End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2667 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -4,6 +4,7 @@ license: apache-2.0
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
model-index:
|
| 9 |
- name: seed_code_multiple_samples_all_scale_up_4K
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# seed_code_multiple_samples_all_scale_up_4K
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 4 |
base_model: Qwen/Qwen2.5-7B-Instruct
|
| 5 |
tags:
|
| 6 |
- llama-factory
|
| 7 |
+
- full
|
| 8 |
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: seed_code_multiple_samples_all_scale_up_4K
|
|
|
|
| 16 |
|
| 17 |
# seed_code_multiple_samples_all_scale_up_4K
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the mlfoundations-dev/seed_code_multiple_samples_all_scale_up_4K dataset.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 372577161248768.0,
|
| 4 |
+
"train_loss": 0.8553513642946879,
|
| 5 |
+
"train_runtime": 20932.817,
|
| 6 |
+
"train_samples_per_second": 1.72,
|
| 7 |
+
"train_steps_per_second": 0.018
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 372577161248768.0,
|
| 4 |
+
"train_loss": 0.8553513642946879,
|
| 5 |
+
"train_runtime": 20932.817,
|
| 6 |
+
"train_samples_per_second": 1.72,
|
| 7 |
+
"train_steps_per_second": 0.018
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2667 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 375,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008,
|
| 13 |
+
"grad_norm": 6.951127529144287,
|
| 14 |
+
"learning_rate": 2.6315789473684213e-07,
|
| 15 |
+
"loss": 1.2947,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.016,
|
| 20 |
+
"grad_norm": 6.943446636199951,
|
| 21 |
+
"learning_rate": 5.263157894736843e-07,
|
| 22 |
+
"loss": 1.2776,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.024,
|
| 27 |
+
"grad_norm": 6.9143147468566895,
|
| 28 |
+
"learning_rate": 7.894736842105263e-07,
|
| 29 |
+
"loss": 1.272,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.032,
|
| 34 |
+
"grad_norm": 6.79904317855835,
|
| 35 |
+
"learning_rate": 1.0526315789473685e-06,
|
| 36 |
+
"loss": 1.2738,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04,
|
| 41 |
+
"grad_norm": 6.621460437774658,
|
| 42 |
+
"learning_rate": 1.3157894736842106e-06,
|
| 43 |
+
"loss": 1.2432,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.048,
|
| 48 |
+
"grad_norm": 6.126567840576172,
|
| 49 |
+
"learning_rate": 1.5789473684210526e-06,
|
| 50 |
+
"loss": 1.2115,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.056,
|
| 55 |
+
"grad_norm": 5.147610187530518,
|
| 56 |
+
"learning_rate": 1.8421052631578948e-06,
|
| 57 |
+
"loss": 1.2493,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.064,
|
| 62 |
+
"grad_norm": 4.801726818084717,
|
| 63 |
+
"learning_rate": 2.105263157894737e-06,
|
| 64 |
+
"loss": 1.2107,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.072,
|
| 69 |
+
"grad_norm": 4.432796001434326,
|
| 70 |
+
"learning_rate": 2.368421052631579e-06,
|
| 71 |
+
"loss": 1.1829,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.08,
|
| 76 |
+
"grad_norm": 2.735476016998291,
|
| 77 |
+
"learning_rate": 2.631578947368421e-06,
|
| 78 |
+
"loss": 1.1258,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.088,
|
| 83 |
+
"grad_norm": 2.625633716583252,
|
| 84 |
+
"learning_rate": 2.8947368421052634e-06,
|
| 85 |
+
"loss": 1.1722,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.096,
|
| 90 |
+
"grad_norm": 2.6213021278381348,
|
| 91 |
+
"learning_rate": 3.157894736842105e-06,
|
| 92 |
+
"loss": 1.1329,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.104,
|
| 97 |
+
"grad_norm": 4.202518463134766,
|
| 98 |
+
"learning_rate": 3.421052631578948e-06,
|
| 99 |
+
"loss": 1.1283,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.112,
|
| 104 |
+
"grad_norm": 4.3722615242004395,
|
| 105 |
+
"learning_rate": 3.6842105263157896e-06,
|
| 106 |
+
"loss": 1.1362,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.12,
|
| 111 |
+
"grad_norm": 4.284238338470459,
|
| 112 |
+
"learning_rate": 3.947368421052632e-06,
|
| 113 |
+
"loss": 1.122,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.128,
|
| 118 |
+
"grad_norm": 3.9620132446289062,
|
| 119 |
+
"learning_rate": 4.210526315789474e-06,
|
| 120 |
+
"loss": 1.098,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.136,
|
| 125 |
+
"grad_norm": 2.9353275299072266,
|
| 126 |
+
"learning_rate": 4.473684210526316e-06,
|
| 127 |
+
"loss": 1.0722,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.144,
|
| 132 |
+
"grad_norm": 2.8066329956054688,
|
| 133 |
+
"learning_rate": 4.736842105263158e-06,
|
| 134 |
+
"loss": 1.0788,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.152,
|
| 139 |
+
"grad_norm": 2.426243543624878,
|
| 140 |
+
"learning_rate": 5e-06,
|
| 141 |
+
"loss": 1.0823,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.16,
|
| 146 |
+
"grad_norm": 1.9383955001831055,
|
| 147 |
+
"learning_rate": 5.263157894736842e-06,
|
| 148 |
+
"loss": 1.0643,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.168,
|
| 153 |
+
"grad_norm": 1.5883516073226929,
|
| 154 |
+
"learning_rate": 5.526315789473685e-06,
|
| 155 |
+
"loss": 1.0186,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.176,
|
| 160 |
+
"grad_norm": 1.5131279230117798,
|
| 161 |
+
"learning_rate": 5.789473684210527e-06,
|
| 162 |
+
"loss": 1.0316,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.184,
|
| 167 |
+
"grad_norm": 1.5010865926742554,
|
| 168 |
+
"learning_rate": 6.0526315789473685e-06,
|
| 169 |
+
"loss": 1.0223,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.192,
|
| 174 |
+
"grad_norm": 1.1548634767532349,
|
| 175 |
+
"learning_rate": 6.31578947368421e-06,
|
| 176 |
+
"loss": 0.9821,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.2,
|
| 181 |
+
"grad_norm": 1.17434561252594,
|
| 182 |
+
"learning_rate": 6.578947368421054e-06,
|
| 183 |
+
"loss": 0.9551,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.208,
|
| 188 |
+
"grad_norm": 1.1539326906204224,
|
| 189 |
+
"learning_rate": 6.842105263157896e-06,
|
| 190 |
+
"loss": 0.9725,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.216,
|
| 195 |
+
"grad_norm": 1.1245882511138916,
|
| 196 |
+
"learning_rate": 7.1052631578947375e-06,
|
| 197 |
+
"loss": 0.9684,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.224,
|
| 202 |
+
"grad_norm": 1.1288598775863647,
|
| 203 |
+
"learning_rate": 7.368421052631579e-06,
|
| 204 |
+
"loss": 0.9963,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.232,
|
| 209 |
+
"grad_norm": 1.0191954374313354,
|
| 210 |
+
"learning_rate": 7.631578947368423e-06,
|
| 211 |
+
"loss": 1.0004,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.24,
|
| 216 |
+
"grad_norm": 0.9676952362060547,
|
| 217 |
+
"learning_rate": 7.894736842105265e-06,
|
| 218 |
+
"loss": 0.9658,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.248,
|
| 223 |
+
"grad_norm": 1.1406854391098022,
|
| 224 |
+
"learning_rate": 8.157894736842106e-06,
|
| 225 |
+
"loss": 0.9879,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.256,
|
| 230 |
+
"grad_norm": 0.9583864212036133,
|
| 231 |
+
"learning_rate": 8.421052631578948e-06,
|
| 232 |
+
"loss": 0.9472,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.264,
|
| 237 |
+
"grad_norm": 0.9609050750732422,
|
| 238 |
+
"learning_rate": 8.68421052631579e-06,
|
| 239 |
+
"loss": 0.978,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.272,
|
| 244 |
+
"grad_norm": 0.8762646913528442,
|
| 245 |
+
"learning_rate": 8.947368421052632e-06,
|
| 246 |
+
"loss": 0.9468,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.28,
|
| 251 |
+
"grad_norm": 0.8770004510879517,
|
| 252 |
+
"learning_rate": 9.210526315789474e-06,
|
| 253 |
+
"loss": 0.9643,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.288,
|
| 258 |
+
"grad_norm": 0.8456284403800964,
|
| 259 |
+
"learning_rate": 9.473684210526315e-06,
|
| 260 |
+
"loss": 0.9534,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.296,
|
| 265 |
+
"grad_norm": 0.9351610541343689,
|
| 266 |
+
"learning_rate": 9.736842105263159e-06,
|
| 267 |
+
"loss": 0.9389,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.304,
|
| 272 |
+
"grad_norm": 0.8555827736854553,
|
| 273 |
+
"learning_rate": 1e-05,
|
| 274 |
+
"loss": 0.9065,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.312,
|
| 279 |
+
"grad_norm": 0.8321928977966309,
|
| 280 |
+
"learning_rate": 9.99978274148479e-06,
|
| 281 |
+
"loss": 0.9469,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.32,
|
| 286 |
+
"grad_norm": 0.7721774578094482,
|
| 287 |
+
"learning_rate": 9.999130984819662e-06,
|
| 288 |
+
"loss": 0.9243,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.328,
|
| 293 |
+
"grad_norm": 0.7928785085678101,
|
| 294 |
+
"learning_rate": 9.998044786644492e-06,
|
| 295 |
+
"loss": 0.9615,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.336,
|
| 300 |
+
"grad_norm": 0.6871277093887329,
|
| 301 |
+
"learning_rate": 9.9965242413536e-06,
|
| 302 |
+
"loss": 0.9437,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.344,
|
| 307 |
+
"grad_norm": 0.7224629521369934,
|
| 308 |
+
"learning_rate": 9.994569481087552e-06,
|
| 309 |
+
"loss": 0.9486,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.352,
|
| 314 |
+
"grad_norm": 0.8585007190704346,
|
| 315 |
+
"learning_rate": 9.992180675721671e-06,
|
| 316 |
+
"loss": 0.9382,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.36,
|
| 321 |
+
"grad_norm": 0.7822074294090271,
|
| 322 |
+
"learning_rate": 9.989358032851283e-06,
|
| 323 |
+
"loss": 0.9179,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.368,
|
| 328 |
+
"grad_norm": 0.7207604646682739,
|
| 329 |
+
"learning_rate": 9.986101797773667e-06,
|
| 330 |
+
"loss": 0.9395,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.376,
|
| 335 |
+
"grad_norm": 0.889811635017395,
|
| 336 |
+
"learning_rate": 9.98241225346674e-06,
|
| 337 |
+
"loss": 0.886,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.384,
|
| 342 |
+
"grad_norm": 0.8683562874794006,
|
| 343 |
+
"learning_rate": 9.978289720564471e-06,
|
| 344 |
+
"loss": 0.8996,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.392,
|
| 349 |
+
"grad_norm": 0.7052727341651917,
|
| 350 |
+
"learning_rate": 9.97373455732901e-06,
|
| 351 |
+
"loss": 0.9052,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4,
|
| 356 |
+
"grad_norm": 0.8796941041946411,
|
| 357 |
+
"learning_rate": 9.968747159619556e-06,
|
| 358 |
+
"loss": 0.9217,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.408,
|
| 363 |
+
"grad_norm": 0.7054159641265869,
|
| 364 |
+
"learning_rate": 9.963327960857962e-06,
|
| 365 |
+
"loss": 0.9534,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.416,
|
| 370 |
+
"grad_norm": 0.7010794281959534,
|
| 371 |
+
"learning_rate": 9.957477431991053e-06,
|
| 372 |
+
"loss": 0.904,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.424,
|
| 377 |
+
"grad_norm": 0.7052696347236633,
|
| 378 |
+
"learning_rate": 9.95119608144972e-06,
|
| 379 |
+
"loss": 0.9093,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.432,
|
| 384 |
+
"grad_norm": 0.6790952086448669,
|
| 385 |
+
"learning_rate": 9.944484455104716e-06,
|
| 386 |
+
"loss": 0.9236,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.44,
|
| 391 |
+
"grad_norm": 0.6545783877372742,
|
| 392 |
+
"learning_rate": 9.937343136219234e-06,
|
| 393 |
+
"loss": 0.9295,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.448,
|
| 398 |
+
"grad_norm": 0.6248077154159546,
|
| 399 |
+
"learning_rate": 9.929772745398207e-06,
|
| 400 |
+
"loss": 0.9171,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.456,
|
| 405 |
+
"grad_norm": 0.6094949245452881,
|
| 406 |
+
"learning_rate": 9.921773940534382e-06,
|
| 407 |
+
"loss": 0.8842,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.464,
|
| 412 |
+
"grad_norm": 0.7010917663574219,
|
| 413 |
+
"learning_rate": 9.913347416751148e-06,
|
| 414 |
+
"loss": 0.9079,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.472,
|
| 419 |
+
"grad_norm": 0.7240394353866577,
|
| 420 |
+
"learning_rate": 9.904493906342124e-06,
|
| 421 |
+
"loss": 0.9433,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.48,
|
| 426 |
+
"grad_norm": 0.71834397315979,
|
| 427 |
+
"learning_rate": 9.895214178707516e-06,
|
| 428 |
+
"loss": 0.9086,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.488,
|
| 433 |
+
"grad_norm": 0.7287778854370117,
|
| 434 |
+
"learning_rate": 9.885509040287267e-06,
|
| 435 |
+
"loss": 0.8941,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.496,
|
| 440 |
+
"grad_norm": 0.6156062483787537,
|
| 441 |
+
"learning_rate": 9.875379334490962e-06,
|
| 442 |
+
"loss": 0.8547,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.504,
|
| 447 |
+
"grad_norm": 0.6256719827651978,
|
| 448 |
+
"learning_rate": 9.864825941624538e-06,
|
| 449 |
+
"loss": 0.8925,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.512,
|
| 454 |
+
"grad_norm": 0.6274106502532959,
|
| 455 |
+
"learning_rate": 9.853849778813777e-06,
|
| 456 |
+
"loss": 0.9073,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.52,
|
| 461 |
+
"grad_norm": 0.6543683409690857,
|
| 462 |
+
"learning_rate": 9.842451799924616e-06,
|
| 463 |
+
"loss": 0.9398,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.528,
|
| 468 |
+
"grad_norm": 0.6689018607139587,
|
| 469 |
+
"learning_rate": 9.830632995480243e-06,
|
| 470 |
+
"loss": 0.9149,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.536,
|
| 475 |
+
"grad_norm": 0.6306348443031311,
|
| 476 |
+
"learning_rate": 9.818394392575018e-06,
|
| 477 |
+
"loss": 0.8812,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.544,
|
| 482 |
+
"grad_norm": 0.7672179937362671,
|
| 483 |
+
"learning_rate": 9.805737054785223e-06,
|
| 484 |
+
"loss": 0.9074,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.552,
|
| 489 |
+
"grad_norm": 0.7313401103019714,
|
| 490 |
+
"learning_rate": 9.792662082076618e-06,
|
| 491 |
+
"loss": 0.8983,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.56,
|
| 496 |
+
"grad_norm": 0.6449005007743835,
|
| 497 |
+
"learning_rate": 9.779170610708872e-06,
|
| 498 |
+
"loss": 0.95,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.568,
|
| 503 |
+
"grad_norm": 0.7076218128204346,
|
| 504 |
+
"learning_rate": 9.765263813136796e-06,
|
| 505 |
+
"loss": 0.8961,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.576,
|
| 510 |
+
"grad_norm": 0.6033435463905334,
|
| 511 |
+
"learning_rate": 9.750942897908468e-06,
|
| 512 |
+
"loss": 0.8845,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.584,
|
| 517 |
+
"grad_norm": 0.6751191020011902,
|
| 518 |
+
"learning_rate": 9.736209109560201e-06,
|
| 519 |
+
"loss": 0.9402,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.592,
|
| 524 |
+
"grad_norm": 0.6485176682472229,
|
| 525 |
+
"learning_rate": 9.721063728508384e-06,
|
| 526 |
+
"loss": 0.8936,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.6,
|
| 531 |
+
"grad_norm": 0.6805006265640259,
|
| 532 |
+
"learning_rate": 9.705508070938219e-06,
|
| 533 |
+
"loss": 0.9238,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.608,
|
| 538 |
+
"grad_norm": 0.7293475270271301,
|
| 539 |
+
"learning_rate": 9.689543488689332e-06,
|
| 540 |
+
"loss": 0.8917,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.616,
|
| 545 |
+
"grad_norm": 0.6713281869888306,
|
| 546 |
+
"learning_rate": 9.673171369138297e-06,
|
| 547 |
+
"loss": 0.9051,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.624,
|
| 552 |
+
"grad_norm": 0.7109168171882629,
|
| 553 |
+
"learning_rate": 9.656393135078067e-06,
|
| 554 |
+
"loss": 0.9147,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.632,
|
| 559 |
+
"grad_norm": 0.7239603400230408,
|
| 560 |
+
"learning_rate": 9.639210244594335e-06,
|
| 561 |
+
"loss": 0.878,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.64,
|
| 566 |
+
"grad_norm": 0.6019280552864075,
|
| 567 |
+
"learning_rate": 9.621624190938802e-06,
|
| 568 |
+
"loss": 0.8919,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.648,
|
| 573 |
+
"grad_norm": 0.7745220065116882,
|
| 574 |
+
"learning_rate": 9.603636502399436e-06,
|
| 575 |
+
"loss": 0.8745,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.656,
|
| 580 |
+
"grad_norm": 0.8410776853561401,
|
| 581 |
+
"learning_rate": 9.585248742167638e-06,
|
| 582 |
+
"loss": 0.9138,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.664,
|
| 587 |
+
"grad_norm": 0.5263102054595947,
|
| 588 |
+
"learning_rate": 9.566462508202403e-06,
|
| 589 |
+
"loss": 0.8924,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.672,
|
| 594 |
+
"grad_norm": 0.9095488786697388,
|
| 595 |
+
"learning_rate": 9.547279433091446e-06,
|
| 596 |
+
"loss": 0.9034,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.68,
|
| 601 |
+
"grad_norm": 0.6388502717018127,
|
| 602 |
+
"learning_rate": 9.527701183909336e-06,
|
| 603 |
+
"loss": 0.9433,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.688,
|
| 608 |
+
"grad_norm": 0.5810441970825195,
|
| 609 |
+
"learning_rate": 9.507729462072615e-06,
|
| 610 |
+
"loss": 0.9063,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.696,
|
| 615 |
+
"grad_norm": 0.6178363561630249,
|
| 616 |
+
"learning_rate": 9.48736600319193e-06,
|
| 617 |
+
"loss": 0.8911,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.704,
|
| 622 |
+
"grad_norm": 0.6862829923629761,
|
| 623 |
+
"learning_rate": 9.466612576921223e-06,
|
| 624 |
+
"loss": 0.8723,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.712,
|
| 629 |
+
"grad_norm": 0.6216509938240051,
|
| 630 |
+
"learning_rate": 9.445470986803922e-06,
|
| 631 |
+
"loss": 0.9128,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.72,
|
| 636 |
+
"grad_norm": 0.6560526490211487,
|
| 637 |
+
"learning_rate": 9.423943070116219e-06,
|
| 638 |
+
"loss": 0.8976,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.728,
|
| 643 |
+
"grad_norm": 0.6992607116699219,
|
| 644 |
+
"learning_rate": 9.402030697707398e-06,
|
| 645 |
+
"loss": 0.9017,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.736,
|
| 650 |
+
"grad_norm": 0.6775327920913696,
|
| 651 |
+
"learning_rate": 9.37973577383726e-06,
|
| 652 |
+
"loss": 0.9186,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.744,
|
| 657 |
+
"grad_norm": 0.5652820467948914,
|
| 658 |
+
"learning_rate": 9.357060236010626e-06,
|
| 659 |
+
"loss": 0.9105,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.752,
|
| 664 |
+
"grad_norm": 0.6999420523643494,
|
| 665 |
+
"learning_rate": 9.334006054808966e-06,
|
| 666 |
+
"loss": 0.9206,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.76,
|
| 671 |
+
"grad_norm": 0.6290011405944824,
|
| 672 |
+
"learning_rate": 9.310575233719155e-06,
|
| 673 |
+
"loss": 0.9198,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.768,
|
| 678 |
+
"grad_norm": 0.7411805987358093,
|
| 679 |
+
"learning_rate": 9.28676980895935e-06,
|
| 680 |
+
"loss": 0.8981,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.776,
|
| 685 |
+
"grad_norm": 0.600430965423584,
|
| 686 |
+
"learning_rate": 9.262591849302049e-06,
|
| 687 |
+
"loss": 0.8867,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.784,
|
| 692 |
+
"grad_norm": 0.6421351432800293,
|
| 693 |
+
"learning_rate": 9.238043455894294e-06,
|
| 694 |
+
"loss": 0.9053,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.792,
|
| 699 |
+
"grad_norm": 0.7420929670333862,
|
| 700 |
+
"learning_rate": 9.213126762075088e-06,
|
| 701 |
+
"loss": 0.8963,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.8,
|
| 706 |
+
"grad_norm": 0.5919941663742065,
|
| 707 |
+
"learning_rate": 9.187843933189994e-06,
|
| 708 |
+
"loss": 0.8234,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.808,
|
| 713 |
+
"grad_norm": 0.7523407340049744,
|
| 714 |
+
"learning_rate": 9.162197166402957e-06,
|
| 715 |
+
"loss": 0.9065,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.816,
|
| 720 |
+
"grad_norm": 0.6169841885566711,
|
| 721 |
+
"learning_rate": 9.136188690505363e-06,
|
| 722 |
+
"loss": 0.9195,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.824,
|
| 727 |
+
"grad_norm": 0.6386623382568359,
|
| 728 |
+
"learning_rate": 9.109820765722357e-06,
|
| 729 |
+
"loss": 0.8867,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.832,
|
| 734 |
+
"grad_norm": 0.5823110342025757,
|
| 735 |
+
"learning_rate": 9.083095683516414e-06,
|
| 736 |
+
"loss": 0.9039,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.84,
|
| 741 |
+
"grad_norm": 0.5956842303276062,
|
| 742 |
+
"learning_rate": 9.056015766388205e-06,
|
| 743 |
+
"loss": 0.9004,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.848,
|
| 748 |
+
"grad_norm": 0.5407257676124573,
|
| 749 |
+
"learning_rate": 9.028583367674767e-06,
|
| 750 |
+
"loss": 0.8644,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.856,
|
| 755 |
+
"grad_norm": 0.6211457252502441,
|
| 756 |
+
"learning_rate": 9.00080087134498e-06,
|
| 757 |
+
"loss": 0.9033,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.864,
|
| 762 |
+
"grad_norm": 0.5631619095802307,
|
| 763 |
+
"learning_rate": 8.972670691792409e-06,
|
| 764 |
+
"loss": 0.846,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.872,
|
| 769 |
+
"grad_norm": 0.5860468745231628,
|
| 770 |
+
"learning_rate": 8.944195273625472e-06,
|
| 771 |
+
"loss": 0.9196,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.88,
|
| 776 |
+
"grad_norm": 0.5991073846817017,
|
| 777 |
+
"learning_rate": 8.915377091454992e-06,
|
| 778 |
+
"loss": 0.8949,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.888,
|
| 783 |
+
"grad_norm": 0.5442657470703125,
|
| 784 |
+
"learning_rate": 8.886218649679162e-06,
|
| 785 |
+
"loss": 0.8848,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.896,
|
| 790 |
+
"grad_norm": 0.49064090847969055,
|
| 791 |
+
"learning_rate": 8.856722482265886e-06,
|
| 792 |
+
"loss": 0.8571,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.904,
|
| 797 |
+
"grad_norm": 0.6109375953674316,
|
| 798 |
+
"learning_rate": 8.826891152532579e-06,
|
| 799 |
+
"loss": 0.874,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.912,
|
| 804 |
+
"grad_norm": 0.5844782590866089,
|
| 805 |
+
"learning_rate": 8.796727252923403e-06,
|
| 806 |
+
"loss": 0.8715,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.92,
|
| 811 |
+
"grad_norm": 0.5409021377563477,
|
| 812 |
+
"learning_rate": 8.766233404783975e-06,
|
| 813 |
+
"loss": 0.8836,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.928,
|
| 818 |
+
"grad_norm": 0.5358095169067383,
|
| 819 |
+
"learning_rate": 8.735412258133562e-06,
|
| 820 |
+
"loss": 0.8686,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.936,
|
| 825 |
+
"grad_norm": 0.6133047938346863,
|
| 826 |
+
"learning_rate": 8.704266491434787e-06,
|
| 827 |
+
"loss": 0.8661,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.944,
|
| 832 |
+
"grad_norm": 0.6640852093696594,
|
| 833 |
+
"learning_rate": 8.672798811360863e-06,
|
| 834 |
+
"loss": 0.8357,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.952,
|
| 839 |
+
"grad_norm": 0.6169326901435852,
|
| 840 |
+
"learning_rate": 8.641011952560372e-06,
|
| 841 |
+
"loss": 0.8542,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.96,
|
| 846 |
+
"grad_norm": 0.6841595768928528,
|
| 847 |
+
"learning_rate": 8.608908677419606e-06,
|
| 848 |
+
"loss": 0.898,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.968,
|
| 853 |
+
"grad_norm": 0.6243302822113037,
|
| 854 |
+
"learning_rate": 8.576491775822527e-06,
|
| 855 |
+
"loss": 0.8579,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.976,
|
| 860 |
+
"grad_norm": 0.5553723573684692,
|
| 861 |
+
"learning_rate": 8.543764064908295e-06,
|
| 862 |
+
"loss": 0.8628,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.984,
|
| 867 |
+
"grad_norm": 0.6372663974761963,
|
| 868 |
+
"learning_rate": 8.510728388826464e-06,
|
| 869 |
+
"loss": 0.8672,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.992,
|
| 874 |
+
"grad_norm": 0.6535351872444153,
|
| 875 |
+
"learning_rate": 8.477387618489808e-06,
|
| 876 |
+
"loss": 0.849,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.0,
|
| 881 |
+
"grad_norm": 0.6568986773490906,
|
| 882 |
+
"learning_rate": 8.443744651324828e-06,
|
| 883 |
+
"loss": 0.8512,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.008,
|
| 888 |
+
"grad_norm": 0.6753949522972107,
|
| 889 |
+
"learning_rate": 8.409802411019962e-06,
|
| 890 |
+
"loss": 0.8784,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.016,
|
| 895 |
+
"grad_norm": 0.7170552015304565,
|
| 896 |
+
"learning_rate": 8.375563847271506e-06,
|
| 897 |
+
"loss": 0.8431,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.024,
|
| 902 |
+
"grad_norm": 0.6562941074371338,
|
| 903 |
+
"learning_rate": 8.341031935527267e-06,
|
| 904 |
+
"loss": 0.8489,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.032,
|
| 909 |
+
"grad_norm": 0.6256612539291382,
|
| 910 |
+
"learning_rate": 8.306209676727994e-06,
|
| 911 |
+
"loss": 0.8586,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.04,
|
| 916 |
+
"grad_norm": 0.7772700190544128,
|
| 917 |
+
"learning_rate": 8.271100097046585e-06,
|
| 918 |
+
"loss": 0.8498,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.048,
|
| 923 |
+
"grad_norm": 0.6408899426460266,
|
| 924 |
+
"learning_rate": 8.235706247625098e-06,
|
| 925 |
+
"loss": 0.8249,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.056,
|
| 930 |
+
"grad_norm": 0.5903573632240295,
|
| 931 |
+
"learning_rate": 8.200031204309604e-06,
|
| 932 |
+
"loss": 0.8051,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.064,
|
| 937 |
+
"grad_norm": 0.7947317957878113,
|
| 938 |
+
"learning_rate": 8.16407806738288e-06,
|
| 939 |
+
"loss": 0.8263,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.072,
|
| 944 |
+
"grad_norm": 0.6758819818496704,
|
| 945 |
+
"learning_rate": 8.127849961294984e-06,
|
| 946 |
+
"loss": 0.801,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.08,
|
| 951 |
+
"grad_norm": 0.682586669921875,
|
| 952 |
+
"learning_rate": 8.091350034391732e-06,
|
| 953 |
+
"loss": 0.8507,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.088,
|
| 958 |
+
"grad_norm": 0.7249560356140137,
|
| 959 |
+
"learning_rate": 8.05458145864109e-06,
|
| 960 |
+
"loss": 0.836,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.096,
|
| 965 |
+
"grad_norm": 0.6341940760612488,
|
| 966 |
+
"learning_rate": 8.017547429357532e-06,
|
| 967 |
+
"loss": 0.846,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.104,
|
| 972 |
+
"grad_norm": 0.7217475175857544,
|
| 973 |
+
"learning_rate": 7.980251164924342e-06,
|
| 974 |
+
"loss": 0.8489,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.112,
|
| 979 |
+
"grad_norm": 0.5944041609764099,
|
| 980 |
+
"learning_rate": 7.94269590651393e-06,
|
| 981 |
+
"loss": 0.8149,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.12,
|
| 986 |
+
"grad_norm": 0.6496677994728088,
|
| 987 |
+
"learning_rate": 7.904884917806174e-06,
|
| 988 |
+
"loss": 0.8401,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.1280000000000001,
|
| 993 |
+
"grad_norm": 0.604494571685791,
|
| 994 |
+
"learning_rate": 7.866821484704777e-06,
|
| 995 |
+
"loss": 0.8379,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.1360000000000001,
|
| 1000 |
+
"grad_norm": 0.5416861772537231,
|
| 1001 |
+
"learning_rate": 7.828508915051724e-06,
|
| 1002 |
+
"loss": 0.817,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.144,
|
| 1007 |
+
"grad_norm": 0.5813770890235901,
|
| 1008 |
+
"learning_rate": 7.789950538339813e-06,
|
| 1009 |
+
"loss": 0.8452,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.152,
|
| 1014 |
+
"grad_norm": 0.5333764553070068,
|
| 1015 |
+
"learning_rate": 7.751149705423313e-06,
|
| 1016 |
+
"loss": 0.8404,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.16,
|
| 1021 |
+
"grad_norm": 0.5556795597076416,
|
| 1022 |
+
"learning_rate": 7.712109788226763e-06,
|
| 1023 |
+
"loss": 0.8142,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.168,
|
| 1028 |
+
"grad_norm": 0.6207058429718018,
|
| 1029 |
+
"learning_rate": 7.672834179451943e-06,
|
| 1030 |
+
"loss": 0.8746,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.176,
|
| 1035 |
+
"grad_norm": 0.5369215607643127,
|
| 1036 |
+
"learning_rate": 7.633326292283028e-06,
|
| 1037 |
+
"loss": 0.8469,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.184,
|
| 1042 |
+
"grad_norm": 0.5977884531021118,
|
| 1043 |
+
"learning_rate": 7.593589560089984e-06,
|
| 1044 |
+
"loss": 0.8328,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.192,
|
| 1049 |
+
"grad_norm": 0.6059456467628479,
|
| 1050 |
+
"learning_rate": 7.553627436130183e-06,
|
| 1051 |
+
"loss": 0.8582,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.2,
|
| 1056 |
+
"grad_norm": 0.5836285948753357,
|
| 1057 |
+
"learning_rate": 7.513443393248312e-06,
|
| 1058 |
+
"loss": 0.8558,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.208,
|
| 1063 |
+
"grad_norm": 0.6144264936447144,
|
| 1064 |
+
"learning_rate": 7.473040923574567e-06,
|
| 1065 |
+
"loss": 0.8549,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.216,
|
| 1070 |
+
"grad_norm": 0.626693844795227,
|
| 1071 |
+
"learning_rate": 7.432423538221179e-06,
|
| 1072 |
+
"loss": 0.8473,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.224,
|
| 1077 |
+
"grad_norm": 0.6432238221168518,
|
| 1078 |
+
"learning_rate": 7.391594766977277e-06,
|
| 1079 |
+
"loss": 0.8281,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.232,
|
| 1084 |
+
"grad_norm": 0.6119612455368042,
|
| 1085 |
+
"learning_rate": 7.350558158002154e-06,
|
| 1086 |
+
"loss": 0.8419,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.24,
|
| 1091 |
+
"grad_norm": 0.6895710229873657,
|
| 1092 |
+
"learning_rate": 7.3093172775169e-06,
|
| 1093 |
+
"loss": 0.8408,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.248,
|
| 1098 |
+
"grad_norm": 0.6269011497497559,
|
| 1099 |
+
"learning_rate": 7.2678757094945e-06,
|
| 1100 |
+
"loss": 0.8345,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.256,
|
| 1105 |
+
"grad_norm": 0.5723068118095398,
|
| 1106 |
+
"learning_rate": 7.226237055348369e-06,
|
| 1107 |
+
"loss": 0.8654,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.264,
|
| 1112 |
+
"grad_norm": 0.6293651461601257,
|
| 1113 |
+
"learning_rate": 7.184404933619377e-06,
|
| 1114 |
+
"loss": 0.8387,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.272,
|
| 1119 |
+
"grad_norm": 0.6416763067245483,
|
| 1120 |
+
"learning_rate": 7.142382979661386e-06,
|
| 1121 |
+
"loss": 0.8317,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.28,
|
| 1126 |
+
"grad_norm": 0.5021880865097046,
|
| 1127 |
+
"learning_rate": 7.100174845325327e-06,
|
| 1128 |
+
"loss": 0.8281,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.288,
|
| 1133 |
+
"grad_norm": 0.567487895488739,
|
| 1134 |
+
"learning_rate": 7.057784198641835e-06,
|
| 1135 |
+
"loss": 0.8646,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.296,
|
| 1140 |
+
"grad_norm": 0.5684318542480469,
|
| 1141 |
+
"learning_rate": 7.015214723502496e-06,
|
| 1142 |
+
"loss": 0.8293,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.304,
|
| 1147 |
+
"grad_norm": 0.5071728825569153,
|
| 1148 |
+
"learning_rate": 6.972470119339692e-06,
|
| 1149 |
+
"loss": 0.8097,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.312,
|
| 1154 |
+
"grad_norm": 0.6408015489578247,
|
| 1155 |
+
"learning_rate": 6.929554100805118e-06,
|
| 1156 |
+
"loss": 0.85,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.32,
|
| 1161 |
+
"grad_norm": 0.5955577492713928,
|
| 1162 |
+
"learning_rate": 6.886470397446958e-06,
|
| 1163 |
+
"loss": 0.8343,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.328,
|
| 1168 |
+
"grad_norm": 0.4974721670150757,
|
| 1169 |
+
"learning_rate": 6.843222753385785e-06,
|
| 1170 |
+
"loss": 0.7749,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.336,
|
| 1175 |
+
"grad_norm": 0.6704075336456299,
|
| 1176 |
+
"learning_rate": 6.799814926989171e-06,
|
| 1177 |
+
"loss": 0.8165,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.3439999999999999,
|
| 1182 |
+
"grad_norm": 0.5451616644859314,
|
| 1183 |
+
"learning_rate": 6.756250690545079e-06,
|
| 1184 |
+
"loss": 0.8349,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.3519999999999999,
|
| 1189 |
+
"grad_norm": 0.5713831782341003,
|
| 1190 |
+
"learning_rate": 6.712533829934042e-06,
|
| 1191 |
+
"loss": 0.8224,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.3599999999999999,
|
| 1196 |
+
"grad_norm": 0.5288674831390381,
|
| 1197 |
+
"learning_rate": 6.6686681443001485e-06,
|
| 1198 |
+
"loss": 0.8518,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.3679999999999999,
|
| 1203 |
+
"grad_norm": 0.6580489873886108,
|
| 1204 |
+
"learning_rate": 6.62465744572089e-06,
|
| 1205 |
+
"loss": 0.8736,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.376,
|
| 1210 |
+
"grad_norm": 0.5576747059822083,
|
| 1211 |
+
"learning_rate": 6.580505558875878e-06,
|
| 1212 |
+
"loss": 0.8171,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.384,
|
| 1217 |
+
"grad_norm": 0.6031728386878967,
|
| 1218 |
+
"learning_rate": 6.536216320714466e-06,
|
| 1219 |
+
"loss": 0.8325,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.392,
|
| 1224 |
+
"grad_norm": 0.49704813957214355,
|
| 1225 |
+
"learning_rate": 6.491793580122301e-06,
|
| 1226 |
+
"loss": 0.8348,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.4,
|
| 1231 |
+
"grad_norm": 0.5445089340209961,
|
| 1232 |
+
"learning_rate": 6.447241197586847e-06,
|
| 1233 |
+
"loss": 0.8679,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.408,
|
| 1238 |
+
"grad_norm": 0.5517768263816833,
|
| 1239 |
+
"learning_rate": 6.402563044861899e-06,
|
| 1240 |
+
"loss": 0.8414,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.416,
|
| 1245 |
+
"grad_norm": 0.5023953318595886,
|
| 1246 |
+
"learning_rate": 6.357763004631104e-06,
|
| 1247 |
+
"loss": 0.7706,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.424,
|
| 1252 |
+
"grad_norm": 0.5183501243591309,
|
| 1253 |
+
"learning_rate": 6.312844970170551e-06,
|
| 1254 |
+
"loss": 0.8145,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.432,
|
| 1259 |
+
"grad_norm": 0.5883236527442932,
|
| 1260 |
+
"learning_rate": 6.267812845010431e-06,
|
| 1261 |
+
"loss": 0.8071,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.44,
|
| 1266 |
+
"grad_norm": 0.5993454456329346,
|
| 1267 |
+
"learning_rate": 6.2226705425958e-06,
|
| 1268 |
+
"loss": 0.8338,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.448,
|
| 1273 |
+
"grad_norm": 0.5319005250930786,
|
| 1274 |
+
"learning_rate": 6.177421985946499e-06,
|
| 1275 |
+
"loss": 0.856,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.456,
|
| 1280 |
+
"grad_norm": 0.6088922619819641,
|
| 1281 |
+
"learning_rate": 6.132071107316221e-06,
|
| 1282 |
+
"loss": 0.8405,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.464,
|
| 1287 |
+
"grad_norm": 0.565377950668335,
|
| 1288 |
+
"learning_rate": 6.0866218478507875e-06,
|
| 1289 |
+
"loss": 0.848,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.472,
|
| 1294 |
+
"grad_norm": 0.5202862024307251,
|
| 1295 |
+
"learning_rate": 6.041078157245649e-06,
|
| 1296 |
+
"loss": 0.8139,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.48,
|
| 1301 |
+
"grad_norm": 0.5979906320571899,
|
| 1302 |
+
"learning_rate": 5.995443993402647e-06,
|
| 1303 |
+
"loss": 0.8267,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.488,
|
| 1308 |
+
"grad_norm": 0.5841382145881653,
|
| 1309 |
+
"learning_rate": 5.949723322086053e-06,
|
| 1310 |
+
"loss": 0.8011,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.496,
|
| 1315 |
+
"grad_norm": 0.5658543705940247,
|
| 1316 |
+
"learning_rate": 5.9039201165779315e-06,
|
| 1317 |
+
"loss": 0.786,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.504,
|
| 1322 |
+
"grad_norm": 0.5966721773147583,
|
| 1323 |
+
"learning_rate": 5.858038357332851e-06,
|
| 1324 |
+
"loss": 0.8083,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.512,
|
| 1329 |
+
"grad_norm": 0.5970141291618347,
|
| 1330 |
+
"learning_rate": 5.812082031631966e-06,
|
| 1331 |
+
"loss": 0.8421,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.52,
|
| 1336 |
+
"grad_norm": 0.5003347992897034,
|
| 1337 |
+
"learning_rate": 5.766055133236513e-06,
|
| 1338 |
+
"loss": 0.823,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.528,
|
| 1343 |
+
"grad_norm": 0.5702365040779114,
|
| 1344 |
+
"learning_rate": 5.7199616620407325e-06,
|
| 1345 |
+
"loss": 0.8309,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.536,
|
| 1350 |
+
"grad_norm": 0.6028835773468018,
|
| 1351 |
+
"learning_rate": 5.673805623724272e-06,
|
| 1352 |
+
"loss": 0.8088,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.544,
|
| 1357 |
+
"grad_norm": 0.5072699189186096,
|
| 1358 |
+
"learning_rate": 5.627591029404072e-06,
|
| 1359 |
+
"loss": 0.8199,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.552,
|
| 1364 |
+
"grad_norm": 0.5302557945251465,
|
| 1365 |
+
"learning_rate": 5.581321895285787e-06,
|
| 1366 |
+
"loss": 0.8194,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.56,
|
| 1371 |
+
"grad_norm": 0.48366543650627136,
|
| 1372 |
+
"learning_rate": 5.535002242314772e-06,
|
| 1373 |
+
"loss": 0.8176,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.568,
|
| 1378 |
+
"grad_norm": 0.6495940089225769,
|
| 1379 |
+
"learning_rate": 5.488636095826636e-06,
|
| 1380 |
+
"loss": 0.8078,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.576,
|
| 1385 |
+
"grad_norm": 0.4819113314151764,
|
| 1386 |
+
"learning_rate": 5.4422274851974356e-06,
|
| 1387 |
+
"loss": 0.8397,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.584,
|
| 1392 |
+
"grad_norm": 0.5706694722175598,
|
| 1393 |
+
"learning_rate": 5.395780443493508e-06,
|
| 1394 |
+
"loss": 0.8549,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.592,
|
| 1399 |
+
"grad_norm": 0.5659922957420349,
|
| 1400 |
+
"learning_rate": 5.34929900712098e-06,
|
| 1401 |
+
"loss": 0.7686,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.6,
|
| 1406 |
+
"grad_norm": 0.5568146705627441,
|
| 1407 |
+
"learning_rate": 5.302787215474992e-06,
|
| 1408 |
+
"loss": 0.8329,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.608,
|
| 1413 |
+
"grad_norm": 0.5732328295707703,
|
| 1414 |
+
"learning_rate": 5.256249110588659e-06,
|
| 1415 |
+
"loss": 0.833,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.616,
|
| 1420 |
+
"grad_norm": 0.5560725927352905,
|
| 1421 |
+
"learning_rate": 5.209688736781811e-06,
|
| 1422 |
+
"loss": 0.812,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.624,
|
| 1427 |
+
"grad_norm": 0.49767476320266724,
|
| 1428 |
+
"learning_rate": 5.163110140309518e-06,
|
| 1429 |
+
"loss": 0.8408,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.6320000000000001,
|
| 1434 |
+
"grad_norm": 0.5641460418701172,
|
| 1435 |
+
"learning_rate": 5.116517369010467e-06,
|
| 1436 |
+
"loss": 0.8294,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.6400000000000001,
|
| 1441 |
+
"grad_norm": 0.6225250363349915,
|
| 1442 |
+
"learning_rate": 5.069914471955179e-06,
|
| 1443 |
+
"loss": 0.7811,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.6480000000000001,
|
| 1448 |
+
"grad_norm": 0.5748852491378784,
|
| 1449 |
+
"learning_rate": 5.023305499094145e-06,
|
| 1450 |
+
"loss": 0.8307,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.6560000000000001,
|
| 1455 |
+
"grad_norm": 0.5865118503570557,
|
| 1456 |
+
"learning_rate": 4.976694500905858e-06,
|
| 1457 |
+
"loss": 0.8046,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.6640000000000001,
|
| 1462 |
+
"grad_norm": 0.5250575542449951,
|
| 1463 |
+
"learning_rate": 4.930085528044823e-06,
|
| 1464 |
+
"loss": 0.8315,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.6720000000000002,
|
| 1469 |
+
"grad_norm": 0.5795397162437439,
|
| 1470 |
+
"learning_rate": 4.883482630989536e-06,
|
| 1471 |
+
"loss": 0.8472,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.6800000000000002,
|
| 1476 |
+
"grad_norm": 0.6399658918380737,
|
| 1477 |
+
"learning_rate": 4.8368898596904834e-06,
|
| 1478 |
+
"loss": 0.8163,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.688,
|
| 1483 |
+
"grad_norm": 0.5113149881362915,
|
| 1484 |
+
"learning_rate": 4.790311263218191e-06,
|
| 1485 |
+
"loss": 0.8362,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.696,
|
| 1490 |
+
"grad_norm": 0.4925471544265747,
|
| 1491 |
+
"learning_rate": 4.743750889411342e-06,
|
| 1492 |
+
"loss": 0.8364,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.704,
|
| 1497 |
+
"grad_norm": 0.5226654410362244,
|
| 1498 |
+
"learning_rate": 4.697212784525009e-06,
|
| 1499 |
+
"loss": 0.8156,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.712,
|
| 1504 |
+
"grad_norm": 0.49383535981178284,
|
| 1505 |
+
"learning_rate": 4.65070099287902e-06,
|
| 1506 |
+
"loss": 0.8363,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.72,
|
| 1511 |
+
"grad_norm": 0.5424585342407227,
|
| 1512 |
+
"learning_rate": 4.604219556506492e-06,
|
| 1513 |
+
"loss": 0.8252,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.728,
|
| 1518 |
+
"grad_norm": 0.4512873589992523,
|
| 1519 |
+
"learning_rate": 4.557772514802564e-06,
|
| 1520 |
+
"loss": 0.8188,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.736,
|
| 1525 |
+
"grad_norm": 0.49828025698661804,
|
| 1526 |
+
"learning_rate": 4.511363904173366e-06,
|
| 1527 |
+
"loss": 0.835,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.744,
|
| 1532 |
+
"grad_norm": 0.5537604093551636,
|
| 1533 |
+
"learning_rate": 4.46499775768523e-06,
|
| 1534 |
+
"loss": 0.8072,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.752,
|
| 1539 |
+
"grad_norm": 0.509067714214325,
|
| 1540 |
+
"learning_rate": 4.418678104714214e-06,
|
| 1541 |
+
"loss": 0.8331,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.76,
|
| 1546 |
+
"grad_norm": 0.49603044986724854,
|
| 1547 |
+
"learning_rate": 4.372408970595931e-06,
|
| 1548 |
+
"loss": 0.8031,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.768,
|
| 1553 |
+
"grad_norm": 0.5341010689735413,
|
| 1554 |
+
"learning_rate": 4.326194376275729e-06,
|
| 1555 |
+
"loss": 0.8117,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.776,
|
| 1560 |
+
"grad_norm": 0.49610334634780884,
|
| 1561 |
+
"learning_rate": 4.280038337959268e-06,
|
| 1562 |
+
"loss": 0.8137,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.784,
|
| 1567 |
+
"grad_norm": 0.4939770996570587,
|
| 1568 |
+
"learning_rate": 4.2339448667634885e-06,
|
| 1569 |
+
"loss": 0.7834,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.792,
|
| 1574 |
+
"grad_norm": 0.5622074007987976,
|
| 1575 |
+
"learning_rate": 4.187917968368036e-06,
|
| 1576 |
+
"loss": 0.8227,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.8,
|
| 1581 |
+
"grad_norm": 0.6220787763595581,
|
| 1582 |
+
"learning_rate": 4.141961642667152e-06,
|
| 1583 |
+
"loss": 0.8526,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.808,
|
| 1588 |
+
"grad_norm": 0.5713546276092529,
|
| 1589 |
+
"learning_rate": 4.09607988342207e-06,
|
| 1590 |
+
"loss": 0.8529,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.8159999999999998,
|
| 1595 |
+
"grad_norm": 0.5482434034347534,
|
| 1596 |
+
"learning_rate": 4.0502766779139485e-06,
|
| 1597 |
+
"loss": 0.8142,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.8239999999999998,
|
| 1602 |
+
"grad_norm": 0.6168124079704285,
|
| 1603 |
+
"learning_rate": 4.0045560065973535e-06,
|
| 1604 |
+
"loss": 0.8127,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.8319999999999999,
|
| 1609 |
+
"grad_norm": 0.5725027918815613,
|
| 1610 |
+
"learning_rate": 3.958921842754351e-06,
|
| 1611 |
+
"loss": 0.8405,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.8399999999999999,
|
| 1616 |
+
"grad_norm": 0.5908339023590088,
|
| 1617 |
+
"learning_rate": 3.913378152149214e-06,
|
| 1618 |
+
"loss": 0.7938,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.8479999999999999,
|
| 1623 |
+
"grad_norm": 0.5032182931900024,
|
| 1624 |
+
"learning_rate": 3.86792889268378e-06,
|
| 1625 |
+
"loss": 0.8237,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.8559999999999999,
|
| 1630 |
+
"grad_norm": 0.4820937514305115,
|
| 1631 |
+
"learning_rate": 3.8225780140535025e-06,
|
| 1632 |
+
"loss": 0.8157,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.8639999999999999,
|
| 1637 |
+
"grad_norm": 0.5697328448295593,
|
| 1638 |
+
"learning_rate": 3.777329457404202e-06,
|
| 1639 |
+
"loss": 0.8471,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.8719999999999999,
|
| 1644 |
+
"grad_norm": 0.5110496282577515,
|
| 1645 |
+
"learning_rate": 3.7321871549895715e-06,
|
| 1646 |
+
"loss": 0.8184,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.88,
|
| 1651 |
+
"grad_norm": 0.4492500424385071,
|
| 1652 |
+
"learning_rate": 3.68715502982945e-06,
|
| 1653 |
+
"loss": 0.8048,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.888,
|
| 1658 |
+
"grad_norm": 0.5033578872680664,
|
| 1659 |
+
"learning_rate": 3.6422369953688973e-06,
|
| 1660 |
+
"loss": 0.8462,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.896,
|
| 1665 |
+
"grad_norm": 0.47552290558815,
|
| 1666 |
+
"learning_rate": 3.5974369551381023e-06,
|
| 1667 |
+
"loss": 0.814,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.904,
|
| 1672 |
+
"grad_norm": 0.47769391536712646,
|
| 1673 |
+
"learning_rate": 3.5527588024131542e-06,
|
| 1674 |
+
"loss": 0.8076,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.912,
|
| 1679 |
+
"grad_norm": 0.47970548272132874,
|
| 1680 |
+
"learning_rate": 3.5082064198777e-06,
|
| 1681 |
+
"loss": 0.8063,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.92,
|
| 1686 |
+
"grad_norm": 0.5200381875038147,
|
| 1687 |
+
"learning_rate": 3.463783679285535e-06,
|
| 1688 |
+
"loss": 0.8212,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.928,
|
| 1693 |
+
"grad_norm": 0.4885690212249756,
|
| 1694 |
+
"learning_rate": 3.4194944411241213e-06,
|
| 1695 |
+
"loss": 0.8087,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.936,
|
| 1700 |
+
"grad_norm": 0.523141086101532,
|
| 1701 |
+
"learning_rate": 3.3753425542791106e-06,
|
| 1702 |
+
"loss": 0.8116,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.944,
|
| 1707 |
+
"grad_norm": 0.5129029750823975,
|
| 1708 |
+
"learning_rate": 3.3313318556998523e-06,
|
| 1709 |
+
"loss": 0.8062,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.952,
|
| 1714 |
+
"grad_norm": 0.4830564260482788,
|
| 1715 |
+
"learning_rate": 3.2874661700659586e-06,
|
| 1716 |
+
"loss": 0.8428,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.96,
|
| 1721 |
+
"grad_norm": 0.4503372609615326,
|
| 1722 |
+
"learning_rate": 3.2437493094549223e-06,
|
| 1723 |
+
"loss": 0.8268,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.968,
|
| 1728 |
+
"grad_norm": 0.4484092891216278,
|
| 1729 |
+
"learning_rate": 3.200185073010831e-06,
|
| 1730 |
+
"loss": 0.8086,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.976,
|
| 1735 |
+
"grad_norm": 0.41616371273994446,
|
| 1736 |
+
"learning_rate": 3.1567772466142156e-06,
|
| 1737 |
+
"loss": 0.8212,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.984,
|
| 1742 |
+
"grad_norm": 0.4492393732070923,
|
| 1743 |
+
"learning_rate": 3.1135296025530426e-06,
|
| 1744 |
+
"loss": 0.8217,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.992,
|
| 1749 |
+
"grad_norm": 0.48867785930633545,
|
| 1750 |
+
"learning_rate": 3.070445899194885e-06,
|
| 1751 |
+
"loss": 0.8394,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 2.0,
|
| 1756 |
+
"grad_norm": 0.5146781206130981,
|
| 1757 |
+
"learning_rate": 3.0275298806603102e-06,
|
| 1758 |
+
"loss": 0.8254,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 2.008,
|
| 1763 |
+
"grad_norm": 0.4940868616104126,
|
| 1764 |
+
"learning_rate": 2.984785276497507e-06,
|
| 1765 |
+
"loss": 0.7948,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.016,
|
| 1770 |
+
"grad_norm": 0.5990253686904907,
|
| 1771 |
+
"learning_rate": 2.9422158013581658e-06,
|
| 1772 |
+
"loss": 0.7709,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.024,
|
| 1777 |
+
"grad_norm": 0.52504962682724,
|
| 1778 |
+
"learning_rate": 2.899825154674674e-06,
|
| 1779 |
+
"loss": 0.7659,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.032,
|
| 1784 |
+
"grad_norm": 0.47968220710754395,
|
| 1785 |
+
"learning_rate": 2.8576170203386144e-06,
|
| 1786 |
+
"loss": 0.7978,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 2.04,
|
| 1791 |
+
"grad_norm": 0.5218414068222046,
|
| 1792 |
+
"learning_rate": 2.8155950663806234e-06,
|
| 1793 |
+
"loss": 0.8075,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 2.048,
|
| 1798 |
+
"grad_norm": 0.6022427082061768,
|
| 1799 |
+
"learning_rate": 2.7737629446516325e-06,
|
| 1800 |
+
"loss": 0.7967,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 2.056,
|
| 1805 |
+
"grad_norm": 0.45537319779396057,
|
| 1806 |
+
"learning_rate": 2.732124290505501e-06,
|
| 1807 |
+
"loss": 0.7637,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 2.064,
|
| 1812 |
+
"grad_norm": 0.47411370277404785,
|
| 1813 |
+
"learning_rate": 2.6906827224831024e-06,
|
| 1814 |
+
"loss": 0.7999,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 2.072,
|
| 1819 |
+
"grad_norm": 0.47417914867401123,
|
| 1820 |
+
"learning_rate": 2.6494418419978485e-06,
|
| 1821 |
+
"loss": 0.8002,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 2.08,
|
| 1826 |
+
"grad_norm": 0.47205376625061035,
|
| 1827 |
+
"learning_rate": 2.608405233022724e-06,
|
| 1828 |
+
"loss": 0.7841,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 2.088,
|
| 1833 |
+
"grad_norm": 0.4820186197757721,
|
| 1834 |
+
"learning_rate": 2.5675764617788233e-06,
|
| 1835 |
+
"loss": 0.7808,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 2.096,
|
| 1840 |
+
"grad_norm": 0.44489675760269165,
|
| 1841 |
+
"learning_rate": 2.526959076425434e-06,
|
| 1842 |
+
"loss": 0.7758,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.104,
|
| 1847 |
+
"grad_norm": 0.4890801012516022,
|
| 1848 |
+
"learning_rate": 2.4865566067516896e-06,
|
| 1849 |
+
"loss": 0.8111,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 2.112,
|
| 1854 |
+
"grad_norm": 0.4414970278739929,
|
| 1855 |
+
"learning_rate": 2.4463725638698182e-06,
|
| 1856 |
+
"loss": 0.7976,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 2.12,
|
| 1861 |
+
"grad_norm": 0.45259734988212585,
|
| 1862 |
+
"learning_rate": 2.406410439910017e-06,
|
| 1863 |
+
"loss": 0.7903,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.128,
|
| 1868 |
+
"grad_norm": 0.48778876662254333,
|
| 1869 |
+
"learning_rate": 2.366673707716973e-06,
|
| 1870 |
+
"loss": 0.8107,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 2.136,
|
| 1875 |
+
"grad_norm": 0.44795307517051697,
|
| 1876 |
+
"learning_rate": 2.327165820548059e-06,
|
| 1877 |
+
"loss": 0.7231,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 2.144,
|
| 1882 |
+
"grad_norm": 0.43515029549598694,
|
| 1883 |
+
"learning_rate": 2.287890211773238e-06,
|
| 1884 |
+
"loss": 0.7503,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 2.152,
|
| 1889 |
+
"grad_norm": 0.4571765661239624,
|
| 1890 |
+
"learning_rate": 2.2488502945766893e-06,
|
| 1891 |
+
"loss": 0.8206,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 2.16,
|
| 1896 |
+
"grad_norm": 0.4228002727031708,
|
| 1897 |
+
"learning_rate": 2.210049461660189e-06,
|
| 1898 |
+
"loss": 0.7817,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 2.168,
|
| 1903 |
+
"grad_norm": 0.4443894624710083,
|
| 1904 |
+
"learning_rate": 2.1714910849482777e-06,
|
| 1905 |
+
"loss": 0.7855,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 2.176,
|
| 1910 |
+
"grad_norm": 0.4512254297733307,
|
| 1911 |
+
"learning_rate": 2.1331785152952243e-06,
|
| 1912 |
+
"loss": 0.8024,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 2.184,
|
| 1917 |
+
"grad_norm": 0.5297703742980957,
|
| 1918 |
+
"learning_rate": 2.0951150821938278e-06,
|
| 1919 |
+
"loss": 0.7472,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 2.192,
|
| 1924 |
+
"grad_norm": 0.46972817182540894,
|
| 1925 |
+
"learning_rate": 2.0573040934860717e-06,
|
| 1926 |
+
"loss": 0.7666,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 2.2,
|
| 1931 |
+
"grad_norm": 0.48370957374572754,
|
| 1932 |
+
"learning_rate": 2.0197488350756618e-06,
|
| 1933 |
+
"loss": 0.7913,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.208,
|
| 1938 |
+
"grad_norm": 0.43513721227645874,
|
| 1939 |
+
"learning_rate": 1.98245257064247e-06,
|
| 1940 |
+
"loss": 0.7855,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 2.216,
|
| 1945 |
+
"grad_norm": 0.44624122977256775,
|
| 1946 |
+
"learning_rate": 1.945418541358911e-06,
|
| 1947 |
+
"loss": 0.7948,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.224,
|
| 1952 |
+
"grad_norm": 0.5190390944480896,
|
| 1953 |
+
"learning_rate": 1.9086499656082685e-06,
|
| 1954 |
+
"loss": 0.7953,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 2.232,
|
| 1959 |
+
"grad_norm": 0.4471817910671234,
|
| 1960 |
+
"learning_rate": 1.872150038705015e-06,
|
| 1961 |
+
"loss": 0.7723,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 2.24,
|
| 1966 |
+
"grad_norm": 0.4081578254699707,
|
| 1967 |
+
"learning_rate": 1.835921932617119e-06,
|
| 1968 |
+
"loss": 0.7863,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 2.248,
|
| 1973 |
+
"grad_norm": 0.4299153983592987,
|
| 1974 |
+
"learning_rate": 1.7999687956903955e-06,
|
| 1975 |
+
"loss": 0.7459,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.2560000000000002,
|
| 1980 |
+
"grad_norm": 0.44286420941352844,
|
| 1981 |
+
"learning_rate": 1.7642937523749038e-06,
|
| 1982 |
+
"loss": 0.7549,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 2.2640000000000002,
|
| 1987 |
+
"grad_norm": 0.5079202651977539,
|
| 1988 |
+
"learning_rate": 1.7288999029534177e-06,
|
| 1989 |
+
"loss": 0.7945,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 2.2720000000000002,
|
| 1994 |
+
"grad_norm": 0.5463809967041016,
|
| 1995 |
+
"learning_rate": 1.6937903232720076e-06,
|
| 1996 |
+
"loss": 0.8004,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 2.2800000000000002,
|
| 2001 |
+
"grad_norm": 0.4327748715877533,
|
| 2002 |
+
"learning_rate": 1.6589680644727347e-06,
|
| 2003 |
+
"loss": 0.7723,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 2.288,
|
| 2008 |
+
"grad_norm": 0.4358852505683899,
|
| 2009 |
+
"learning_rate": 1.6244361527284953e-06,
|
| 2010 |
+
"loss": 0.7721,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 2.296,
|
| 2015 |
+
"grad_norm": 0.41679710149765015,
|
| 2016 |
+
"learning_rate": 1.5901975889800387e-06,
|
| 2017 |
+
"loss": 0.7891,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 2.304,
|
| 2022 |
+
"grad_norm": 0.4202753007411957,
|
| 2023 |
+
"learning_rate": 1.556255348675174e-06,
|
| 2024 |
+
"loss": 0.7585,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 2.312,
|
| 2029 |
+
"grad_norm": 0.4579218029975891,
|
| 2030 |
+
"learning_rate": 1.522612381510195e-06,
|
| 2031 |
+
"loss": 0.7672,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 2.32,
|
| 2036 |
+
"grad_norm": 0.47101694345474243,
|
| 2037 |
+
"learning_rate": 1.489271611173538e-06,
|
| 2038 |
+
"loss": 0.7861,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 2.328,
|
| 2043 |
+
"grad_norm": 0.4235668182373047,
|
| 2044 |
+
"learning_rate": 1.4562359350917054e-06,
|
| 2045 |
+
"loss": 0.791,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 2.336,
|
| 2050 |
+
"grad_norm": 0.44197484850883484,
|
| 2051 |
+
"learning_rate": 1.423508224177474e-06,
|
| 2052 |
+
"loss": 0.7861,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 2.344,
|
| 2057 |
+
"grad_norm": 0.43885377049446106,
|
| 2058 |
+
"learning_rate": 1.3910913225803946e-06,
|
| 2059 |
+
"loss": 0.7965,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 2.352,
|
| 2064 |
+
"grad_norm": 0.4231855273246765,
|
| 2065 |
+
"learning_rate": 1.35898804743963e-06,
|
| 2066 |
+
"loss": 0.7815,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.36,
|
| 2071 |
+
"grad_norm": 0.48017239570617676,
|
| 2072 |
+
"learning_rate": 1.3272011886391368e-06,
|
| 2073 |
+
"loss": 0.8028,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 2.368,
|
| 2078 |
+
"grad_norm": 0.46980807185173035,
|
| 2079 |
+
"learning_rate": 1.295733508565213e-06,
|
| 2080 |
+
"loss": 0.7692,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.376,
|
| 2085 |
+
"grad_norm": 0.45487645268440247,
|
| 2086 |
+
"learning_rate": 1.2645877418664394e-06,
|
| 2087 |
+
"loss": 0.7546,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 2.384,
|
| 2092 |
+
"grad_norm": 0.44050565361976624,
|
| 2093 |
+
"learning_rate": 1.2337665952160266e-06,
|
| 2094 |
+
"loss": 0.798,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 2.392,
|
| 2099 |
+
"grad_norm": 0.4430494010448456,
|
| 2100 |
+
"learning_rate": 1.2032727470765982e-06,
|
| 2101 |
+
"loss": 0.7849,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.4,
|
| 2106 |
+
"grad_norm": 0.4752819836139679,
|
| 2107 |
+
"learning_rate": 1.1731088474674235e-06,
|
| 2108 |
+
"loss": 0.8045,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.408,
|
| 2113 |
+
"grad_norm": 0.4466467499732971,
|
| 2114 |
+
"learning_rate": 1.1432775177341165e-06,
|
| 2115 |
+
"loss": 0.7841,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 2.416,
|
| 2120 |
+
"grad_norm": 0.4148922562599182,
|
| 2121 |
+
"learning_rate": 1.11378135032084e-06,
|
| 2122 |
+
"loss": 0.7993,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.424,
|
| 2127 |
+
"grad_norm": 0.44224417209625244,
|
| 2128 |
+
"learning_rate": 1.08462290854501e-06,
|
| 2129 |
+
"loss": 0.7655,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 2.432,
|
| 2134 |
+
"grad_norm": 0.42925554513931274,
|
| 2135 |
+
"learning_rate": 1.0558047263745297e-06,
|
| 2136 |
+
"loss": 0.7607,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 2.44,
|
| 2141 |
+
"grad_norm": 0.4702661633491516,
|
| 2142 |
+
"learning_rate": 1.0273293082075914e-06,
|
| 2143 |
+
"loss": 0.7425,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.448,
|
| 2148 |
+
"grad_norm": 0.44665732979774475,
|
| 2149 |
+
"learning_rate": 9.991991286550207e-07,
|
| 2150 |
+
"loss": 0.8086,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.456,
|
| 2155 |
+
"grad_norm": 0.44039228558540344,
|
| 2156 |
+
"learning_rate": 9.71416632325235e-07,
|
| 2157 |
+
"loss": 0.7751,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.464,
|
| 2162 |
+
"grad_norm": 0.4385325610637665,
|
| 2163 |
+
"learning_rate": 9.439842336117954e-07,
|
| 2164 |
+
"loss": 0.8141,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.472,
|
| 2169 |
+
"grad_norm": 0.43111181259155273,
|
| 2170 |
+
"learning_rate": 9.169043164835867e-07,
|
| 2171 |
+
"loss": 0.762,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 2.48,
|
| 2176 |
+
"grad_norm": 0.4057021737098694,
|
| 2177 |
+
"learning_rate": 8.901792342776439e-07,
|
| 2178 |
+
"loss": 0.8327,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 2.488,
|
| 2183 |
+
"grad_norm": 0.4096888303756714,
|
| 2184 |
+
"learning_rate": 8.638113094946382e-07,
|
| 2185 |
+
"loss": 0.8251,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.496,
|
| 2190 |
+
"grad_norm": 0.4527164101600647,
|
| 2191 |
+
"learning_rate": 8.378028335970451e-07,
|
| 2192 |
+
"loss": 0.7605,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 2.504,
|
| 2197 |
+
"grad_norm": 0.4072956144809723,
|
| 2198 |
+
"learning_rate": 8.121560668100065e-07,
|
| 2199 |
+
"loss": 0.7836,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 2.512,
|
| 2204 |
+
"grad_norm": 0.4347890615463257,
|
| 2205 |
+
"learning_rate": 7.868732379249122e-07,
|
| 2206 |
+
"loss": 0.7888,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.52,
|
| 2211 |
+
"grad_norm": 0.44405418634414673,
|
| 2212 |
+
"learning_rate": 7.619565441057075e-07,
|
| 2213 |
+
"loss": 0.7476,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 2.528,
|
| 2218 |
+
"grad_norm": 0.4290863573551178,
|
| 2219 |
+
"learning_rate": 7.37408150697953e-07,
|
| 2220 |
+
"loss": 0.776,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 2.536,
|
| 2225 |
+
"grad_norm": 0.4163723289966583,
|
| 2226 |
+
"learning_rate": 7.132301910406503e-07,
|
| 2227 |
+
"loss": 0.7841,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 2.544,
|
| 2232 |
+
"grad_norm": 0.39156779646873474,
|
| 2233 |
+
"learning_rate": 6.894247662808456e-07,
|
| 2234 |
+
"loss": 0.8141,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 2.552,
|
| 2239 |
+
"grad_norm": 0.3718986511230469,
|
| 2240 |
+
"learning_rate": 6.659939451910341e-07,
|
| 2241 |
+
"loss": 0.8007,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 2.56,
|
| 2246 |
+
"grad_norm": 0.39346760511398315,
|
| 2247 |
+
"learning_rate": 6.429397639893758e-07,
|
| 2248 |
+
"loss": 0.7884,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 2.568,
|
| 2253 |
+
"grad_norm": 0.38905784487724304,
|
| 2254 |
+
"learning_rate": 6.202642261627411e-07,
|
| 2255 |
+
"loss": 0.7697,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 2.576,
|
| 2260 |
+
"grad_norm": 0.4061625301837921,
|
| 2261 |
+
"learning_rate": 5.979693022926025e-07,
|
| 2262 |
+
"loss": 0.7715,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 2.584,
|
| 2267 |
+
"grad_norm": 0.4345610737800598,
|
| 2268 |
+
"learning_rate": 5.760569298837825e-07,
|
| 2269 |
+
"loss": 0.785,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 2.592,
|
| 2274 |
+
"grad_norm": 0.3941546678543091,
|
| 2275 |
+
"learning_rate": 5.54529013196079e-07,
|
| 2276 |
+
"loss": 0.8025,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 2.6,
|
| 2281 |
+
"grad_norm": 0.36712539196014404,
|
| 2282 |
+
"learning_rate": 5.333874230787772e-07,
|
| 2283 |
+
"loss": 0.8007,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 2.608,
|
| 2288 |
+
"grad_norm": 0.3812052607536316,
|
| 2289 |
+
"learning_rate": 5.126339968080696e-07,
|
| 2290 |
+
"loss": 0.7889,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.616,
|
| 2295 |
+
"grad_norm": 0.4085177481174469,
|
| 2296 |
+
"learning_rate": 4.922705379273862e-07,
|
| 2297 |
+
"loss": 0.7641,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 2.624,
|
| 2302 |
+
"grad_norm": 0.42239484190940857,
|
| 2303 |
+
"learning_rate": 4.7229881609066387e-07,
|
| 2304 |
+
"loss": 0.7979,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 2.632,
|
| 2309 |
+
"grad_norm": 0.4121035635471344,
|
| 2310 |
+
"learning_rate": 4.5272056690855494e-07,
|
| 2311 |
+
"loss": 0.7858,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 2.64,
|
| 2316 |
+
"grad_norm": 0.38635125756263733,
|
| 2317 |
+
"learning_rate": 4.335374917975982e-07,
|
| 2318 |
+
"loss": 0.8095,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 2.648,
|
| 2323 |
+
"grad_norm": 0.38656213879585266,
|
| 2324 |
+
"learning_rate": 4.147512578323615e-07,
|
| 2325 |
+
"loss": 0.7669,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 2.656,
|
| 2330 |
+
"grad_norm": 0.41909682750701904,
|
| 2331 |
+
"learning_rate": 3.9636349760056427e-07,
|
| 2332 |
+
"loss": 0.7788,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.664,
|
| 2337 |
+
"grad_norm": 0.4400654137134552,
|
| 2338 |
+
"learning_rate": 3.783758090611983e-07,
|
| 2339 |
+
"loss": 0.7382,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 2.672,
|
| 2344 |
+
"grad_norm": 0.3858467638492584,
|
| 2345 |
+
"learning_rate": 3.6078975540566716e-07,
|
| 2346 |
+
"loss": 0.767,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 2.68,
|
| 2351 |
+
"grad_norm": 0.36701512336730957,
|
| 2352 |
+
"learning_rate": 3.4360686492193263e-07,
|
| 2353 |
+
"loss": 0.7928,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 2.6879999999999997,
|
| 2358 |
+
"grad_norm": 0.38759228587150574,
|
| 2359 |
+
"learning_rate": 3.268286308617041e-07,
|
| 2360 |
+
"loss": 0.8181,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 2.6959999999999997,
|
| 2365 |
+
"grad_norm": 0.38279959559440613,
|
| 2366 |
+
"learning_rate": 3.104565113106689e-07,
|
| 2367 |
+
"loss": 0.7943,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 2.7039999999999997,
|
| 2372 |
+
"grad_norm": 0.4579544961452484,
|
| 2373 |
+
"learning_rate": 2.9449192906178205e-07,
|
| 2374 |
+
"loss": 0.7605,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.7119999999999997,
|
| 2379 |
+
"grad_norm": 0.40908747911453247,
|
| 2380 |
+
"learning_rate": 2.789362714916172e-07,
|
| 2381 |
+
"loss": 0.8095,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 2.7199999999999998,
|
| 2386 |
+
"grad_norm": 0.37257906794548035,
|
| 2387 |
+
"learning_rate": 2.6379089043980064e-07,
|
| 2388 |
+
"loss": 0.7988,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 2.7279999999999998,
|
| 2393 |
+
"grad_norm": 0.431416779756546,
|
| 2394 |
+
"learning_rate": 2.4905710209153224e-07,
|
| 2395 |
+
"loss": 0.7819,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.7359999999999998,
|
| 2400 |
+
"grad_norm": 0.4172951579093933,
|
| 2401 |
+
"learning_rate": 2.3473618686320477e-07,
|
| 2402 |
+
"loss": 0.7597,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 2.7439999999999998,
|
| 2407 |
+
"grad_norm": 0.41569289565086365,
|
| 2408 |
+
"learning_rate": 2.208293892911284e-07,
|
| 2409 |
+
"loss": 0.7797,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 2.752,
|
| 2414 |
+
"grad_norm": 0.42253097891807556,
|
| 2415 |
+
"learning_rate": 2.0733791792338197e-07,
|
| 2416 |
+
"loss": 0.787,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 2.76,
|
| 2421 |
+
"grad_norm": 0.36844950914382935,
|
| 2422 |
+
"learning_rate": 1.9426294521477874e-07,
|
| 2423 |
+
"loss": 0.8194,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 2.768,
|
| 2428 |
+
"grad_norm": 0.4011365473270416,
|
| 2429 |
+
"learning_rate": 1.8160560742498223e-07,
|
| 2430 |
+
"loss": 0.7881,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 2.776,
|
| 2435 |
+
"grad_norm": 0.44703608751296997,
|
| 2436 |
+
"learning_rate": 1.6936700451975818e-07,
|
| 2437 |
+
"loss": 0.7567,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 2.784,
|
| 2442 |
+
"grad_norm": 0.39994892477989197,
|
| 2443 |
+
"learning_rate": 1.5754820007538473e-07,
|
| 2444 |
+
"loss": 0.7683,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 2.792,
|
| 2449 |
+
"grad_norm": 0.4645727872848511,
|
| 2450 |
+
"learning_rate": 1.461502211862237e-07,
|
| 2451 |
+
"loss": 0.7094,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 2.8,
|
| 2456 |
+
"grad_norm": 0.3852178156375885,
|
| 2457 |
+
"learning_rate": 1.3517405837546404e-07,
|
| 2458 |
+
"loss": 0.8263,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 2.808,
|
| 2463 |
+
"grad_norm": 0.4058782160282135,
|
| 2464 |
+
"learning_rate": 1.2462066550903818e-07,
|
| 2465 |
+
"loss": 0.7728,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 2.816,
|
| 2470 |
+
"grad_norm": 0.4285104274749756,
|
| 2471 |
+
"learning_rate": 1.1449095971273305e-07,
|
| 2472 |
+
"loss": 0.8048,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 2.824,
|
| 2477 |
+
"grad_norm": 0.39162999391555786,
|
| 2478 |
+
"learning_rate": 1.0478582129248516e-07,
|
| 2479 |
+
"loss": 0.8227,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 2.832,
|
| 2484 |
+
"grad_norm": 0.39359840750694275,
|
| 2485 |
+
"learning_rate": 9.550609365787888e-08,
|
| 2486 |
+
"loss": 0.7782,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 2.84,
|
| 2491 |
+
"grad_norm": 0.4036370813846588,
|
| 2492 |
+
"learning_rate": 8.66525832488535e-08,
|
| 2493 |
+
"loss": 0.7619,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 2.848,
|
| 2498 |
+
"grad_norm": 0.42931556701660156,
|
| 2499 |
+
"learning_rate": 7.822605946561923e-08,
|
| 2500 |
+
"loss": 0.7714,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 2.856,
|
| 2505 |
+
"grad_norm": 0.4127103090286255,
|
| 2506 |
+
"learning_rate": 7.022725460179459e-08,
|
| 2507 |
+
"loss": 0.7252,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 2.864,
|
| 2512 |
+
"grad_norm": 0.3982616364955902,
|
| 2513 |
+
"learning_rate": 6.265686378076729e-08,
|
| 2514 |
+
"loss": 0.8099,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 2.872,
|
| 2519 |
+
"grad_norm": 0.4003240168094635,
|
| 2520 |
+
"learning_rate": 5.5515544895284324e-08,
|
| 2521 |
+
"loss": 0.7717,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 2.88,
|
| 2526 |
+
"grad_norm": 0.3641519844532013,
|
| 2527 |
+
"learning_rate": 4.880391855028088e-08,
|
| 2528 |
+
"loss": 0.7815,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 2.888,
|
| 2533 |
+
"grad_norm": 0.3692556619644165,
|
| 2534 |
+
"learning_rate": 4.252256800894694e-08,
|
| 2535 |
+
"loss": 0.8232,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 2.896,
|
| 2540 |
+
"grad_norm": 0.3871900141239166,
|
| 2541 |
+
"learning_rate": 3.6672039142039426e-08,
|
| 2542 |
+
"loss": 0.7802,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 2.904,
|
| 2547 |
+
"grad_norm": 0.39646491408348083,
|
| 2548 |
+
"learning_rate": 3.125284038044407e-08,
|
| 2549 |
+
"loss": 0.7824,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 2.912,
|
| 2554 |
+
"grad_norm": 0.40282607078552246,
|
| 2555 |
+
"learning_rate": 2.6265442670991293e-08,
|
| 2556 |
+
"loss": 0.7715,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 2.92,
|
| 2561 |
+
"grad_norm": 0.4330500662326813,
|
| 2562 |
+
"learning_rate": 2.1710279435530058e-08,
|
| 2563 |
+
"loss": 0.7895,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 2.928,
|
| 2568 |
+
"grad_norm": 0.3595670461654663,
|
| 2569 |
+
"learning_rate": 1.7587746533260786e-08,
|
| 2570 |
+
"loss": 0.8252,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 2.936,
|
| 2575 |
+
"grad_norm": 0.41389673948287964,
|
| 2576 |
+
"learning_rate": 1.3898202226333424e-08,
|
| 2577 |
+
"loss": 0.7914,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 2.944,
|
| 2582 |
+
"grad_norm": 0.42384371161460876,
|
| 2583 |
+
"learning_rate": 1.0641967148716236e-08,
|
| 2584 |
+
"loss": 0.775,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 2.952,
|
| 2589 |
+
"grad_norm": 0.4105876088142395,
|
| 2590 |
+
"learning_rate": 7.819324278328099e-09,
|
| 2591 |
+
"loss": 0.8268,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 2.96,
|
| 2596 |
+
"grad_norm": 0.4283929467201233,
|
| 2597 |
+
"learning_rate": 5.430518912448169e-09,
|
| 2598 |
+
"loss": 0.785,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 2.968,
|
| 2603 |
+
"grad_norm": 0.4095473289489746,
|
| 2604 |
+
"learning_rate": 3.4757586464001513e-09,
|
| 2605 |
+
"loss": 0.7683,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 2.976,
|
| 2610 |
+
"grad_norm": 0.38301780819892883,
|
| 2611 |
+
"learning_rate": 1.9552133555084117e-09,
|
| 2612 |
+
"loss": 0.7837,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 2.984,
|
| 2617 |
+
"grad_norm": 0.4291246831417084,
|
| 2618 |
+
"learning_rate": 8.690151803386615e-10,
|
| 2619 |
+
"loss": 0.7784,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 2.992,
|
| 2624 |
+
"grad_norm": 0.4616396725177765,
|
| 2625 |
+
"learning_rate": 2.1725851521103847e-10,
|
| 2626 |
+
"loss": 0.753,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 3.0,
|
| 2631 |
+
"grad_norm": 0.4092578887939453,
|
| 2632 |
+
"learning_rate": 0.0,
|
| 2633 |
+
"loss": 0.7939,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 3.0,
|
| 2638 |
+
"step": 375,
|
| 2639 |
+
"total_flos": 372577161248768.0,
|
| 2640 |
+
"train_loss": 0.8553513642946879,
|
| 2641 |
+
"train_runtime": 20932.817,
|
| 2642 |
+
"train_samples_per_second": 1.72,
|
| 2643 |
+
"train_steps_per_second": 0.018
|
| 2644 |
+
}
|
| 2645 |
+
],
|
| 2646 |
+
"logging_steps": 1.0,
|
| 2647 |
+
"max_steps": 375,
|
| 2648 |
+
"num_input_tokens_seen": 0,
|
| 2649 |
+
"num_train_epochs": 3,
|
| 2650 |
+
"save_steps": 500,
|
| 2651 |
+
"stateful_callbacks": {
|
| 2652 |
+
"TrainerControl": {
|
| 2653 |
+
"args": {
|
| 2654 |
+
"should_epoch_stop": false,
|
| 2655 |
+
"should_evaluate": false,
|
| 2656 |
+
"should_log": false,
|
| 2657 |
+
"should_save": true,
|
| 2658 |
+
"should_training_stop": true
|
| 2659 |
+
},
|
| 2660 |
+
"attributes": {}
|
| 2661 |
+
}
|
| 2662 |
+
},
|
| 2663 |
+
"total_flos": 372577161248768.0,
|
| 2664 |
+
"train_batch_size": 1,
|
| 2665 |
+
"trial_name": null,
|
| 2666 |
+
"trial_params": null
|
| 2667 |
+
}
|
training_loss.png
ADDED
|