 |
+
},
|
| 2487 |
+
{
|
| 2488 |
+
"epoch": 2.9303106633081444,
|
| 2489 |
+
"grad_norm": 0.5175715407014831,
|
| 2490 |
+
"learning_rate": 8.217995701418624e-08,
|
| 2491 |
+
"loss": 1.0454,
|
| 2492 |
+
"step": 3490
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 2.938706968933669,
|
| 2496 |
+
"grad_norm": 0.5240794519085307,
|
| 2497 |
+
"learning_rate": 6.35783656927097e-08,
|
| 2498 |
+
"loss": 1.056,
|
| 2499 |
+
"step": 3500
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 2.947103274559194,
|
| 2503 |
+
"grad_norm": 0.49756212695927055,
|
| 2504 |
+
"learning_rate": 4.7357824110044743e-08,
|
| 2505 |
+
"loss": 1.0689,
|
| 2506 |
+
"step": 3510
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 2.9554995801847186,
|
| 2510 |
+
"grad_norm": 0.5251282063460313,
|
| 2511 |
+
"learning_rate": 3.351988108168391e-08,
|
| 2512 |
+
"loss": 1.0603,
|
| 2513 |
+
"step": 3520
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 2.9638958858102438,
|
| 2517 |
+
"grad_norm": 0.5422569799063949,
|
| 2518 |
+
"learning_rate": 2.2065857921133025e-08,
|
| 2519 |
+
"loss": 1.0822,
|
| 2520 |
+
"step": 3530
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 2.972292191435768,
|
| 2524 |
+
"grad_norm": 0.543354269594394,
|
| 2525 |
+
"learning_rate": 1.2996848313734378e-08,
|
| 2526 |
+
"loss": 1.075,
|
| 2527 |
+
"step": 3540
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 2.9806884970612932,
|
| 2531 |
+
"grad_norm": 0.5348306207009784,
|
| 2532 |
+
"learning_rate": 6.31371821224469e-09,
|
| 2533 |
+
"loss": 1.0692,
|
| 2534 |
+
"step": 3550
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 2.9890848026868175,
|
| 2538 |
+
"grad_norm": 0.5488049489103773,
|
| 2539 |
+
"learning_rate": 2.017105754134607e-09,
|
| 2540 |
+
"loss": 1.0895,
|
| 2541 |
+
"step": 3560
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 2.9974811083123427,
|
| 2545 |
+
"grad_norm": 0.552482965655854,
|
| 2546 |
+
"learning_rate": 1.0742120067630535e-10,
|
| 2547 |
+
"loss": 1.0465,
|
| 2548 |
+
"step": 3570
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 3.0,
|
| 2552 |
+
"step": 3573,
|
| 2553 |
+
"total_flos": 234145141751808.0,
|
| 2554 |
+
"train_loss": 1.2264163225196905,
|
| 2555 |
+
"train_runtime": 4742.9935,
|
| 2556 |
+
"train_samples_per_second": 24.104,
|
| 2557 |
+
"train_steps_per_second": 0.753
|
| 2558 |
+
}
|
| 2559 |
+
],
|
| 2560 |
+
"logging_steps": 10,
|
| 2561 |
+
"max_steps": 3573,
|
| 2562 |
+
"num_input_tokens_seen": 0,
|
| 2563 |
+
"num_train_epochs": 3,
|
| 2564 |
+
"save_steps": 500,
|
| 2565 |
+
"stateful_callbacks": {
|
| 2566 |
+
"TrainerControl": {
|
| 2567 |
+
"args": {
|
| 2568 |
+
"should_epoch_stop": false,
|
| 2569 |
+
"should_evaluate": false,
|
| 2570 |
+
"should_log": false,
|
| 2571 |
+
"should_save": true,
|
| 2572 |
+
"should_training_stop": true
|
| 2573 |
+
},
|
| 2574 |
+
"attributes": {}
|
| 2575 |
+
}
|
| 2576 |
+
},
|
| 2577 |
+
"total_flos": 234145141751808.0,
|
| 2578 |
+
"train_batch_size": 8,
|
| 2579 |
+
"trial_name": null,
|
| 2580 |
+
"trial_params": null
|
| 2581 |
+
}
|
training_eval_loss.png
ADDED
|
training_loss.png
ADDED
|