End of training
Browse files- README.md +2 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2205 -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_gpt_verification_scale_up_4K
|
|
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# seed_code_multiple_samples_gpt_verification_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_gpt_verification_scale_up_4K
|
|
|
|
| 16 |
|
| 17 |
# seed_code_multiple_samples_gpt_verification_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_gpt_verification_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": 266439779352576.0,
|
| 4 |
+
"train_loss": 0.8324689010586168,
|
| 5 |
+
"train_runtime": 15120.0259,
|
| 6 |
+
"train_samples_per_second": 1.962,
|
| 7 |
+
"train_steps_per_second": 0.02
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 266439779352576.0,
|
| 4 |
+
"train_loss": 0.8324689010586168,
|
| 5 |
+
"train_runtime": 15120.0259,
|
| 6 |
+
"train_samples_per_second": 1.962,
|
| 7 |
+
"train_steps_per_second": 0.02
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2205 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 309,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.009708737864077669,
|
| 13 |
+
"grad_norm": 6.576948642730713,
|
| 14 |
+
"learning_rate": 3.2258064516129035e-07,
|
| 15 |
+
"loss": 1.2227,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.019417475728155338,
|
| 20 |
+
"grad_norm": 6.762261867523193,
|
| 21 |
+
"learning_rate": 6.451612903225807e-07,
|
| 22 |
+
"loss": 1.2675,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02912621359223301,
|
| 27 |
+
"grad_norm": 6.975041389465332,
|
| 28 |
+
"learning_rate": 9.67741935483871e-07,
|
| 29 |
+
"loss": 1.2517,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.038834951456310676,
|
| 34 |
+
"grad_norm": 6.626192569732666,
|
| 35 |
+
"learning_rate": 1.2903225806451614e-06,
|
| 36 |
+
"loss": 1.2515,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04854368932038835,
|
| 41 |
+
"grad_norm": 6.378674030303955,
|
| 42 |
+
"learning_rate": 1.6129032258064516e-06,
|
| 43 |
+
"loss": 1.2498,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05825242718446602,
|
| 48 |
+
"grad_norm": 6.095025539398193,
|
| 49 |
+
"learning_rate": 1.935483870967742e-06,
|
| 50 |
+
"loss": 1.2283,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06796116504854369,
|
| 55 |
+
"grad_norm": 4.670423984527588,
|
| 56 |
+
"learning_rate": 2.2580645161290324e-06,
|
| 57 |
+
"loss": 1.1489,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.07766990291262135,
|
| 62 |
+
"grad_norm": 4.401129245758057,
|
| 63 |
+
"learning_rate": 2.580645161290323e-06,
|
| 64 |
+
"loss": 1.1759,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.08737864077669903,
|
| 69 |
+
"grad_norm": 2.814140796661377,
|
| 70 |
+
"learning_rate": 2.903225806451613e-06,
|
| 71 |
+
"loss": 1.128,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.0970873786407767,
|
| 76 |
+
"grad_norm": 2.6264705657958984,
|
| 77 |
+
"learning_rate": 3.225806451612903e-06,
|
| 78 |
+
"loss": 1.1088,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.10679611650485436,
|
| 83 |
+
"grad_norm": 2.619597911834717,
|
| 84 |
+
"learning_rate": 3.548387096774194e-06,
|
| 85 |
+
"loss": 1.1275,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.11650485436893204,
|
| 90 |
+
"grad_norm": 4.305774688720703,
|
| 91 |
+
"learning_rate": 3.870967741935484e-06,
|
| 92 |
+
"loss": 1.1287,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1262135922330097,
|
| 97 |
+
"grad_norm": 4.335093975067139,
|
| 98 |
+
"learning_rate": 4.193548387096774e-06,
|
| 99 |
+
"loss": 1.1011,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.13592233009708737,
|
| 104 |
+
"grad_norm": 4.46498966217041,
|
| 105 |
+
"learning_rate": 4.516129032258065e-06,
|
| 106 |
+
"loss": 1.1392,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.14563106796116504,
|
| 111 |
+
"grad_norm": 3.7746875286102295,
|
| 112 |
+
"learning_rate": 4.838709677419355e-06,
|
| 113 |
+
"loss": 1.0678,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.1553398058252427,
|
| 118 |
+
"grad_norm": 2.9575371742248535,
|
| 119 |
+
"learning_rate": 5.161290322580646e-06,
|
| 120 |
+
"loss": 1.0683,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1650485436893204,
|
| 125 |
+
"grad_norm": 2.7806265354156494,
|
| 126 |
+
"learning_rate": 5.483870967741935e-06,
|
| 127 |
+
"loss": 1.0367,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.17475728155339806,
|
| 132 |
+
"grad_norm": 2.41747784614563,
|
| 133 |
+
"learning_rate": 5.806451612903226e-06,
|
| 134 |
+
"loss": 1.0401,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.18446601941747573,
|
| 139 |
+
"grad_norm": 2.0261120796203613,
|
| 140 |
+
"learning_rate": 6.129032258064517e-06,
|
| 141 |
+
"loss": 1.0364,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.1941747572815534,
|
| 146 |
+
"grad_norm": 1.6902999877929688,
|
| 147 |
+
"learning_rate": 6.451612903225806e-06,
|
| 148 |
+
"loss": 0.9807,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.20388349514563106,
|
| 153 |
+
"grad_norm": 1.6579384803771973,
|
| 154 |
+
"learning_rate": 6.774193548387097e-06,
|
| 155 |
+
"loss": 1.0094,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.21359223300970873,
|
| 160 |
+
"grad_norm": 1.3350183963775635,
|
| 161 |
+
"learning_rate": 7.096774193548388e-06,
|
| 162 |
+
"loss": 0.9756,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.22330097087378642,
|
| 167 |
+
"grad_norm": 1.4252748489379883,
|
| 168 |
+
"learning_rate": 7.4193548387096784e-06,
|
| 169 |
+
"loss": 1.0005,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.23300970873786409,
|
| 174 |
+
"grad_norm": 1.2831058502197266,
|
| 175 |
+
"learning_rate": 7.741935483870968e-06,
|
| 176 |
+
"loss": 0.9875,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.24271844660194175,
|
| 181 |
+
"grad_norm": 1.1137187480926514,
|
| 182 |
+
"learning_rate": 8.064516129032258e-06,
|
| 183 |
+
"loss": 0.9444,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2524271844660194,
|
| 188 |
+
"grad_norm": 1.186092495918274,
|
| 189 |
+
"learning_rate": 8.387096774193549e-06,
|
| 190 |
+
"loss": 0.9467,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.2621359223300971,
|
| 195 |
+
"grad_norm": 0.910400390625,
|
| 196 |
+
"learning_rate": 8.70967741935484e-06,
|
| 197 |
+
"loss": 0.9549,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.27184466019417475,
|
| 202 |
+
"grad_norm": 1.0872230529785156,
|
| 203 |
+
"learning_rate": 9.03225806451613e-06,
|
| 204 |
+
"loss": 0.9433,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2815533980582524,
|
| 209 |
+
"grad_norm": 1.0862929821014404,
|
| 210 |
+
"learning_rate": 9.35483870967742e-06,
|
| 211 |
+
"loss": 0.9353,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2912621359223301,
|
| 216 |
+
"grad_norm": 1.1340218782424927,
|
| 217 |
+
"learning_rate": 9.67741935483871e-06,
|
| 218 |
+
"loss": 0.9254,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.30097087378640774,
|
| 223 |
+
"grad_norm": 1.1494905948638916,
|
| 224 |
+
"learning_rate": 1e-05,
|
| 225 |
+
"loss": 0.9231,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.3106796116504854,
|
| 230 |
+
"grad_norm": 0.9529262781143188,
|
| 231 |
+
"learning_rate": 9.999680739242022e-06,
|
| 232 |
+
"loss": 0.9367,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.32038834951456313,
|
| 237 |
+
"grad_norm": 0.9129893183708191,
|
| 238 |
+
"learning_rate": 9.99872299773906e-06,
|
| 239 |
+
"loss": 0.9194,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.3300970873786408,
|
| 244 |
+
"grad_norm": 1.1129016876220703,
|
| 245 |
+
"learning_rate": 9.997126897798826e-06,
|
| 246 |
+
"loss": 0.8978,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.33980582524271846,
|
| 251 |
+
"grad_norm": 1.0057542324066162,
|
| 252 |
+
"learning_rate": 9.994892643250147e-06,
|
| 253 |
+
"loss": 0.9046,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.34951456310679613,
|
| 258 |
+
"grad_norm": 0.8979021906852722,
|
| 259 |
+
"learning_rate": 9.99202051941695e-06,
|
| 260 |
+
"loss": 0.9191,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.3592233009708738,
|
| 265 |
+
"grad_norm": 0.8347776532173157,
|
| 266 |
+
"learning_rate": 9.9885108930818e-06,
|
| 267 |
+
"loss": 0.8796,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.36893203883495146,
|
| 272 |
+
"grad_norm": 0.8598542213439941,
|
| 273 |
+
"learning_rate": 9.984364212439089e-06,
|
| 274 |
+
"loss": 0.9132,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3786407766990291,
|
| 279 |
+
"grad_norm": 0.8244529962539673,
|
| 280 |
+
"learning_rate": 9.979581007037776e-06,
|
| 281 |
+
"loss": 0.8933,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.3883495145631068,
|
| 286 |
+
"grad_norm": 0.733172595500946,
|
| 287 |
+
"learning_rate": 9.974161887713775e-06,
|
| 288 |
+
"loss": 0.9021,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.39805825242718446,
|
| 293 |
+
"grad_norm": 0.9536871910095215,
|
| 294 |
+
"learning_rate": 9.968107546511942e-06,
|
| 295 |
+
"loss": 0.9121,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.4077669902912621,
|
| 300 |
+
"grad_norm": 0.7781872153282166,
|
| 301 |
+
"learning_rate": 9.961418756597703e-06,
|
| 302 |
+
"loss": 0.9045,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.4174757281553398,
|
| 307 |
+
"grad_norm": 0.8622099757194519,
|
| 308 |
+
"learning_rate": 9.95409637215831e-06,
|
| 309 |
+
"loss": 0.9019,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.42718446601941745,
|
| 314 |
+
"grad_norm": 0.694644570350647,
|
| 315 |
+
"learning_rate": 9.94614132829377e-06,
|
| 316 |
+
"loss": 0.8502,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.4368932038834951,
|
| 321 |
+
"grad_norm": 0.7023694515228271,
|
| 322 |
+
"learning_rate": 9.937554640897414e-06,
|
| 323 |
+
"loss": 0.9056,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.44660194174757284,
|
| 328 |
+
"grad_norm": 0.7865616679191589,
|
| 329 |
+
"learning_rate": 9.928337406526172e-06,
|
| 330 |
+
"loss": 0.8951,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.4563106796116505,
|
| 335 |
+
"grad_norm": 0.8198397159576416,
|
| 336 |
+
"learning_rate": 9.918490802260538e-06,
|
| 337 |
+
"loss": 0.919,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.46601941747572817,
|
| 342 |
+
"grad_norm": 0.6332915425300598,
|
| 343 |
+
"learning_rate": 9.90801608555425e-06,
|
| 344 |
+
"loss": 0.9086,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.47572815533980584,
|
| 349 |
+
"grad_norm": 0.665955662727356,
|
| 350 |
+
"learning_rate": 9.896914594073703e-06,
|
| 351 |
+
"loss": 0.8707,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4854368932038835,
|
| 356 |
+
"grad_norm": 0.8108275532722473,
|
| 357 |
+
"learning_rate": 9.885187745527132e-06,
|
| 358 |
+
"loss": 0.8903,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.49514563106796117,
|
| 363 |
+
"grad_norm": 0.7815718054771423,
|
| 364 |
+
"learning_rate": 9.87283703748356e-06,
|
| 365 |
+
"loss": 0.8803,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5048543689320388,
|
| 370 |
+
"grad_norm": 0.7477515935897827,
|
| 371 |
+
"learning_rate": 9.859864047181551e-06,
|
| 372 |
+
"loss": 0.9034,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.5145631067961165,
|
| 377 |
+
"grad_norm": 0.8046391606330872,
|
| 378 |
+
"learning_rate": 9.846270431327793e-06,
|
| 379 |
+
"loss": 0.8962,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.5242718446601942,
|
| 384 |
+
"grad_norm": 0.6419571042060852,
|
| 385 |
+
"learning_rate": 9.832057925885526e-06,
|
| 386 |
+
"loss": 0.8834,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5339805825242718,
|
| 391 |
+
"grad_norm": 0.885932981967926,
|
| 392 |
+
"learning_rate": 9.817228345852853e-06,
|
| 393 |
+
"loss": 0.9023,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5436893203883495,
|
| 398 |
+
"grad_norm": 0.5639209747314453,
|
| 399 |
+
"learning_rate": 9.801783585030959e-06,
|
| 400 |
+
"loss": 0.9119,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.5533980582524272,
|
| 405 |
+
"grad_norm": 0.8406121730804443,
|
| 406 |
+
"learning_rate": 9.785725615782262e-06,
|
| 407 |
+
"loss": 0.8605,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5631067961165048,
|
| 412 |
+
"grad_norm": 0.6397046446800232,
|
| 413 |
+
"learning_rate": 9.769056488778538e-06,
|
| 414 |
+
"loss": 0.87,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5728155339805825,
|
| 419 |
+
"grad_norm": 0.6496418714523315,
|
| 420 |
+
"learning_rate": 9.751778332739033e-06,
|
| 421 |
+
"loss": 0.8863,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5825242718446602,
|
| 426 |
+
"grad_norm": 0.8817011713981628,
|
| 427 |
+
"learning_rate": 9.733893354158628e-06,
|
| 428 |
+
"loss": 0.8757,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.5922330097087378,
|
| 433 |
+
"grad_norm": 0.6433860063552856,
|
| 434 |
+
"learning_rate": 9.715403837026046e-06,
|
| 435 |
+
"loss": 0.8613,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.6019417475728155,
|
| 440 |
+
"grad_norm": 0.9720786213874817,
|
| 441 |
+
"learning_rate": 9.69631214253219e-06,
|
| 442 |
+
"loss": 0.8831,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.6116504854368932,
|
| 447 |
+
"grad_norm": 0.6686399579048157,
|
| 448 |
+
"learning_rate": 9.676620708768608e-06,
|
| 449 |
+
"loss": 0.8823,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6213592233009708,
|
| 454 |
+
"grad_norm": 0.6597081422805786,
|
| 455 |
+
"learning_rate": 9.656332050416118e-06,
|
| 456 |
+
"loss": 0.8351,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.6310679611650486,
|
| 461 |
+
"grad_norm": 0.781461775302887,
|
| 462 |
+
"learning_rate": 9.635448758423703e-06,
|
| 463 |
+
"loss": 0.8792,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.6407766990291263,
|
| 468 |
+
"grad_norm": 0.5933522582054138,
|
| 469 |
+
"learning_rate": 9.613973499677613e-06,
|
| 470 |
+
"loss": 0.912,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.6504854368932039,
|
| 475 |
+
"grad_norm": 0.7272161841392517,
|
| 476 |
+
"learning_rate": 9.591909016660806e-06,
|
| 477 |
+
"loss": 0.8788,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.6601941747572816,
|
| 482 |
+
"grad_norm": 0.6520808339118958,
|
| 483 |
+
"learning_rate": 9.569258127102708e-06,
|
| 484 |
+
"loss": 0.8965,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6699029126213593,
|
| 489 |
+
"grad_norm": 0.7806766033172607,
|
| 490 |
+
"learning_rate": 9.546023723619387e-06,
|
| 491 |
+
"loss": 0.8917,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.6796116504854369,
|
| 496 |
+
"grad_norm": 0.7100508213043213,
|
| 497 |
+
"learning_rate": 9.522208773344147e-06,
|
| 498 |
+
"loss": 0.8872,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.6893203883495146,
|
| 503 |
+
"grad_norm": 0.7099691033363342,
|
| 504 |
+
"learning_rate": 9.497816317548625e-06,
|
| 505 |
+
"loss": 0.8793,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.6990291262135923,
|
| 510 |
+
"grad_norm": 0.8257980346679688,
|
| 511 |
+
"learning_rate": 9.472849471254386e-06,
|
| 512 |
+
"loss": 0.8539,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.7087378640776699,
|
| 517 |
+
"grad_norm": 0.6728125214576721,
|
| 518 |
+
"learning_rate": 9.447311422835141e-06,
|
| 519 |
+
"loss": 0.8669,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.7184466019417476,
|
| 524 |
+
"grad_norm": 0.7229968905448914,
|
| 525 |
+
"learning_rate": 9.421205433609568e-06,
|
| 526 |
+
"loss": 0.8991,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.7281553398058253,
|
| 531 |
+
"grad_norm": 0.7105078101158142,
|
| 532 |
+
"learning_rate": 9.39453483742483e-06,
|
| 533 |
+
"loss": 0.8704,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.7378640776699029,
|
| 538 |
+
"grad_norm": 0.7349528670310974,
|
| 539 |
+
"learning_rate": 9.367303040230828e-06,
|
| 540 |
+
"loss": 0.8866,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.7475728155339806,
|
| 545 |
+
"grad_norm": 0.5864897966384888,
|
| 546 |
+
"learning_rate": 9.33951351964525e-06,
|
| 547 |
+
"loss": 0.8671,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.7572815533980582,
|
| 552 |
+
"grad_norm": 0.6392523646354675,
|
| 553 |
+
"learning_rate": 9.311169824509454e-06,
|
| 554 |
+
"loss": 0.8435,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.7669902912621359,
|
| 559 |
+
"grad_norm": 0.5775032639503479,
|
| 560 |
+
"learning_rate": 9.28227557443528e-06,
|
| 561 |
+
"loss": 0.8518,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.7766990291262136,
|
| 566 |
+
"grad_norm": 0.5982815623283386,
|
| 567 |
+
"learning_rate": 9.252834459342801e-06,
|
| 568 |
+
"loss": 0.9034,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.7864077669902912,
|
| 573 |
+
"grad_norm": 0.6583596467971802,
|
| 574 |
+
"learning_rate": 9.222850238989104e-06,
|
| 575 |
+
"loss": 0.8638,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.7961165048543689,
|
| 580 |
+
"grad_norm": 0.621867299079895,
|
| 581 |
+
"learning_rate": 9.192326742488153e-06,
|
| 582 |
+
"loss": 0.8658,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.8058252427184466,
|
| 587 |
+
"grad_norm": 0.6219890117645264,
|
| 588 |
+
"learning_rate": 9.161267867821802e-06,
|
| 589 |
+
"loss": 0.8769,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.8155339805825242,
|
| 594 |
+
"grad_norm": 0.6069071292877197,
|
| 595 |
+
"learning_rate": 9.129677581342e-06,
|
| 596 |
+
"loss": 0.8677,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.8252427184466019,
|
| 601 |
+
"grad_norm": 0.660553514957428,
|
| 602 |
+
"learning_rate": 9.097559917264268e-06,
|
| 603 |
+
"loss": 0.8465,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.8349514563106796,
|
| 608 |
+
"grad_norm": 0.5944454669952393,
|
| 609 |
+
"learning_rate": 9.064918977152517e-06,
|
| 610 |
+
"loss": 0.8741,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.8446601941747572,
|
| 615 |
+
"grad_norm": 0.6104128360748291,
|
| 616 |
+
"learning_rate": 9.031758929395259e-06,
|
| 617 |
+
"loss": 0.8248,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.8543689320388349,
|
| 622 |
+
"grad_norm": 0.5879605412483215,
|
| 623 |
+
"learning_rate": 8.998084008673284e-06,
|
| 624 |
+
"loss": 0.8398,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.8640776699029126,
|
| 629 |
+
"grad_norm": 0.6651065349578857,
|
| 630 |
+
"learning_rate": 8.963898515418885e-06,
|
| 631 |
+
"loss": 0.8687,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.8737864077669902,
|
| 636 |
+
"grad_norm": 0.6086816191673279,
|
| 637 |
+
"learning_rate": 8.929206815266653e-06,
|
| 638 |
+
"loss": 0.8639,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.883495145631068,
|
| 643 |
+
"grad_norm": 0.6449173092842102,
|
| 644 |
+
"learning_rate": 8.89401333849598e-06,
|
| 645 |
+
"loss": 0.8483,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.8932038834951457,
|
| 650 |
+
"grad_norm": 0.6413861513137817,
|
| 651 |
+
"learning_rate": 8.8583225794653e-06,
|
| 652 |
+
"loss": 0.9039,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.9029126213592233,
|
| 657 |
+
"grad_norm": 0.5411512851715088,
|
| 658 |
+
"learning_rate": 8.82213909603812e-06,
|
| 659 |
+
"loss": 0.8814,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.912621359223301,
|
| 664 |
+
"grad_norm": 0.6556549072265625,
|
| 665 |
+
"learning_rate": 8.78546750900098e-06,
|
| 666 |
+
"loss": 0.8944,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.9223300970873787,
|
| 671 |
+
"grad_norm": 0.6550806760787964,
|
| 672 |
+
"learning_rate": 8.748312501473351e-06,
|
| 673 |
+
"loss": 0.8819,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.9320388349514563,
|
| 678 |
+
"grad_norm": 0.6445295810699463,
|
| 679 |
+
"learning_rate": 8.710678818309576e-06,
|
| 680 |
+
"loss": 0.8582,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.941747572815534,
|
| 685 |
+
"grad_norm": 0.7256100177764893,
|
| 686 |
+
"learning_rate": 8.672571265492944e-06,
|
| 687 |
+
"loss": 0.8361,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.9514563106796117,
|
| 692 |
+
"grad_norm": 0.599454402923584,
|
| 693 |
+
"learning_rate": 8.63399470952193e-06,
|
| 694 |
+
"loss": 0.8679,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.9611650485436893,
|
| 699 |
+
"grad_norm": 0.7651983499526978,
|
| 700 |
+
"learning_rate": 8.594954076788736e-06,
|
| 701 |
+
"loss": 0.8182,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.970873786407767,
|
| 706 |
+
"grad_norm": 0.7521474361419678,
|
| 707 |
+
"learning_rate": 8.555454352950161e-06,
|
| 708 |
+
"loss": 0.8622,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.9805825242718447,
|
| 713 |
+
"grad_norm": 0.6626675724983215,
|
| 714 |
+
"learning_rate": 8.515500582290914e-06,
|
| 715 |
+
"loss": 0.845,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.9902912621359223,
|
| 720 |
+
"grad_norm": 0.7361329793930054,
|
| 721 |
+
"learning_rate": 8.475097867079437e-06,
|
| 722 |
+
"loss": 0.8672,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.0,
|
| 727 |
+
"grad_norm": 0.5690233707427979,
|
| 728 |
+
"learning_rate": 8.434251366916323e-06,
|
| 729 |
+
"loss": 0.8989,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.0097087378640777,
|
| 734 |
+
"grad_norm": 0.778568685054779,
|
| 735 |
+
"learning_rate": 8.392966298075413e-06,
|
| 736 |
+
"loss": 0.8642,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.0194174757281553,
|
| 741 |
+
"grad_norm": 0.5900771021842957,
|
| 742 |
+
"learning_rate": 8.351247932837655e-06,
|
| 743 |
+
"loss": 0.8001,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.029126213592233,
|
| 748 |
+
"grad_norm": 0.628034770488739,
|
| 749 |
+
"learning_rate": 8.309101598817812e-06,
|
| 750 |
+
"loss": 0.8101,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.0388349514563107,
|
| 755 |
+
"grad_norm": 0.860468864440918,
|
| 756 |
+
"learning_rate": 8.266532678284103e-06,
|
| 757 |
+
"loss": 0.7936,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.0485436893203883,
|
| 762 |
+
"grad_norm": 0.7311742901802063,
|
| 763 |
+
"learning_rate": 8.223546607470863e-06,
|
| 764 |
+
"loss": 0.8278,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.058252427184466,
|
| 769 |
+
"grad_norm": 0.6832053661346436,
|
| 770 |
+
"learning_rate": 8.18014887588431e-06,
|
| 771 |
+
"loss": 0.8272,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.0679611650485437,
|
| 776 |
+
"grad_norm": 0.8355731964111328,
|
| 777 |
+
"learning_rate": 8.13634502560152e-06,
|
| 778 |
+
"loss": 0.8485,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.0776699029126213,
|
| 783 |
+
"grad_norm": 0.6029130816459656,
|
| 784 |
+
"learning_rate": 8.092140650562665e-06,
|
| 785 |
+
"loss": 0.7722,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.087378640776699,
|
| 790 |
+
"grad_norm": 0.8511623740196228,
|
| 791 |
+
"learning_rate": 8.047541395856661e-06,
|
| 792 |
+
"loss": 0.8361,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.0970873786407767,
|
| 797 |
+
"grad_norm": 0.6003599166870117,
|
| 798 |
+
"learning_rate": 8.002552957000254e-06,
|
| 799 |
+
"loss": 0.8157,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.1067961165048543,
|
| 804 |
+
"grad_norm": 0.8679140210151672,
|
| 805 |
+
"learning_rate": 7.957181079210676e-06,
|
| 806 |
+
"loss": 0.819,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.116504854368932,
|
| 811 |
+
"grad_norm": 0.65385901927948,
|
| 812 |
+
"learning_rate": 7.911431556671967e-06,
|
| 813 |
+
"loss": 0.8191,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.1262135922330097,
|
| 818 |
+
"grad_norm": 0.5810941457748413,
|
| 819 |
+
"learning_rate": 7.865310231795026e-06,
|
| 820 |
+
"loss": 0.8084,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.1359223300970873,
|
| 825 |
+
"grad_norm": 0.7386210560798645,
|
| 826 |
+
"learning_rate": 7.818822994471504e-06,
|
| 827 |
+
"loss": 0.8088,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.145631067961165,
|
| 832 |
+
"grad_norm": 0.6052452325820923,
|
| 833 |
+
"learning_rate": 7.771975781321655e-06,
|
| 834 |
+
"loss": 0.8236,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.1553398058252426,
|
| 839 |
+
"grad_norm": 0.6413220167160034,
|
| 840 |
+
"learning_rate": 7.72477457493619e-06,
|
| 841 |
+
"loss": 0.8427,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.1650485436893203,
|
| 846 |
+
"grad_norm": 0.6168550848960876,
|
| 847 |
+
"learning_rate": 7.677225403112277e-06,
|
| 848 |
+
"loss": 0.8204,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.174757281553398,
|
| 853 |
+
"grad_norm": 0.6849708557128906,
|
| 854 |
+
"learning_rate": 7.629334338083774e-06,
|
| 855 |
+
"loss": 0.7887,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.1844660194174756,
|
| 860 |
+
"grad_norm": 0.7301358580589294,
|
| 861 |
+
"learning_rate": 7.58110749574577e-06,
|
| 862 |
+
"loss": 0.8137,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.1941747572815533,
|
| 867 |
+
"grad_norm": 0.6189250349998474,
|
| 868 |
+
"learning_rate": 7.532551034873558e-06,
|
| 869 |
+
"loss": 0.7985,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.203883495145631,
|
| 874 |
+
"grad_norm": 0.70215904712677,
|
| 875 |
+
"learning_rate": 7.483671156336142e-06,
|
| 876 |
+
"loss": 0.8108,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.2135922330097086,
|
| 881 |
+
"grad_norm": 0.642704963684082,
|
| 882 |
+
"learning_rate": 7.43447410230435e-06,
|
| 883 |
+
"loss": 0.8161,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.2233009708737863,
|
| 888 |
+
"grad_norm": 0.6741234660148621,
|
| 889 |
+
"learning_rate": 7.384966155453686e-06,
|
| 890 |
+
"loss": 0.7965,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.233009708737864,
|
| 895 |
+
"grad_norm": 0.577409029006958,
|
| 896 |
+
"learning_rate": 7.335153638162005e-06,
|
| 897 |
+
"loss": 0.7989,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.2427184466019416,
|
| 902 |
+
"grad_norm": 0.5523732900619507,
|
| 903 |
+
"learning_rate": 7.285042911702116e-06,
|
| 904 |
+
"loss": 0.7812,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.2524271844660193,
|
| 909 |
+
"grad_norm": 0.6712595224380493,
|
| 910 |
+
"learning_rate": 7.234640375429427e-06,
|
| 911 |
+
"loss": 0.8034,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.262135922330097,
|
| 916 |
+
"grad_norm": 0.6166908144950867,
|
| 917 |
+
"learning_rate": 7.183952465964711e-06,
|
| 918 |
+
"loss": 0.8004,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.2718446601941746,
|
| 923 |
+
"grad_norm": 0.7181272506713867,
|
| 924 |
+
"learning_rate": 7.132985656372126e-06,
|
| 925 |
+
"loss": 0.8122,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.2815533980582523,
|
| 930 |
+
"grad_norm": 0.5397993922233582,
|
| 931 |
+
"learning_rate": 7.0817464553325764e-06,
|
| 932 |
+
"loss": 0.8256,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.29126213592233,
|
| 937 |
+
"grad_norm": 0.599078357219696,
|
| 938 |
+
"learning_rate": 7.030241406312528e-06,
|
| 939 |
+
"loss": 0.8012,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.3009708737864076,
|
| 944 |
+
"grad_norm": 0.6068196296691895,
|
| 945 |
+
"learning_rate": 6.978477086728375e-06,
|
| 946 |
+
"loss": 0.7475,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.3106796116504853,
|
| 951 |
+
"grad_norm": 0.6208373308181763,
|
| 952 |
+
"learning_rate": 6.926460107106483e-06,
|
| 953 |
+
"loss": 0.7631,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.3203883495145632,
|
| 958 |
+
"grad_norm": 0.62087482213974,
|
| 959 |
+
"learning_rate": 6.874197110238986e-06,
|
| 960 |
+
"loss": 0.7626,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.3300970873786409,
|
| 965 |
+
"grad_norm": 0.6587031483650208,
|
| 966 |
+
"learning_rate": 6.8216947703354815e-06,
|
| 967 |
+
"loss": 0.809,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.3398058252427185,
|
| 972 |
+
"grad_norm": 0.7958360314369202,
|
| 973 |
+
"learning_rate": 6.7689597921707065e-06,
|
| 974 |
+
"loss": 0.8311,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.3495145631067962,
|
| 979 |
+
"grad_norm": 0.6715186834335327,
|
| 980 |
+
"learning_rate": 6.715998910228296e-06,
|
| 981 |
+
"loss": 0.7949,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.3592233009708738,
|
| 986 |
+
"grad_norm": 0.6351740956306458,
|
| 987 |
+
"learning_rate": 6.6628188878407806e-06,
|
| 988 |
+
"loss": 0.794,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.3689320388349515,
|
| 993 |
+
"grad_norm": 0.8707325458526611,
|
| 994 |
+
"learning_rate": 6.609426516325859e-06,
|
| 995 |
+
"loss": 0.8329,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.3786407766990292,
|
| 1000 |
+
"grad_norm": 0.7321479320526123,
|
| 1001 |
+
"learning_rate": 6.555828614119132e-06,
|
| 1002 |
+
"loss": 0.8136,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.3883495145631068,
|
| 1007 |
+
"grad_norm": 0.5844182968139648,
|
| 1008 |
+
"learning_rate": 6.502032025903356e-06,
|
| 1009 |
+
"loss": 0.7576,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.3980582524271845,
|
| 1014 |
+
"grad_norm": 0.6121401786804199,
|
| 1015 |
+
"learning_rate": 6.4480436217343366e-06,
|
| 1016 |
+
"loss": 0.8205,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.4077669902912622,
|
| 1021 |
+
"grad_norm": 0.6373964548110962,
|
| 1022 |
+
"learning_rate": 6.393870296163616e-06,
|
| 1023 |
+
"loss": 0.8104,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.4174757281553398,
|
| 1028 |
+
"grad_norm": 0.5534393787384033,
|
| 1029 |
+
"learning_rate": 6.339518967357985e-06,
|
| 1030 |
+
"loss": 0.7822,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.4271844660194175,
|
| 1035 |
+
"grad_norm": 0.593232274055481,
|
| 1036 |
+
"learning_rate": 6.284996576216014e-06,
|
| 1037 |
+
"loss": 0.8227,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.4368932038834952,
|
| 1042 |
+
"grad_norm": 0.5890830755233765,
|
| 1043 |
+
"learning_rate": 6.230310085481677e-06,
|
| 1044 |
+
"loss": 0.8005,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.4466019417475728,
|
| 1049 |
+
"grad_norm": 0.5876540541648865,
|
| 1050 |
+
"learning_rate": 6.175466478855161e-06,
|
| 1051 |
+
"loss": 0.7936,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.4563106796116505,
|
| 1056 |
+
"grad_norm": 0.5593201518058777,
|
| 1057 |
+
"learning_rate": 6.1204727601010396e-06,
|
| 1058 |
+
"loss": 0.7473,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.4660194174757282,
|
| 1063 |
+
"grad_norm": 0.5841484069824219,
|
| 1064 |
+
"learning_rate": 6.065335952153846e-06,
|
| 1065 |
+
"loss": 0.7974,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.4757281553398058,
|
| 1070 |
+
"grad_norm": 0.49436667561531067,
|
| 1071 |
+
"learning_rate": 6.010063096221215e-06,
|
| 1072 |
+
"loss": 0.7921,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.4854368932038835,
|
| 1077 |
+
"grad_norm": 0.6187512874603271,
|
| 1078 |
+
"learning_rate": 5.954661250884704e-06,
|
| 1079 |
+
"loss": 0.7822,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.4951456310679612,
|
| 1084 |
+
"grad_norm": 0.6759915351867676,
|
| 1085 |
+
"learning_rate": 5.899137491198364e-06,
|
| 1086 |
+
"loss": 0.8273,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.5048543689320388,
|
| 1091 |
+
"grad_norm": 0.5890839099884033,
|
| 1092 |
+
"learning_rate": 5.843498907785236e-06,
|
| 1093 |
+
"loss": 0.7979,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.5145631067961165,
|
| 1098 |
+
"grad_norm": 0.5246779322624207,
|
| 1099 |
+
"learning_rate": 5.78775260593185e-06,
|
| 1100 |
+
"loss": 0.8198,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.5242718446601942,
|
| 1105 |
+
"grad_norm": 0.5603223443031311,
|
| 1106 |
+
"learning_rate": 5.731905704680834e-06,
|
| 1107 |
+
"loss": 0.8023,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.5339805825242718,
|
| 1112 |
+
"grad_norm": 0.5730907917022705,
|
| 1113 |
+
"learning_rate": 5.6759653359218e-06,
|
| 1114 |
+
"loss": 0.8231,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.5436893203883495,
|
| 1119 |
+
"grad_norm": 0.5317082405090332,
|
| 1120 |
+
"learning_rate": 5.6199386434805615e-06,
|
| 1121 |
+
"loss": 0.7984,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.5533980582524272,
|
| 1126 |
+
"grad_norm": 0.5330399870872498,
|
| 1127 |
+
"learning_rate": 5.563832782206835e-06,
|
| 1128 |
+
"loss": 0.7975,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.5631067961165048,
|
| 1133 |
+
"grad_norm": 0.6568978428840637,
|
| 1134 |
+
"learning_rate": 5.507654917060541e-06,
|
| 1135 |
+
"loss": 0.824,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.5728155339805825,
|
| 1140 |
+
"grad_norm": 0.6758802533149719,
|
| 1141 |
+
"learning_rate": 5.451412222196801e-06,
|
| 1142 |
+
"loss": 0.7844,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.5825242718446602,
|
| 1147 |
+
"grad_norm": 0.5091277360916138,
|
| 1148 |
+
"learning_rate": 5.395111880049775e-06,
|
| 1149 |
+
"loss": 0.7946,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.5922330097087378,
|
| 1154 |
+
"grad_norm": 0.5247997045516968,
|
| 1155 |
+
"learning_rate": 5.338761080415425e-06,
|
| 1156 |
+
"loss": 0.8035,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.6019417475728155,
|
| 1161 |
+
"grad_norm": 0.5019735097885132,
|
| 1162 |
+
"learning_rate": 5.28236701953335e-06,
|
| 1163 |
+
"loss": 0.7802,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.6116504854368932,
|
| 1168 |
+
"grad_norm": 0.5382864475250244,
|
| 1169 |
+
"learning_rate": 5.225936899167803e-06,
|
| 1170 |
+
"loss": 0.8301,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.6213592233009708,
|
| 1175 |
+
"grad_norm": 0.503472089767456,
|
| 1176 |
+
"learning_rate": 5.169477925687981e-06,
|
| 1177 |
+
"loss": 0.7875,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.6310679611650487,
|
| 1182 |
+
"grad_norm": 0.5378626585006714,
|
| 1183 |
+
"learning_rate": 5.112997309147753e-06,
|
| 1184 |
+
"loss": 0.8003,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.6407766990291264,
|
| 1189 |
+
"grad_norm": 0.4998520314693451,
|
| 1190 |
+
"learning_rate": 5.0565022623649e-06,
|
| 1191 |
+
"loss": 0.8094,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.650485436893204,
|
| 1196 |
+
"grad_norm": 0.5108544826507568,
|
| 1197 |
+
"learning_rate": 5e-06,
|
| 1198 |
+
"loss": 0.7756,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.6601941747572817,
|
| 1203 |
+
"grad_norm": 0.4933609962463379,
|
| 1204 |
+
"learning_rate": 4.943497737635103e-06,
|
| 1205 |
+
"loss": 0.8081,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.6699029126213594,
|
| 1210 |
+
"grad_norm": 0.5090092420578003,
|
| 1211 |
+
"learning_rate": 4.887002690852249e-06,
|
| 1212 |
+
"loss": 0.8025,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.679611650485437,
|
| 1217 |
+
"grad_norm": 0.5769246816635132,
|
| 1218 |
+
"learning_rate": 4.830522074312019e-06,
|
| 1219 |
+
"loss": 0.8068,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.6893203883495147,
|
| 1224 |
+
"grad_norm": 0.5017176866531372,
|
| 1225 |
+
"learning_rate": 4.774063100832199e-06,
|
| 1226 |
+
"loss": 0.8078,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.6990291262135924,
|
| 1231 |
+
"grad_norm": 0.5930182337760925,
|
| 1232 |
+
"learning_rate": 4.717632980466652e-06,
|
| 1233 |
+
"loss": 0.8529,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.70873786407767,
|
| 1238 |
+
"grad_norm": 0.5795880556106567,
|
| 1239 |
+
"learning_rate": 4.661238919584578e-06,
|
| 1240 |
+
"loss": 0.7914,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.7184466019417477,
|
| 1245 |
+
"grad_norm": 0.4433108866214752,
|
| 1246 |
+
"learning_rate": 4.6048881199502265e-06,
|
| 1247 |
+
"loss": 0.8377,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.7281553398058254,
|
| 1252 |
+
"grad_norm": 0.5359524488449097,
|
| 1253 |
+
"learning_rate": 4.548587777803198e-06,
|
| 1254 |
+
"loss": 0.841,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.737864077669903,
|
| 1259 |
+
"grad_norm": 0.5342350006103516,
|
| 1260 |
+
"learning_rate": 4.49234508293946e-06,
|
| 1261 |
+
"loss": 0.8118,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.7475728155339807,
|
| 1266 |
+
"grad_norm": 0.4875918924808502,
|
| 1267 |
+
"learning_rate": 4.436167217793167e-06,
|
| 1268 |
+
"loss": 0.8091,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.7572815533980584,
|
| 1273 |
+
"grad_norm": 0.44451451301574707,
|
| 1274 |
+
"learning_rate": 4.38006135651944e-06,
|
| 1275 |
+
"loss": 0.8012,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.766990291262136,
|
| 1280 |
+
"grad_norm": 0.6147545576095581,
|
| 1281 |
+
"learning_rate": 4.3240346640782014e-06,
|
| 1282 |
+
"loss": 0.7779,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.7766990291262137,
|
| 1287 |
+
"grad_norm": 0.4527565538883209,
|
| 1288 |
+
"learning_rate": 4.268094295319167e-06,
|
| 1289 |
+
"loss": 0.8184,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.7864077669902914,
|
| 1294 |
+
"grad_norm": 0.48011067509651184,
|
| 1295 |
+
"learning_rate": 4.212247394068151e-06,
|
| 1296 |
+
"loss": 0.8109,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.796116504854369,
|
| 1301 |
+
"grad_norm": 0.5363189578056335,
|
| 1302 |
+
"learning_rate": 4.1565010922147644e-06,
|
| 1303 |
+
"loss": 0.8068,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.8058252427184467,
|
| 1308 |
+
"grad_norm": 0.5342061519622803,
|
| 1309 |
+
"learning_rate": 4.100862508801639e-06,
|
| 1310 |
+
"loss": 0.8158,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.8155339805825244,
|
| 1315 |
+
"grad_norm": 0.4906405508518219,
|
| 1316 |
+
"learning_rate": 4.045338749115299e-06,
|
| 1317 |
+
"loss": 0.8012,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.825242718446602,
|
| 1322 |
+
"grad_norm": 0.4942418038845062,
|
| 1323 |
+
"learning_rate": 3.989936903778785e-06,
|
| 1324 |
+
"loss": 0.8276,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.8349514563106797,
|
| 1329 |
+
"grad_norm": 0.47334739565849304,
|
| 1330 |
+
"learning_rate": 3.934664047846157e-06,
|
| 1331 |
+
"loss": 0.7817,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.8446601941747574,
|
| 1336 |
+
"grad_norm": 0.4942063093185425,
|
| 1337 |
+
"learning_rate": 3.879527239898962e-06,
|
| 1338 |
+
"loss": 0.8325,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.854368932038835,
|
| 1343 |
+
"grad_norm": 0.5283282995223999,
|
| 1344 |
+
"learning_rate": 3.8245335211448404e-06,
|
| 1345 |
+
"loss": 0.7786,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.8640776699029127,
|
| 1350 |
+
"grad_norm": 0.5127359628677368,
|
| 1351 |
+
"learning_rate": 3.769689914518326e-06,
|
| 1352 |
+
"loss": 0.8142,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.8737864077669903,
|
| 1357 |
+
"grad_norm": 0.49381014704704285,
|
| 1358 |
+
"learning_rate": 3.715003423783986e-06,
|
| 1359 |
+
"loss": 0.7969,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.883495145631068,
|
| 1364 |
+
"grad_norm": 0.46980249881744385,
|
| 1365 |
+
"learning_rate": 3.660481032642016e-06,
|
| 1366 |
+
"loss": 0.7949,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.8932038834951457,
|
| 1371 |
+
"grad_norm": 0.5440580248832703,
|
| 1372 |
+
"learning_rate": 3.6061297038363853e-06,
|
| 1373 |
+
"loss": 0.7786,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.9029126213592233,
|
| 1378 |
+
"grad_norm": 0.5657265782356262,
|
| 1379 |
+
"learning_rate": 3.5519563782656642e-06,
|
| 1380 |
+
"loss": 0.8131,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.912621359223301,
|
| 1385 |
+
"grad_norm": 0.560371458530426,
|
| 1386 |
+
"learning_rate": 3.497967974096647e-06,
|
| 1387 |
+
"loss": 0.7723,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.9223300970873787,
|
| 1392 |
+
"grad_norm": 0.4916374087333679,
|
| 1393 |
+
"learning_rate": 3.4441713858808684e-06,
|
| 1394 |
+
"loss": 0.793,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.9320388349514563,
|
| 1399 |
+
"grad_norm": 0.5217000246047974,
|
| 1400 |
+
"learning_rate": 3.3905734836741415e-06,
|
| 1401 |
+
"loss": 0.8259,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.941747572815534,
|
| 1406 |
+
"grad_norm": 0.6433777213096619,
|
| 1407 |
+
"learning_rate": 3.3371811121592203e-06,
|
| 1408 |
+
"loss": 0.7691,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.9514563106796117,
|
| 1413 |
+
"grad_norm": 0.5687490701675415,
|
| 1414 |
+
"learning_rate": 3.2840010897717045e-06,
|
| 1415 |
+
"loss": 0.7592,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.9611650485436893,
|
| 1420 |
+
"grad_norm": 0.5231119990348816,
|
| 1421 |
+
"learning_rate": 3.2310402078292956e-06,
|
| 1422 |
+
"loss": 0.806,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.970873786407767,
|
| 1427 |
+
"grad_norm": 0.5205345153808594,
|
| 1428 |
+
"learning_rate": 3.178305229664519e-06,
|
| 1429 |
+
"loss": 0.8302,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.9805825242718447,
|
| 1434 |
+
"grad_norm": 0.5407435894012451,
|
| 1435 |
+
"learning_rate": 3.125802889761016e-06,
|
| 1436 |
+
"loss": 0.7726,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.9902912621359223,
|
| 1441 |
+
"grad_norm": 0.628868818283081,
|
| 1442 |
+
"learning_rate": 3.073539892893519e-06,
|
| 1443 |
+
"loss": 0.8098,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.0,
|
| 1448 |
+
"grad_norm": 0.5963810682296753,
|
| 1449 |
+
"learning_rate": 3.021522913271627e-06,
|
| 1450 |
+
"loss": 0.7834,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 2.0097087378640777,
|
| 1455 |
+
"grad_norm": 0.5925841927528381,
|
| 1456 |
+
"learning_rate": 2.969758593687475e-06,
|
| 1457 |
+
"loss": 0.7591,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 2.0194174757281553,
|
| 1462 |
+
"grad_norm": 0.47755351662635803,
|
| 1463 |
+
"learning_rate": 2.9182535446674244e-06,
|
| 1464 |
+
"loss": 0.7803,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 2.029126213592233,
|
| 1469 |
+
"grad_norm": 0.5342966914176941,
|
| 1470 |
+
"learning_rate": 2.8670143436278757e-06,
|
| 1471 |
+
"loss": 0.7553,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 2.0388349514563107,
|
| 1476 |
+
"grad_norm": 0.5112743377685547,
|
| 1477 |
+
"learning_rate": 2.8160475340352913e-06,
|
| 1478 |
+
"loss": 0.7849,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 2.0485436893203883,
|
| 1483 |
+
"grad_norm": 0.5908800363540649,
|
| 1484 |
+
"learning_rate": 2.765359624570574e-06,
|
| 1485 |
+
"loss": 0.7618,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 2.058252427184466,
|
| 1490 |
+
"grad_norm": 0.4771662652492523,
|
| 1491 |
+
"learning_rate": 2.714957088297886e-06,
|
| 1492 |
+
"loss": 0.6928,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 2.0679611650485437,
|
| 1497 |
+
"grad_norm": 0.5731369853019714,
|
| 1498 |
+
"learning_rate": 2.664846361837997e-06,
|
| 1499 |
+
"loss": 0.7647,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 2.0776699029126213,
|
| 1504 |
+
"grad_norm": 0.4691663384437561,
|
| 1505 |
+
"learning_rate": 2.6150338445463146e-06,
|
| 1506 |
+
"loss": 0.7537,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 2.087378640776699,
|
| 1511 |
+
"grad_norm": 0.5235918760299683,
|
| 1512 |
+
"learning_rate": 2.565525897695651e-06,
|
| 1513 |
+
"loss": 0.7775,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 2.0970873786407767,
|
| 1518 |
+
"grad_norm": 0.49530091881752014,
|
| 1519 |
+
"learning_rate": 2.51632884366386e-06,
|
| 1520 |
+
"loss": 0.7784,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 2.1067961165048543,
|
| 1525 |
+
"grad_norm": 0.5703685283660889,
|
| 1526 |
+
"learning_rate": 2.4674489651264433e-06,
|
| 1527 |
+
"loss": 0.7207,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 2.116504854368932,
|
| 1532 |
+
"grad_norm": 0.45397961139678955,
|
| 1533 |
+
"learning_rate": 2.418892504254231e-06,
|
| 1534 |
+
"loss": 0.7493,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 2.1262135922330097,
|
| 1539 |
+
"grad_norm": 0.512279212474823,
|
| 1540 |
+
"learning_rate": 2.3706656619162278e-06,
|
| 1541 |
+
"loss": 0.7775,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 2.1359223300970873,
|
| 1546 |
+
"grad_norm": 0.5071080327033997,
|
| 1547 |
+
"learning_rate": 2.322774596887726e-06,
|
| 1548 |
+
"loss": 0.7594,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 2.145631067961165,
|
| 1553 |
+
"grad_norm": 0.5117892026901245,
|
| 1554 |
+
"learning_rate": 2.275225425063813e-06,
|
| 1555 |
+
"loss": 0.7265,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 2.1553398058252426,
|
| 1560 |
+
"grad_norm": 0.4993806779384613,
|
| 1561 |
+
"learning_rate": 2.2280242186783473e-06,
|
| 1562 |
+
"loss": 0.7149,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 2.1650485436893203,
|
| 1567 |
+
"grad_norm": 0.4563067853450775,
|
| 1568 |
+
"learning_rate": 2.1811770055284968e-06,
|
| 1569 |
+
"loss": 0.7902,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 2.174757281553398,
|
| 1574 |
+
"grad_norm": 0.47798827290534973,
|
| 1575 |
+
"learning_rate": 2.134689768204975e-06,
|
| 1576 |
+
"loss": 0.7583,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 2.1844660194174756,
|
| 1581 |
+
"grad_norm": 0.46698057651519775,
|
| 1582 |
+
"learning_rate": 2.0885684433280336e-06,
|
| 1583 |
+
"loss": 0.7717,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 2.1941747572815533,
|
| 1588 |
+
"grad_norm": 0.46322739124298096,
|
| 1589 |
+
"learning_rate": 2.042818920789326e-06,
|
| 1590 |
+
"loss": 0.7622,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 2.203883495145631,
|
| 1595 |
+
"grad_norm": 0.4613381028175354,
|
| 1596 |
+
"learning_rate": 1.9974470429997482e-06,
|
| 1597 |
+
"loss": 0.7508,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 2.2135922330097086,
|
| 1602 |
+
"grad_norm": 0.4670693278312683,
|
| 1603 |
+
"learning_rate": 1.9524586041433393e-06,
|
| 1604 |
+
"loss": 0.7132,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 2.2233009708737863,
|
| 1609 |
+
"grad_norm": 0.45155465602874756,
|
| 1610 |
+
"learning_rate": 1.907859349437336e-06,
|
| 1611 |
+
"loss": 0.7544,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.233009708737864,
|
| 1616 |
+
"grad_norm": 0.4583199620246887,
|
| 1617 |
+
"learning_rate": 1.8636549743984815e-06,
|
| 1618 |
+
"loss": 0.7485,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 2.2427184466019416,
|
| 1623 |
+
"grad_norm": 0.4572857618331909,
|
| 1624 |
+
"learning_rate": 1.8198511241156902e-06,
|
| 1625 |
+
"loss": 0.7552,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 2.2524271844660193,
|
| 1630 |
+
"grad_norm": 0.47898781299591064,
|
| 1631 |
+
"learning_rate": 1.776453392529139e-06,
|
| 1632 |
+
"loss": 0.7683,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.262135922330097,
|
| 1637 |
+
"grad_norm": 0.47700902819633484,
|
| 1638 |
+
"learning_rate": 1.7334673217158976e-06,
|
| 1639 |
+
"loss": 0.7444,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 2.2718446601941746,
|
| 1644 |
+
"grad_norm": 0.45447424054145813,
|
| 1645 |
+
"learning_rate": 1.6908984011821883e-06,
|
| 1646 |
+
"loss": 0.7495,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 2.2815533980582523,
|
| 1651 |
+
"grad_norm": 0.4574134647846222,
|
| 1652 |
+
"learning_rate": 1.6487520671623469e-06,
|
| 1653 |
+
"loss": 0.7545,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 2.29126213592233,
|
| 1658 |
+
"grad_norm": 0.4218170642852783,
|
| 1659 |
+
"learning_rate": 1.6070337019245896e-06,
|
| 1660 |
+
"loss": 0.7521,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 2.3009708737864076,
|
| 1665 |
+
"grad_norm": 0.41513901948928833,
|
| 1666 |
+
"learning_rate": 1.5657486330836786e-06,
|
| 1667 |
+
"loss": 0.759,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 2.3106796116504853,
|
| 1672 |
+
"grad_norm": 0.4804514944553375,
|
| 1673 |
+
"learning_rate": 1.5249021329205638e-06,
|
| 1674 |
+
"loss": 0.757,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 2.320388349514563,
|
| 1679 |
+
"grad_norm": 0.48271673917770386,
|
| 1680 |
+
"learning_rate": 1.4844994177090871e-06,
|
| 1681 |
+
"loss": 0.7533,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 2.3300970873786406,
|
| 1686 |
+
"grad_norm": 0.4723464846611023,
|
| 1687 |
+
"learning_rate": 1.4445456470498392e-06,
|
| 1688 |
+
"loss": 0.7859,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 2.3398058252427183,
|
| 1693 |
+
"grad_norm": 0.4236687123775482,
|
| 1694 |
+
"learning_rate": 1.4050459232112652e-06,
|
| 1695 |
+
"loss": 0.7873,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 2.349514563106796,
|
| 1700 |
+
"grad_norm": 0.4499347507953644,
|
| 1701 |
+
"learning_rate": 1.3660052904780707e-06,
|
| 1702 |
+
"loss": 0.7664,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 2.3592233009708736,
|
| 1707 |
+
"grad_norm": 0.4442787170410156,
|
| 1708 |
+
"learning_rate": 1.3274287345070564e-06,
|
| 1709 |
+
"loss": 0.7408,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 2.3689320388349513,
|
| 1714 |
+
"grad_norm": 0.45914286375045776,
|
| 1715 |
+
"learning_rate": 1.2893211816904243e-06,
|
| 1716 |
+
"loss": 0.7218,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 2.378640776699029,
|
| 1721 |
+
"grad_norm": 0.42868661880493164,
|
| 1722 |
+
"learning_rate": 1.2516874985266508e-06,
|
| 1723 |
+
"loss": 0.7466,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 2.3883495145631066,
|
| 1728 |
+
"grad_norm": 0.46668022871017456,
|
| 1729 |
+
"learning_rate": 1.2145324909990202e-06,
|
| 1730 |
+
"loss": 0.7116,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 2.3980582524271843,
|
| 1735 |
+
"grad_norm": 0.47152259945869446,
|
| 1736 |
+
"learning_rate": 1.1778609039618804e-06,
|
| 1737 |
+
"loss": 0.7446,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.407766990291262,
|
| 1742 |
+
"grad_norm": 0.4311952292919159,
|
| 1743 |
+
"learning_rate": 1.1416774205347015e-06,
|
| 1744 |
+
"loss": 0.7546,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 2.4174757281553396,
|
| 1749 |
+
"grad_norm": 0.47215738892555237,
|
| 1750 |
+
"learning_rate": 1.1059866615040205e-06,
|
| 1751 |
+
"loss": 0.7843,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 2.4271844660194173,
|
| 1756 |
+
"grad_norm": 0.42326658964157104,
|
| 1757 |
+
"learning_rate": 1.0707931847333487e-06,
|
| 1758 |
+
"loss": 0.7519,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 2.436893203883495,
|
| 1763 |
+
"grad_norm": 0.43430864810943604,
|
| 1764 |
+
"learning_rate": 1.036101484581117e-06,
|
| 1765 |
+
"loss": 0.7283,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 2.4466019417475726,
|
| 1770 |
+
"grad_norm": 0.45083683729171753,
|
| 1771 |
+
"learning_rate": 1.0019159913267156e-06,
|
| 1772 |
+
"loss": 0.7529,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.4563106796116507,
|
| 1777 |
+
"grad_norm": 0.49279189109802246,
|
| 1778 |
+
"learning_rate": 9.68241070604743e-07,
|
| 1779 |
+
"loss": 0.763,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 2.466019417475728,
|
| 1784 |
+
"grad_norm": 0.44121843576431274,
|
| 1785 |
+
"learning_rate": 9.350810228474855e-07,
|
| 1786 |
+
"loss": 0.7093,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 2.475728155339806,
|
| 1791 |
+
"grad_norm": 0.46301037073135376,
|
| 1792 |
+
"learning_rate": 9.024400827357344e-07,
|
| 1793 |
+
"loss": 0.7699,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 2.4854368932038833,
|
| 1798 |
+
"grad_norm": 0.4788742661476135,
|
| 1799 |
+
"learning_rate": 8.703224186580012e-07,
|
| 1800 |
+
"loss": 0.7811,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 2.4951456310679614,
|
| 1805 |
+
"grad_norm": 0.4840719401836395,
|
| 1806 |
+
"learning_rate": 8.387321321781977e-07,
|
| 1807 |
+
"loss": 0.7633,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 2.5048543689320386,
|
| 1812 |
+
"grad_norm": 0.4576054811477661,
|
| 1813 |
+
"learning_rate": 8.07673257511849e-07,
|
| 1814 |
+
"loss": 0.7755,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 2.5145631067961167,
|
| 1819 |
+
"grad_norm": 0.43629974126815796,
|
| 1820 |
+
"learning_rate": 7.771497610108981e-07,
|
| 1821 |
+
"loss": 0.7574,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 2.524271844660194,
|
| 1826 |
+
"grad_norm": 0.4582881033420563,
|
| 1827 |
+
"learning_rate": 7.471655406572003e-07,
|
| 1828 |
+
"loss": 0.7703,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 2.533980582524272,
|
| 1833 |
+
"grad_norm": 0.4240984618663788,
|
| 1834 |
+
"learning_rate": 7.177244255647209e-07,
|
| 1835 |
+
"loss": 0.7567,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 2.5436893203883493,
|
| 1840 |
+
"grad_norm": 0.4236100912094116,
|
| 1841 |
+
"learning_rate": 6.888301754905469e-07,
|
| 1842 |
+
"loss": 0.7912,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.5533980582524274,
|
| 1847 |
+
"grad_norm": 0.40091320872306824,
|
| 1848 |
+
"learning_rate": 6.604864803547511e-07,
|
| 1849 |
+
"loss": 0.7474,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 2.5631067961165046,
|
| 1854 |
+
"grad_norm": 0.44925597310066223,
|
| 1855 |
+
"learning_rate": 6.326969597691724e-07,
|
| 1856 |
+
"loss": 0.7381,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 2.5728155339805827,
|
| 1861 |
+
"grad_norm": 0.4370911121368408,
|
| 1862 |
+
"learning_rate": 6.054651625751717e-07,
|
| 1863 |
+
"loss": 0.7587,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.58252427184466,
|
| 1868 |
+
"grad_norm": 0.43402257561683655,
|
| 1869 |
+
"learning_rate": 5.787945663904332e-07,
|
| 1870 |
+
"loss": 0.7715,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 2.592233009708738,
|
| 1875 |
+
"grad_norm": 0.4456167221069336,
|
| 1876 |
+
"learning_rate": 5.526885771648599e-07,
|
| 1877 |
+
"loss": 0.7469,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 2.6019417475728153,
|
| 1882 |
+
"grad_norm": 0.44401729106903076,
|
| 1883 |
+
"learning_rate": 5.271505287456153e-07,
|
| 1884 |
+
"loss": 0.7223,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 2.6116504854368934,
|
| 1889 |
+
"grad_norm": 0.4610472023487091,
|
| 1890 |
+
"learning_rate": 5.021836824513759e-07,
|
| 1891 |
+
"loss": 0.7802,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 2.6213592233009706,
|
| 1896 |
+
"grad_norm": 0.4633938670158386,
|
| 1897 |
+
"learning_rate": 4.777912266558532e-07,
|
| 1898 |
+
"loss": 0.7417,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 2.6310679611650487,
|
| 1903 |
+
"grad_norm": 0.4150261878967285,
|
| 1904 |
+
"learning_rate": 4.5397627638061604e-07,
|
| 1905 |
+
"loss": 0.7705,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 2.6407766990291264,
|
| 1910 |
+
"grad_norm": 0.45528119802474976,
|
| 1911 |
+
"learning_rate": 4.307418728972934e-07,
|
| 1912 |
+
"loss": 0.7391,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 2.650485436893204,
|
| 1917 |
+
"grad_norm": 0.4313214123249054,
|
| 1918 |
+
"learning_rate": 4.080909833391944e-07,
|
| 1919 |
+
"loss": 0.7624,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 2.6601941747572817,
|
| 1924 |
+
"grad_norm": 0.403175950050354,
|
| 1925 |
+
"learning_rate": 3.8602650032238675e-07,
|
| 1926 |
+
"loss": 0.7797,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 2.6699029126213594,
|
| 1931 |
+
"grad_norm": 0.4482267498970032,
|
| 1932 |
+
"learning_rate": 3.6455124157629805e-07,
|
| 1933 |
+
"loss": 0.7373,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 2.679611650485437,
|
| 1938 |
+
"grad_norm": 0.42840349674224854,
|
| 1939 |
+
"learning_rate": 3.436679495838835e-07,
|
| 1940 |
+
"loss": 0.7349,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 2.6893203883495147,
|
| 1945 |
+
"grad_norm": 0.41052892804145813,
|
| 1946 |
+
"learning_rate": 3.233792912313943e-07,
|
| 1947 |
+
"loss": 0.754,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.6990291262135924,
|
| 1952 |
+
"grad_norm": 0.4851820170879364,
|
| 1953 |
+
"learning_rate": 3.0368785746780925e-07,
|
| 1954 |
+
"loss": 0.7887,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 2.70873786407767,
|
| 1959 |
+
"grad_norm": 0.43496590852737427,
|
| 1960 |
+
"learning_rate": 2.8459616297395464e-07,
|
| 1961 |
+
"loss": 0.7558,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 2.7184466019417477,
|
| 1966 |
+
"grad_norm": 0.44541043043136597,
|
| 1967 |
+
"learning_rate": 2.6610664584137413e-07,
|
| 1968 |
+
"loss": 0.7088,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 2.7281553398058254,
|
| 1973 |
+
"grad_norm": 0.45385488867759705,
|
| 1974 |
+
"learning_rate": 2.482216672609677e-07,
|
| 1975 |
+
"loss": 0.7279,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.737864077669903,
|
| 1980 |
+
"grad_norm": 0.4380588233470917,
|
| 1981 |
+
"learning_rate": 2.3094351122146307e-07,
|
| 1982 |
+
"loss": 0.7354,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 2.7475728155339807,
|
| 1987 |
+
"grad_norm": 0.4383409321308136,
|
| 1988 |
+
"learning_rate": 2.142743842177386e-07,
|
| 1989 |
+
"loss": 0.7427,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 2.7572815533980584,
|
| 1994 |
+
"grad_norm": 0.4154166579246521,
|
| 1995 |
+
"learning_rate": 1.98216414969043e-07,
|
| 1996 |
+
"loss": 0.7533,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 2.766990291262136,
|
| 2001 |
+
"grad_norm": 0.40241143107414246,
|
| 2002 |
+
"learning_rate": 1.8277165414714858e-07,
|
| 2003 |
+
"loss": 0.7471,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 2.7766990291262137,
|
| 2008 |
+
"grad_norm": 0.4261252284049988,
|
| 2009 |
+
"learning_rate": 1.6794207411447548e-07,
|
| 2010 |
+
"loss": 0.7628,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 2.7864077669902914,
|
| 2015 |
+
"grad_norm": 0.39932355284690857,
|
| 2016 |
+
"learning_rate": 1.5372956867220678e-07,
|
| 2017 |
+
"loss": 0.7747,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 2.796116504854369,
|
| 2022 |
+
"grad_norm": 0.40878960490226746,
|
| 2023 |
+
"learning_rate": 1.4013595281844872e-07,
|
| 2024 |
+
"loss": 0.7583,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 2.8058252427184467,
|
| 2029 |
+
"grad_norm": 0.42195621132850647,
|
| 2030 |
+
"learning_rate": 1.2716296251644e-07,
|
| 2031 |
+
"loss": 0.7675,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 2.8155339805825244,
|
| 2036 |
+
"grad_norm": 0.41512733697891235,
|
| 2037 |
+
"learning_rate": 1.1481225447286803e-07,
|
| 2038 |
+
"loss": 0.7228,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 2.825242718446602,
|
| 2043 |
+
"grad_norm": 0.42841315269470215,
|
| 2044 |
+
"learning_rate": 1.0308540592629756e-07,
|
| 2045 |
+
"loss": 0.7937,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 2.8349514563106797,
|
| 2050 |
+
"grad_norm": 0.445648193359375,
|
| 2051 |
+
"learning_rate": 9.198391444575072e-08,
|
| 2052 |
+
"loss": 0.7037,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 2.8446601941747574,
|
| 2057 |
+
"grad_norm": 0.4242344796657562,
|
| 2058 |
+
"learning_rate": 8.150919773946165e-08,
|
| 2059 |
+
"loss": 0.7795,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 2.854368932038835,
|
| 2064 |
+
"grad_norm": 0.4234623610973358,
|
| 2065 |
+
"learning_rate": 7.166259347382854e-08,
|
| 2066 |
+
"loss": 0.7292,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 2.8640776699029127,
|
| 2071 |
+
"grad_norm": 0.3901122808456421,
|
| 2072 |
+
"learning_rate": 6.244535910258697e-08,
|
| 2073 |
+
"loss": 0.776,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 2.8737864077669903,
|
| 2078 |
+
"grad_norm": 0.41637924313545227,
|
| 2079 |
+
"learning_rate": 5.3858671706230605e-08,
|
| 2080 |
+
"loss": 0.7917,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.883495145631068,
|
| 2085 |
+
"grad_norm": 0.39965224266052246,
|
| 2086 |
+
"learning_rate": 4.590362784169022e-08,
|
| 2087 |
+
"loss": 0.8002,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 2.8932038834951457,
|
| 2092 |
+
"grad_norm": 0.48560866713523865,
|
| 2093 |
+
"learning_rate": 3.858124340229863e-08,
|
| 2094 |
+
"loss": 0.7326,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 2.9029126213592233,
|
| 2099 |
+
"grad_norm": 0.41089245676994324,
|
| 2100 |
+
"learning_rate": 3.1892453488058803e-08,
|
| 2101 |
+
"loss": 0.7604,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 2.912621359223301,
|
| 2106 |
+
"grad_norm": 0.4088441729545593,
|
| 2107 |
+
"learning_rate": 2.5838112286226123e-08,
|
| 2108 |
+
"loss": 0.7339,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 2.9223300970873787,
|
| 2113 |
+
"grad_norm": 0.43762901425361633,
|
| 2114 |
+
"learning_rate": 2.0418992962224495e-08,
|
| 2115 |
+
"loss": 0.7299,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 2.9320388349514563,
|
| 2120 |
+
"grad_norm": 0.41060447692871094,
|
| 2121 |
+
"learning_rate": 1.563578756091144e-08,
|
| 2122 |
+
"loss": 0.7919,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.941747572815534,
|
| 2127 |
+
"grad_norm": 0.4254586398601532,
|
| 2128 |
+
"learning_rate": 1.1489106918200487e-08,
|
| 2129 |
+
"loss": 0.6998,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 2.9514563106796117,
|
| 2134 |
+
"grad_norm": 0.4328642785549164,
|
| 2135 |
+
"learning_rate": 7.979480583052423e-09,
|
| 2136 |
+
"loss": 0.7362,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 2.9611650485436893,
|
| 2141 |
+
"grad_norm": 0.40661028027534485,
|
| 2142 |
+
"learning_rate": 5.107356749853298e-09,
|
| 2143 |
+
"loss": 0.7416,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 2.970873786407767,
|
| 2148 |
+
"grad_norm": 0.4029662609100342,
|
| 2149 |
+
"learning_rate": 2.8731022011757593e-09,
|
| 2150 |
+
"loss": 0.7835,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 2.9805825242718447,
|
| 2155 |
+
"grad_norm": 0.4190027117729187,
|
| 2156 |
+
"learning_rate": 1.2770022609409628e-09,
|
| 2157 |
+
"loss": 0.7464,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 2.9902912621359223,
|
| 2162 |
+
"grad_norm": 0.40702641010284424,
|
| 2163 |
+
"learning_rate": 3.1926075797827914e-10,
|
| 2164 |
+
"loss": 0.7806,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 3.0,
|
| 2169 |
+
"grad_norm": 0.40051835775375366,
|
| 2170 |
+
"learning_rate": 0.0,
|
| 2171 |
+
"loss": 0.7573,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 3.0,
|
| 2176 |
+
"step": 309,
|
| 2177 |
+
"total_flos": 266439779352576.0,
|
| 2178 |
+
"train_loss": 0.8324689010586168,
|
| 2179 |
+
"train_runtime": 15120.0259,
|
| 2180 |
+
"train_samples_per_second": 1.962,
|
| 2181 |
+
"train_steps_per_second": 0.02
|
| 2182 |
+
}
|
| 2183 |
+
],
|
| 2184 |
+
"logging_steps": 1.0,
|
| 2185 |
+
"max_steps": 309,
|
| 2186 |
+
"num_input_tokens_seen": 0,
|
| 2187 |
+
"num_train_epochs": 3,
|
| 2188 |
+
"save_steps": 500,
|
| 2189 |
+
"stateful_callbacks": {
|
| 2190 |
+
"TrainerControl": {
|
| 2191 |
+
"args": {
|
| 2192 |
+
"should_epoch_stop": false,
|
| 2193 |
+
"should_evaluate": false,
|
| 2194 |
+
"should_log": false,
|
| 2195 |
+
"should_save": true,
|
| 2196 |
+
"should_training_stop": true
|
| 2197 |
+
},
|
| 2198 |
+
"attributes": {}
|
| 2199 |
+
}
|
| 2200 |
+
},
|
| 2201 |
+
"total_flos": 266439779352576.0,
|
| 2202 |
+
"train_batch_size": 1,
|
| 2203 |
+
"trial_name": null,
|
| 2204 |
+
"trial_params": null
|
| 2205 |
+
}
|
training_loss.png
ADDED
|