Model save
Browse files- README.md +68 -0
- all_results.json +8 -0
- generation_config.json +9 -0
- train_results.json +8 -0
- trainer_state.json +1602 -0
README.md
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: DeepSeek-R1-Distill-Qwen-1.5B-GRPO_generation_num_2_new
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- grpo
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for DeepSeek-R1-Distill-Qwen-1.5B-GRPO_generation_num_2_new
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B).
|
| 15 |
+
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 16 |
+
|
| 17 |
+
## Quick start
|
| 18 |
+
|
| 19 |
+
```python
|
| 20 |
+
from transformers import pipeline
|
| 21 |
+
|
| 22 |
+
question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
|
| 23 |
+
generator = pipeline("text-generation", model="EricLabile/DeepSeek-R1-Distill-Qwen-1.5B-GRPO_generation_num_2_new", device="cuda")
|
| 24 |
+
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
|
| 25 |
+
print(output["generated_text"])
|
| 26 |
+
```
|
| 27 |
+
|
| 28 |
+
## Training procedure
|
| 29 |
+
|
| 30 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/ericlabile/GRPO-Experiments/runs/pd6pdf2e)
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
|
| 34 |
+
|
| 35 |
+
### Framework versions
|
| 36 |
+
|
| 37 |
+
- TRL: 0.16.0.dev0
|
| 38 |
+
- Transformers: 4.49.0
|
| 39 |
+
- Pytorch: 2.5.1
|
| 40 |
+
- Datasets: 3.4.0
|
| 41 |
+
- Tokenizers: 0.21.1
|
| 42 |
+
|
| 43 |
+
## Citations
|
| 44 |
+
|
| 45 |
+
Cite GRPO as:
|
| 46 |
+
|
| 47 |
+
```bibtex
|
| 48 |
+
@article{zhihong2024deepseekmath,
|
| 49 |
+
title = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
|
| 50 |
+
author = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
|
| 51 |
+
year = 2024,
|
| 52 |
+
eprint = {arXiv:2402.03300},
|
| 53 |
+
}
|
| 54 |
+
|
| 55 |
+
```
|
| 56 |
+
|
| 57 |
+
Cite TRL as:
|
| 58 |
+
|
| 59 |
+
```bibtex
|
| 60 |
+
@misc{vonwerra2022trl,
|
| 61 |
+
title = {{TRL: Transformer Reinforcement Learning}},
|
| 62 |
+
author = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallouédec},
|
| 63 |
+
year = 2020,
|
| 64 |
+
journal = {GitHub repository},
|
| 65 |
+
publisher = {GitHub},
|
| 66 |
+
howpublished = {\url{https://github.com/huggingface/trl}}
|
| 67 |
+
}
|
| 68 |
+
```
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 0.02088056951134394,
|
| 4 |
+
"train_runtime": 4550.0212,
|
| 5 |
+
"train_samples": 7500,
|
| 6 |
+
"train_samples_per_second": 3.297,
|
| 7 |
+
"train_steps_per_second": 0.023
|
| 8 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151646,
|
| 4 |
+
"do_sample": true,
|
| 5 |
+
"eos_token_id": 151643,
|
| 6 |
+
"temperature": 0.6,
|
| 7 |
+
"top_p": 0.95,
|
| 8 |
+
"transformers_version": "4.49.0"
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 0.02088056951134394,
|
| 4 |
+
"train_runtime": 4550.0212,
|
| 5 |
+
"train_samples": 7500,
|
| 6 |
+
"train_samples_per_second": 3.297,
|
| 7 |
+
"train_steps_per_second": 0.023
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1602 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.9952153110047846,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 104,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"clip_ratio": 0.0,
|
| 13 |
+
"completion_length": 885.7360992431641,
|
| 14 |
+
"epoch": 0.019138755980861243,
|
| 15 |
+
"grad_norm": 0.0571745065537259,
|
| 16 |
+
"kl": 0.0,
|
| 17 |
+
"learning_rate": 9.09090909090909e-08,
|
| 18 |
+
"loss": 0.0224,
|
| 19 |
+
"reward": 0.6623263955116272,
|
| 20 |
+
"reward_std": 0.20746709778904915,
|
| 21 |
+
"rewards/accuracy_reward": 0.2500000037252903,
|
| 22 |
+
"rewards/format_reward": 0.0,
|
| 23 |
+
"rewards/tag_count_reward": 0.4123263880610466,
|
| 24 |
+
"step": 1
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"clip_ratio": 0.0,
|
| 28 |
+
"completion_length": 855.6076354980469,
|
| 29 |
+
"epoch": 0.03827751196172249,
|
| 30 |
+
"grad_norm": 0.06704915151383242,
|
| 31 |
+
"kl": 0.0,
|
| 32 |
+
"learning_rate": 1.818181818181818e-07,
|
| 33 |
+
"loss": 0.0199,
|
| 34 |
+
"reward": 0.6901041716337204,
|
| 35 |
+
"reward_std": 0.22956417128443718,
|
| 36 |
+
"rewards/accuracy_reward": 0.2465277761220932,
|
| 37 |
+
"rewards/format_reward": 0.0,
|
| 38 |
+
"rewards/tag_count_reward": 0.4435763880610466,
|
| 39 |
+
"step": 2
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"clip_ratio": 0.0,
|
| 43 |
+
"completion_length": 904.9861297607422,
|
| 44 |
+
"epoch": 0.05741626794258373,
|
| 45 |
+
"grad_norm": 0.05238384111260958,
|
| 46 |
+
"kl": 2.7120113372802734e-05,
|
| 47 |
+
"learning_rate": 2.727272727272727e-07,
|
| 48 |
+
"loss": 0.0109,
|
| 49 |
+
"reward": 0.5494791641831398,
|
| 50 |
+
"reward_std": 0.17554907500743866,
|
| 51 |
+
"rewards/accuracy_reward": 0.17361111380159855,
|
| 52 |
+
"rewards/format_reward": 0.0,
|
| 53 |
+
"rewards/tag_count_reward": 0.375868059694767,
|
| 54 |
+
"step": 3
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"clip_ratio": 0.0,
|
| 58 |
+
"completion_length": 880.2083435058594,
|
| 59 |
+
"epoch": 0.07655502392344497,
|
| 60 |
+
"grad_norm": 0.05735410600314712,
|
| 61 |
+
"kl": 2.5987625122070312e-05,
|
| 62 |
+
"learning_rate": 3.636363636363636e-07,
|
| 63 |
+
"loss": 0.0214,
|
| 64 |
+
"reward": 0.6362847238779068,
|
| 65 |
+
"reward_std": 0.22956416755914688,
|
| 66 |
+
"rewards/accuracy_reward": 0.2256944440305233,
|
| 67 |
+
"rewards/format_reward": 0.0,
|
| 68 |
+
"rewards/tag_count_reward": 0.4105902835726738,
|
| 69 |
+
"step": 4
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"clip_ratio": 0.0,
|
| 73 |
+
"completion_length": 867.8368225097656,
|
| 74 |
+
"epoch": 0.09569377990430622,
|
| 75 |
+
"grad_norm": 0.056581697743245854,
|
| 76 |
+
"kl": 2.8789043426513672e-05,
|
| 77 |
+
"learning_rate": 4.545454545454545e-07,
|
| 78 |
+
"loss": 0.0117,
|
| 79 |
+
"reward": 0.665798619389534,
|
| 80 |
+
"reward_std": 0.1829147655516863,
|
| 81 |
+
"rewards/accuracy_reward": 0.25,
|
| 82 |
+
"rewards/format_reward": 0.0,
|
| 83 |
+
"rewards/tag_count_reward": 0.4157986119389534,
|
| 84 |
+
"step": 5
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"clip_ratio": 0.0,
|
| 88 |
+
"completion_length": 884.0381927490234,
|
| 89 |
+
"epoch": 0.11483253588516747,
|
| 90 |
+
"grad_norm": 0.060995509922658335,
|
| 91 |
+
"kl": 2.4437904357910156e-05,
|
| 92 |
+
"learning_rate": 5.454545454545454e-07,
|
| 93 |
+
"loss": 0.0206,
|
| 94 |
+
"reward": 0.659722238779068,
|
| 95 |
+
"reward_std": 0.24306795001029968,
|
| 96 |
+
"rewards/accuracy_reward": 0.243055559694767,
|
| 97 |
+
"rewards/format_reward": 0.0,
|
| 98 |
+
"rewards/tag_count_reward": 0.4166666716337204,
|
| 99 |
+
"step": 6
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"clip_ratio": 0.0,
|
| 103 |
+
"completion_length": 878.7083435058594,
|
| 104 |
+
"epoch": 0.1339712918660287,
|
| 105 |
+
"grad_norm": 0.06312478683496295,
|
| 106 |
+
"kl": 2.7120113372802734e-05,
|
| 107 |
+
"learning_rate": 6.363636363636363e-07,
|
| 108 |
+
"loss": 0.0152,
|
| 109 |
+
"reward": 0.6484375149011612,
|
| 110 |
+
"reward_std": 0.224653709679842,
|
| 111 |
+
"rewards/accuracy_reward": 0.243055559694767,
|
| 112 |
+
"rewards/format_reward": 0.0,
|
| 113 |
+
"rewards/tag_count_reward": 0.4053819477558136,
|
| 114 |
+
"step": 7
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"clip_ratio": 0.0,
|
| 118 |
+
"completion_length": 871.4930572509766,
|
| 119 |
+
"epoch": 0.15311004784688995,
|
| 120 |
+
"grad_norm": 0.06419727373029128,
|
| 121 |
+
"kl": 2.6404857635498047e-05,
|
| 122 |
+
"learning_rate": 7.272727272727272e-07,
|
| 123 |
+
"loss": 0.0246,
|
| 124 |
+
"reward": 0.6631944477558136,
|
| 125 |
+
"reward_std": 0.2381574772298336,
|
| 126 |
+
"rewards/accuracy_reward": 0.2395833283662796,
|
| 127 |
+
"rewards/format_reward": 0.0,
|
| 128 |
+
"rewards/tag_count_reward": 0.423611119389534,
|
| 129 |
+
"step": 8
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"clip_ratio": 0.0,
|
| 133 |
+
"completion_length": 883.0347290039062,
|
| 134 |
+
"epoch": 0.1722488038277512,
|
| 135 |
+
"grad_norm": 0.07232208605276093,
|
| 136 |
+
"kl": 2.5197863578796387e-05,
|
| 137 |
+
"learning_rate": 8.181818181818182e-07,
|
| 138 |
+
"loss": 0.0217,
|
| 139 |
+
"reward": 0.6892361044883728,
|
| 140 |
+
"reward_std": 0.270075511187315,
|
| 141 |
+
"rewards/accuracy_reward": 0.2847222164273262,
|
| 142 |
+
"rewards/format_reward": 0.0,
|
| 143 |
+
"rewards/tag_count_reward": 0.4045138880610466,
|
| 144 |
+
"step": 9
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"clip_ratio": 0.0,
|
| 148 |
+
"completion_length": 869.8229217529297,
|
| 149 |
+
"epoch": 0.19138755980861244,
|
| 150 |
+
"grad_norm": 0.05684777755384174,
|
| 151 |
+
"kl": 2.7745962142944336e-05,
|
| 152 |
+
"learning_rate": 9.09090909090909e-07,
|
| 153 |
+
"loss": 0.0053,
|
| 154 |
+
"reward": 0.697048619389534,
|
| 155 |
+
"reward_std": 0.2197432443499565,
|
| 156 |
+
"rewards/accuracy_reward": 0.2916666641831398,
|
| 157 |
+
"rewards/format_reward": 0.0,
|
| 158 |
+
"rewards/tag_count_reward": 0.4053819477558136,
|
| 159 |
+
"step": 10
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"clip_ratio": 0.0,
|
| 163 |
+
"completion_length": 850.7152862548828,
|
| 164 |
+
"epoch": 0.21052631578947367,
|
| 165 |
+
"grad_norm": 0.07180408830721081,
|
| 166 |
+
"kl": 2.5004148483276367e-05,
|
| 167 |
+
"learning_rate": 1e-06,
|
| 168 |
+
"loss": 0.0299,
|
| 169 |
+
"reward": 0.706597238779068,
|
| 170 |
+
"reward_std": 0.28726211935281754,
|
| 171 |
+
"rewards/accuracy_reward": 0.2673611156642437,
|
| 172 |
+
"rewards/format_reward": 0.0,
|
| 173 |
+
"rewards/tag_count_reward": 0.439236119389534,
|
| 174 |
+
"step": 11
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"clip_ratio": 0.0,
|
| 178 |
+
"completion_length": 868.28125,
|
| 179 |
+
"epoch": 0.22966507177033493,
|
| 180 |
+
"grad_norm": 0.0646871027868035,
|
| 181 |
+
"kl": 2.7686357498168945e-05,
|
| 182 |
+
"learning_rate": 9.99743270918373e-07,
|
| 183 |
+
"loss": 0.0177,
|
| 184 |
+
"reward": 0.7300347238779068,
|
| 185 |
+
"reward_std": 0.24429556727409363,
|
| 186 |
+
"rewards/accuracy_reward": 0.2881944440305233,
|
| 187 |
+
"rewards/format_reward": 0.0,
|
| 188 |
+
"rewards/tag_count_reward": 0.4418402761220932,
|
| 189 |
+
"step": 12
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"clip_ratio": 0.0,
|
| 193 |
+
"completion_length": 864.7395935058594,
|
| 194 |
+
"epoch": 0.24880382775119617,
|
| 195 |
+
"grad_norm": 0.06556460276965595,
|
| 196 |
+
"kl": 2.7000904083251953e-05,
|
| 197 |
+
"learning_rate": 9.989733766060318e-07,
|
| 198 |
+
"loss": 0.0163,
|
| 199 |
+
"reward": 0.712673619389534,
|
| 200 |
+
"reward_std": 0.20255662500858307,
|
| 201 |
+
"rewards/accuracy_reward": 0.2777777910232544,
|
| 202 |
+
"rewards/format_reward": 0.0,
|
| 203 |
+
"rewards/tag_count_reward": 0.4348958283662796,
|
| 204 |
+
"step": 13
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"clip_ratio": 0.0,
|
| 208 |
+
"completion_length": 879.0590209960938,
|
| 209 |
+
"epoch": 0.2679425837320574,
|
| 210 |
+
"grad_norm": 0.06296015481337248,
|
| 211 |
+
"kl": 3.0308961868286133e-05,
|
| 212 |
+
"learning_rate": 9.976911955263528e-07,
|
| 213 |
+
"loss": 0.0101,
|
| 214 |
+
"reward": 0.6510416716337204,
|
| 215 |
+
"reward_std": 0.22588133066892624,
|
| 216 |
+
"rewards/accuracy_reward": 0.2326388917863369,
|
| 217 |
+
"rewards/format_reward": 0.0,
|
| 218 |
+
"rewards/tag_count_reward": 0.4184027761220932,
|
| 219 |
+
"step": 14
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"clip_ratio": 0.0,
|
| 223 |
+
"completion_length": 885.1006927490234,
|
| 224 |
+
"epoch": 0.28708133971291866,
|
| 225 |
+
"grad_norm": 0.06103661524458471,
|
| 226 |
+
"kl": 3.299117088317871e-05,
|
| 227 |
+
"learning_rate": 9.958981906712076e-07,
|
| 228 |
+
"loss": 0.0203,
|
| 229 |
+
"reward": 0.677083358168602,
|
| 230 |
+
"reward_std": 0.272530734539032,
|
| 231 |
+
"rewards/accuracy_reward": 0.2534722238779068,
|
| 232 |
+
"rewards/format_reward": 0.0,
|
| 233 |
+
"rewards/tag_count_reward": 0.4236111119389534,
|
| 234 |
+
"step": 15
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"clip_ratio": 0.0,
|
| 238 |
+
"completion_length": 874.2257080078125,
|
| 239 |
+
"epoch": 0.3062200956937799,
|
| 240 |
+
"grad_norm": 0.0764541954488641,
|
| 241 |
+
"kl": 5.513429641723633e-05,
|
| 242 |
+
"learning_rate": 9.935964078916617e-07,
|
| 243 |
+
"loss": 0.0251,
|
| 244 |
+
"reward": 0.681423619389534,
|
| 245 |
+
"reward_std": 0.21483278647065163,
|
| 246 |
+
"rewards/accuracy_reward": 0.2708333395421505,
|
| 247 |
+
"rewards/format_reward": 0.0,
|
| 248 |
+
"rewards/tag_count_reward": 0.4105902835726738,
|
| 249 |
+
"step": 16
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"clip_ratio": 0.0,
|
| 253 |
+
"completion_length": 852.8472290039062,
|
| 254 |
+
"epoch": 0.3253588516746411,
|
| 255 |
+
"grad_norm": 0.0763344765825447,
|
| 256 |
+
"kl": 0.00010287761688232422,
|
| 257 |
+
"learning_rate": 9.907884735636224e-07,
|
| 258 |
+
"loss": 0.0222,
|
| 259 |
+
"reward": 0.7560763955116272,
|
| 260 |
+
"reward_std": 0.2909449674189091,
|
| 261 |
+
"rewards/accuracy_reward": 0.3263888880610466,
|
| 262 |
+
"rewards/format_reward": 0.0,
|
| 263 |
+
"rewards/tag_count_reward": 0.4296875074505806,
|
| 264 |
+
"step": 17
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"clip_ratio": 0.0,
|
| 268 |
+
"completion_length": 873.6388854980469,
|
| 269 |
+
"epoch": 0.3444976076555024,
|
| 270 |
+
"grad_norm": 0.07015709920099024,
|
| 271 |
+
"kl": 0.0001728534698486328,
|
| 272 |
+
"learning_rate": 9.874775915910958e-07,
|
| 273 |
+
"loss": 0.0186,
|
| 274 |
+
"reward": 0.7039930671453476,
|
| 275 |
+
"reward_std": 0.224653709679842,
|
| 276 |
+
"rewards/accuracy_reward": 0.274305559694767,
|
| 277 |
+
"rewards/format_reward": 0.0,
|
| 278 |
+
"rewards/tag_count_reward": 0.4296875,
|
| 279 |
+
"step": 18
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"clip_ratio": 0.0,
|
| 283 |
+
"completion_length": 865.6250152587891,
|
| 284 |
+
"epoch": 0.36363636363636365,
|
| 285 |
+
"grad_norm": 0.06286838529602348,
|
| 286 |
+
"kl": 0.0002143383026123047,
|
| 287 |
+
"learning_rate": 9.836675397504734e-07,
|
| 288 |
+
"loss": 0.0188,
|
| 289 |
+
"reward": 0.724826380610466,
|
| 290 |
+
"reward_std": 0.23938511312007904,
|
| 291 |
+
"rewards/accuracy_reward": 0.3090277835726738,
|
| 292 |
+
"rewards/format_reward": 0.0,
|
| 293 |
+
"rewards/tag_count_reward": 0.415798619389534,
|
| 294 |
+
"step": 19
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"clip_ratio": 0.0,
|
| 298 |
+
"completion_length": 851.5555572509766,
|
| 299 |
+
"epoch": 0.3827751196172249,
|
| 300 |
+
"grad_norm": 0.0652308138114061,
|
| 301 |
+
"kl": 0.00026702880859375,
|
| 302 |
+
"learning_rate": 9.793626653800218e-07,
|
| 303 |
+
"loss": 0.019,
|
| 304 |
+
"reward": 0.7230902761220932,
|
| 305 |
+
"reward_std": 0.24184033274650574,
|
| 306 |
+
"rewards/accuracy_reward": 0.288194440305233,
|
| 307 |
+
"rewards/format_reward": 0.0,
|
| 308 |
+
"rewards/tag_count_reward": 0.4348958283662796,
|
| 309 |
+
"step": 20
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"clip_ratio": 0.0,
|
| 313 |
+
"completion_length": 861.40625,
|
| 314 |
+
"epoch": 0.4019138755980861,
|
| 315 |
+
"grad_norm": 0.054173134249339484,
|
| 316 |
+
"kl": 0.00030875205993652344,
|
| 317 |
+
"learning_rate": 9.74567880419491e-07,
|
| 318 |
+
"loss": 0.0092,
|
| 319 |
+
"reward": 0.6545138955116272,
|
| 320 |
+
"reward_std": 0.2086947076022625,
|
| 321 |
+
"rewards/accuracy_reward": 0.2361111119389534,
|
| 322 |
+
"rewards/format_reward": 0.0,
|
| 323 |
+
"rewards/tag_count_reward": 0.4184027835726738,
|
| 324 |
+
"step": 21
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"clip_ratio": 0.0,
|
| 328 |
+
"completion_length": 839.1805725097656,
|
| 329 |
+
"epoch": 0.42105263157894735,
|
| 330 |
+
"grad_norm": 0.06484920729256737,
|
| 331 |
+
"kl": 0.0003552436828613281,
|
| 332 |
+
"learning_rate": 9.692886558055006e-07,
|
| 333 |
+
"loss": 0.019,
|
| 334 |
+
"reward": 0.7100694626569748,
|
| 335 |
+
"reward_std": 0.23079179227352142,
|
| 336 |
+
"rewards/accuracy_reward": 0.2881944440305233,
|
| 337 |
+
"rewards/format_reward": 0.0,
|
| 338 |
+
"rewards/tag_count_reward": 0.4218749925494194,
|
| 339 |
+
"step": 22
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"clip_ratio": 0.0,
|
| 343 |
+
"completion_length": 889.3159790039062,
|
| 344 |
+
"epoch": 0.44019138755980863,
|
| 345 |
+
"grad_norm": 0.0673083530273057,
|
| 346 |
+
"kl": 0.0004329681396484375,
|
| 347 |
+
"learning_rate": 9.635310152291037e-07,
|
| 348 |
+
"loss": 0.02,
|
| 349 |
+
"reward": 0.6163194626569748,
|
| 350 |
+
"reward_std": 0.20132901147007942,
|
| 351 |
+
"rewards/accuracy_reward": 0.2222222201526165,
|
| 352 |
+
"rewards/format_reward": 0.0,
|
| 353 |
+
"rewards/tag_count_reward": 0.3940972238779068,
|
| 354 |
+
"step": 23
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"clip_ratio": 0.0,
|
| 358 |
+
"completion_length": 872.4722290039062,
|
| 359 |
+
"epoch": 0.45933014354066987,
|
| 360 |
+
"grad_norm": 0.05489657330998097,
|
| 361 |
+
"kl": 0.0005602836608886719,
|
| 362 |
+
"learning_rate": 9.573015282626445e-07,
|
| 363 |
+
"loss": 0.0217,
|
| 364 |
+
"reward": 0.6996527910232544,
|
| 365 |
+
"reward_std": 0.22097085788846016,
|
| 366 |
+
"rewards/accuracy_reward": 0.274305559694767,
|
| 367 |
+
"rewards/format_reward": 0.0,
|
| 368 |
+
"rewards/tag_count_reward": 0.4253472313284874,
|
| 369 |
+
"step": 24
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"clip_ratio": 0.0,
|
| 373 |
+
"completion_length": 864.3889007568359,
|
| 374 |
+
"epoch": 0.4784688995215311,
|
| 375 |
+
"grad_norm": 0.06394072732120833,
|
| 376 |
+
"kl": 0.0007648468017578125,
|
| 377 |
+
"learning_rate": 9.506073028637582e-07,
|
| 378 |
+
"loss": 0.0165,
|
| 379 |
+
"reward": 0.6866319477558136,
|
| 380 |
+
"reward_std": 0.26148219406604767,
|
| 381 |
+
"rewards/accuracy_reward": 0.274305559694767,
|
| 382 |
+
"rewards/format_reward": 0.0,
|
| 383 |
+
"rewards/tag_count_reward": 0.4123263880610466,
|
| 384 |
+
"step": 25
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"clip_ratio": 0.0,
|
| 388 |
+
"completion_length": 859.4375152587891,
|
| 389 |
+
"epoch": 0.49760765550239233,
|
| 390 |
+
"grad_norm": 0.06801792884391539,
|
| 391 |
+
"kl": 0.0011510848999023438,
|
| 392 |
+
"learning_rate": 9.434559772650618e-07,
|
| 393 |
+
"loss": 0.0179,
|
| 394 |
+
"reward": 0.7569444477558136,
|
| 395 |
+
"reward_std": 0.2357022576034069,
|
| 396 |
+
"rewards/accuracy_reward": 0.329861119389534,
|
| 397 |
+
"rewards/format_reward": 0.0,
|
| 398 |
+
"rewards/tag_count_reward": 0.4270833358168602,
|
| 399 |
+
"step": 26
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"clip_ratio": 0.0,
|
| 403 |
+
"completion_length": 882.1701354980469,
|
| 404 |
+
"epoch": 0.5167464114832536,
|
| 405 |
+
"grad_norm": 0.06594394459261921,
|
| 406 |
+
"kl": 0.001209259033203125,
|
| 407 |
+
"learning_rate": 9.358557112587928e-07,
|
| 408 |
+
"loss": 0.0201,
|
| 409 |
+
"reward": 0.6857638955116272,
|
| 410 |
+
"reward_std": 0.2651650384068489,
|
| 411 |
+
"rewards/accuracy_reward": 0.2777777761220932,
|
| 412 |
+
"rewards/format_reward": 0.0,
|
| 413 |
+
"rewards/tag_count_reward": 0.407986119389534,
|
| 414 |
+
"step": 27
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"clip_ratio": 0.0,
|
| 418 |
+
"completion_length": 848.2395935058594,
|
| 419 |
+
"epoch": 0.5358851674641149,
|
| 420 |
+
"grad_norm": 0.0697274048868991,
|
| 421 |
+
"kl": 0.001598358154296875,
|
| 422 |
+
"learning_rate": 9.278151768863375e-07,
|
| 423 |
+
"loss": 0.029,
|
| 424 |
+
"reward": 0.7039930671453476,
|
| 425 |
+
"reward_std": 0.2835792750120163,
|
| 426 |
+
"rewards/accuracy_reward": 0.2881944440305233,
|
| 427 |
+
"rewards/format_reward": 0.0,
|
| 428 |
+
"rewards/tag_count_reward": 0.4157986119389534,
|
| 429 |
+
"step": 28
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"clip_ratio": 0.0,
|
| 433 |
+
"completion_length": 858.8889007568359,
|
| 434 |
+
"epoch": 0.5550239234449761,
|
| 435 |
+
"grad_norm": 0.06562680518580778,
|
| 436 |
+
"kl": 0.0018520355224609375,
|
| 437 |
+
"learning_rate": 9.193435485432744e-07,
|
| 438 |
+
"loss": 0.0227,
|
| 439 |
+
"reward": 0.7395833283662796,
|
| 440 |
+
"reward_std": 0.2676202692091465,
|
| 441 |
+
"rewards/accuracy_reward": 0.3159722313284874,
|
| 442 |
+
"rewards/format_reward": 0.0,
|
| 443 |
+
"rewards/tag_count_reward": 0.423611119389534,
|
| 444 |
+
"step": 29
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"clip_ratio": 0.0,
|
| 448 |
+
"completion_length": 809.0694580078125,
|
| 449 |
+
"epoch": 0.5741626794258373,
|
| 450 |
+
"grad_norm": 0.07438927904369452,
|
| 451 |
+
"kl": 0.002056121826171875,
|
| 452 |
+
"learning_rate": 9.104504925112231e-07,
|
| 453 |
+
"loss": 0.0256,
|
| 454 |
+
"reward": 0.8541666716337204,
|
| 455 |
+
"reward_std": 0.2749859616160393,
|
| 456 |
+
"rewards/accuracy_reward": 0.399305559694767,
|
| 457 |
+
"rewards/format_reward": 0.0,
|
| 458 |
+
"rewards/tag_count_reward": 0.454861119389534,
|
| 459 |
+
"step": 30
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"clip_ratio": 0.0,
|
| 463 |
+
"completion_length": 871.6979217529297,
|
| 464 |
+
"epoch": 0.5933014354066986,
|
| 465 |
+
"grad_norm": 0.049564487099522436,
|
| 466 |
+
"kl": 0.00186920166015625,
|
| 467 |
+
"learning_rate": 9.011461559284413e-07,
|
| 468 |
+
"loss": 0.0095,
|
| 469 |
+
"reward": 0.7387152910232544,
|
| 470 |
+
"reward_std": 0.15099675580859184,
|
| 471 |
+
"rewards/accuracy_reward": 0.3402777835726738,
|
| 472 |
+
"rewards/format_reward": 0.0,
|
| 473 |
+
"rewards/tag_count_reward": 0.3984375074505806,
|
| 474 |
+
"step": 31
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"clip_ratio": 0.0,
|
| 478 |
+
"completion_length": 834.6250152587891,
|
| 479 |
+
"epoch": 0.6124401913875598,
|
| 480 |
+
"grad_norm": 0.06896614921110253,
|
| 481 |
+
"kl": 0.0023975372314453125,
|
| 482 |
+
"learning_rate": 8.914411552117559e-07,
|
| 483 |
+
"loss": 0.0227,
|
| 484 |
+
"reward": 0.7560763955116272,
|
| 485 |
+
"reward_std": 0.26884788274765015,
|
| 486 |
+
"rewards/accuracy_reward": 0.336805559694767,
|
| 487 |
+
"rewards/format_reward": 0.0,
|
| 488 |
+
"rewards/tag_count_reward": 0.4192708358168602,
|
| 489 |
+
"step": 32
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"clip_ratio": 0.0,
|
| 493 |
+
"completion_length": 845.7048645019531,
|
| 494 |
+
"epoch": 0.631578947368421,
|
| 495 |
+
"grad_norm": 0.07262913185512221,
|
| 496 |
+
"kl": 0.00289154052734375,
|
| 497 |
+
"learning_rate": 8.813465639430386e-07,
|
| 498 |
+
"loss": 0.0293,
|
| 499 |
+
"reward": 0.7838541716337204,
|
| 500 |
+
"reward_std": 0.26393742114305496,
|
| 501 |
+
"rewards/accuracy_reward": 0.329861119389534,
|
| 502 |
+
"rewards/format_reward": 0.0,
|
| 503 |
+
"rewards/tag_count_reward": 0.4539930522441864,
|
| 504 |
+
"step": 33
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"clip_ratio": 0.0,
|
| 508 |
+
"completion_length": 825.7361145019531,
|
| 509 |
+
"epoch": 0.6507177033492823,
|
| 510 |
+
"grad_norm": 0.06923687691149574,
|
| 511 |
+
"kl": 0.002803802490234375,
|
| 512 |
+
"learning_rate": 8.70873900234049e-07,
|
| 513 |
+
"loss": 0.0156,
|
| 514 |
+
"reward": 0.8272569626569748,
|
| 515 |
+
"reward_std": 0.2590269446372986,
|
| 516 |
+
"rewards/accuracy_reward": 0.392361119389534,
|
| 517 |
+
"rewards/format_reward": 0.0,
|
| 518 |
+
"rewards/tag_count_reward": 0.4348958358168602,
|
| 519 |
+
"step": 34
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"clip_ratio": 0.0,
|
| 523 |
+
"completion_length": 813.9930725097656,
|
| 524 |
+
"epoch": 0.6698564593301436,
|
| 525 |
+
"grad_norm": 0.0634139401535218,
|
| 526 |
+
"kl": 0.002788543701171875,
|
| 527 |
+
"learning_rate": 8.600351135840589e-07,
|
| 528 |
+
"loss": 0.0179,
|
| 529 |
+
"reward": 0.8159722238779068,
|
| 530 |
+
"reward_std": 0.24306795746088028,
|
| 531 |
+
"rewards/accuracy_reward": 0.3715277835726738,
|
| 532 |
+
"rewards/format_reward": 0.0,
|
| 533 |
+
"rewards/tag_count_reward": 0.4444444477558136,
|
| 534 |
+
"step": 35
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"clip_ratio": 0.0,
|
| 538 |
+
"completion_length": 888.8159942626953,
|
| 539 |
+
"epoch": 0.6889952153110048,
|
| 540 |
+
"grad_norm": 0.06115251503429904,
|
| 541 |
+
"kl": 0.00292205810546875,
|
| 542 |
+
"learning_rate": 8.488425712452576e-07,
|
| 543 |
+
"loss": 0.0181,
|
| 544 |
+
"reward": 0.6588541716337204,
|
| 545 |
+
"reward_std": 0.2221984826028347,
|
| 546 |
+
"rewards/accuracy_reward": 0.2777777798473835,
|
| 547 |
+
"rewards/format_reward": 0.0,
|
| 548 |
+
"rewards/tag_count_reward": 0.3810763880610466,
|
| 549 |
+
"step": 36
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"clip_ratio": 0.0,
|
| 553 |
+
"completion_length": 837.6458435058594,
|
| 554 |
+
"epoch": 0.7081339712918661,
|
| 555 |
+
"grad_norm": 0.06174697280825877,
|
| 556 |
+
"kl": 0.002880096435546875,
|
| 557 |
+
"learning_rate": 8.373090441114925e-07,
|
| 558 |
+
"loss": 0.0255,
|
| 559 |
+
"reward": 0.7873263955116272,
|
| 560 |
+
"reward_std": 0.2197432555258274,
|
| 561 |
+
"rewards/accuracy_reward": 0.3541666716337204,
|
| 562 |
+
"rewards/format_reward": 0.0,
|
| 563 |
+
"rewards/tag_count_reward": 0.4331597238779068,
|
| 564 |
+
"step": 37
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"clip_ratio": 0.0,
|
| 568 |
+
"completion_length": 855.8090362548828,
|
| 569 |
+
"epoch": 0.7272727272727273,
|
| 570 |
+
"grad_norm": 0.07806365108296032,
|
| 571 |
+
"kl": 0.00281524658203125,
|
| 572 |
+
"learning_rate": 8.254476921464483e-07,
|
| 573 |
+
"loss": 0.0277,
|
| 574 |
+
"reward": 0.7621527910232544,
|
| 575 |
+
"reward_std": 0.24797841161489487,
|
| 576 |
+
"rewards/accuracy_reward": 0.3298611119389534,
|
| 577 |
+
"rewards/format_reward": 0.0,
|
| 578 |
+
"rewards/tag_count_reward": 0.4322916716337204,
|
| 579 |
+
"step": 38
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"clip_ratio": 0.0,
|
| 583 |
+
"completion_length": 857.6840515136719,
|
| 584 |
+
"epoch": 0.7464114832535885,
|
| 585 |
+
"grad_norm": 0.07203148213149006,
|
| 586 |
+
"kl": 0.003299713134765625,
|
| 587 |
+
"learning_rate": 8.132720493678884e-07,
|
| 588 |
+
"loss": 0.023,
|
| 589 |
+
"reward": 0.7803819328546524,
|
| 590 |
+
"reward_std": 0.24184033647179604,
|
| 591 |
+
"rewards/accuracy_reward": 0.3402777835726738,
|
| 592 |
+
"rewards/format_reward": 0.0,
|
| 593 |
+
"rewards/tag_count_reward": 0.4401041716337204,
|
| 594 |
+
"step": 39
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"clip_ratio": 0.0,
|
| 598 |
+
"completion_length": 838.0764007568359,
|
| 599 |
+
"epoch": 0.7655502392344498,
|
| 600 |
+
"grad_norm": 0.06474279173647825,
|
| 601 |
+
"kl": 0.00308990478515625,
|
| 602 |
+
"learning_rate": 8.007960084050963e-07,
|
| 603 |
+
"loss": 0.0212,
|
| 604 |
+
"reward": 0.8741319477558136,
|
| 605 |
+
"reward_std": 0.23447464406490326,
|
| 606 |
+
"rewards/accuracy_reward": 0.4340277761220932,
|
| 607 |
+
"rewards/format_reward": 0.0,
|
| 608 |
+
"rewards/tag_count_reward": 0.4401041716337204,
|
| 609 |
+
"step": 40
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"clip_ratio": 0.0,
|
| 613 |
+
"completion_length": 859.78125,
|
| 614 |
+
"epoch": 0.784688995215311,
|
| 615 |
+
"grad_norm": 0.06299207628008677,
|
| 616 |
+
"kl": 0.003017425537109375,
|
| 617 |
+
"learning_rate": 7.88033804647133e-07,
|
| 618 |
+
"loss": 0.0193,
|
| 619 |
+
"reward": 0.7256944626569748,
|
| 620 |
+
"reward_std": 0.24306795746088028,
|
| 621 |
+
"rewards/accuracy_reward": 0.3159722238779068,
|
| 622 |
+
"rewards/format_reward": 0.0,
|
| 623 |
+
"rewards/tag_count_reward": 0.4097222238779068,
|
| 624 |
+
"step": 41
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"clip_ratio": 0.0,
|
| 628 |
+
"completion_length": 840.0625152587891,
|
| 629 |
+
"epoch": 0.8038277511961722,
|
| 630 |
+
"grad_norm": 0.07344429258616171,
|
| 631 |
+
"kl": 0.003047943115234375,
|
| 632 |
+
"learning_rate": 7.75e-07,
|
| 633 |
+
"loss": 0.0251,
|
| 634 |
+
"reward": 0.8220486044883728,
|
| 635 |
+
"reward_std": 0.20992231741547585,
|
| 636 |
+
"rewards/accuracy_reward": 0.392361119389534,
|
| 637 |
+
"rewards/format_reward": 0.0,
|
| 638 |
+
"rewards/tag_count_reward": 0.4296875074505806,
|
| 639 |
+
"step": 42
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"clip_ratio": 0.0,
|
| 643 |
+
"completion_length": 819.0382232666016,
|
| 644 |
+
"epoch": 0.8229665071770335,
|
| 645 |
+
"grad_norm": 0.061499395941431584,
|
| 646 |
+
"kl": 0.003070831298828125,
|
| 647 |
+
"learning_rate": 7.6170946627124e-07,
|
| 648 |
+
"loss": 0.0223,
|
| 649 |
+
"reward": 0.8151041716337204,
|
| 650 |
+
"reward_std": 0.20992232859134674,
|
| 651 |
+
"rewards/accuracy_reward": 0.3784722164273262,
|
| 652 |
+
"rewards/format_reward": 0.0,
|
| 653 |
+
"rewards/tag_count_reward": 0.4366319477558136,
|
| 654 |
+
"step": 43
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"clip_ratio": 0.0,
|
| 658 |
+
"completion_length": 863.3264007568359,
|
| 659 |
+
"epoch": 0.8421052631578947,
|
| 660 |
+
"grad_norm": 0.06745163743583922,
|
| 661 |
+
"kl": 0.0031585693359375,
|
| 662 |
+
"learning_rate": 7.481773682009354e-07,
|
| 663 |
+
"loss": 0.0196,
|
| 664 |
+
"reward": 0.7387152910232544,
|
| 665 |
+
"reward_std": 0.20501185581088066,
|
| 666 |
+
"rewards/accuracy_reward": 0.3159722238779068,
|
| 667 |
+
"rewards/format_reward": 0.0,
|
| 668 |
+
"rewards/tag_count_reward": 0.422743059694767,
|
| 669 |
+
"step": 44
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"clip_ratio": 0.0,
|
| 673 |
+
"completion_length": 830.9930725097656,
|
| 674 |
+
"epoch": 0.861244019138756,
|
| 675 |
+
"grad_norm": 0.07628628253732472,
|
| 676 |
+
"kl": 0.0036468505859375,
|
| 677 |
+
"learning_rate": 7.344191461584643e-07,
|
| 678 |
+
"loss": 0.0256,
|
| 679 |
+
"reward": 0.8289930671453476,
|
| 680 |
+
"reward_std": 0.2492060288786888,
|
| 681 |
+
"rewards/accuracy_reward": 0.3819444477558136,
|
| 682 |
+
"rewards/format_reward": 0.0,
|
| 683 |
+
"rewards/tag_count_reward": 0.4470486119389534,
|
| 684 |
+
"step": 45
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"clip_ratio": 0.0,
|
| 688 |
+
"completion_length": 834.5590362548828,
|
| 689 |
+
"epoch": 0.8803827751196173,
|
| 690 |
+
"grad_norm": 0.0697101918978914,
|
| 691 |
+
"kl": 0.004314422607421875,
|
| 692 |
+
"learning_rate": 7.204504985247592e-07,
|
| 693 |
+
"loss": 0.0163,
|
| 694 |
+
"reward": 0.8498263955116272,
|
| 695 |
+
"reward_std": 0.24184034019708633,
|
| 696 |
+
"rewards/accuracy_reward": 0.40625,
|
| 697 |
+
"rewards/format_reward": 0.0,
|
| 698 |
+
"rewards/tag_count_reward": 0.4435763955116272,
|
| 699 |
+
"step": 46
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"clip_ratio": 0.0,
|
| 703 |
+
"completion_length": 846.2569427490234,
|
| 704 |
+
"epoch": 0.8995215311004785,
|
| 705 |
+
"grad_norm": 0.06859248959883628,
|
| 706 |
+
"kl": 0.003566741943359375,
|
| 707 |
+
"learning_rate": 7.06287363780169e-07,
|
| 708 |
+
"loss": 0.0134,
|
| 709 |
+
"reward": 0.7534722238779068,
|
| 710 |
+
"reward_std": 0.22097086533904076,
|
| 711 |
+
"rewards/accuracy_reward": 0.3333333358168602,
|
| 712 |
+
"rewards/format_reward": 0.0,
|
| 713 |
+
"rewards/tag_count_reward": 0.4201388955116272,
|
| 714 |
+
"step": 47
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"clip_ratio": 0.0,
|
| 718 |
+
"completion_length": 813.9062652587891,
|
| 719 |
+
"epoch": 0.9186602870813397,
|
| 720 |
+
"grad_norm": 0.07616625495450065,
|
| 721 |
+
"kl": 0.00397491455078125,
|
| 722 |
+
"learning_rate": 6.91945902318365e-07,
|
| 723 |
+
"loss": 0.0284,
|
| 724 |
+
"reward": 0.862847238779068,
|
| 725 |
+
"reward_std": 0.2700754962861538,
|
| 726 |
+
"rewards/accuracy_reward": 0.4131944477558136,
|
| 727 |
+
"rewards/format_reward": 0.0,
|
| 728 |
+
"rewards/tag_count_reward": 0.4496527761220932,
|
| 729 |
+
"step": 48
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"clip_ratio": 0.0,
|
| 733 |
+
"completion_length": 843.1215209960938,
|
| 734 |
+
"epoch": 0.937799043062201,
|
| 735 |
+
"grad_norm": 0.06987264134329038,
|
| 736 |
+
"kl": 0.003589630126953125,
|
| 737 |
+
"learning_rate": 6.774424780070367e-07,
|
| 738 |
+
"loss": 0.0194,
|
| 739 |
+
"reward": 0.8289930671453476,
|
| 740 |
+
"reward_std": 0.2786688208580017,
|
| 741 |
+
"rewards/accuracy_reward": 0.4027777761220932,
|
| 742 |
+
"rewards/format_reward": 0.0,
|
| 743 |
+
"rewards/tag_count_reward": 0.4262152835726738,
|
| 744 |
+
"step": 49
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"clip_ratio": 0.0,
|
| 748 |
+
"completion_length": 848.1909790039062,
|
| 749 |
+
"epoch": 0.9569377990430622,
|
| 750 |
+
"grad_norm": 0.07535010112426463,
|
| 751 |
+
"kl": 0.00441741943359375,
|
| 752 |
+
"learning_rate": 6.627936395164242e-07,
|
| 753 |
+
"loss": 0.0225,
|
| 754 |
+
"reward": 0.8229166716337204,
|
| 755 |
+
"reward_std": 0.27744119614362717,
|
| 756 |
+
"rewards/accuracy_reward": 0.3645833358168602,
|
| 757 |
+
"rewards/format_reward": 0.0,
|
| 758 |
+
"rewards/tag_count_reward": 0.4583333358168602,
|
| 759 |
+
"step": 50
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"clip_ratio": 0.0,
|
| 763 |
+
"completion_length": 844.8402862548828,
|
| 764 |
+
"epoch": 0.9760765550239234,
|
| 765 |
+
"grad_norm": 0.07005632661144727,
|
| 766 |
+
"kl": 0.004360198974609375,
|
| 767 |
+
"learning_rate": 6.480161014369832e-07,
|
| 768 |
+
"loss": 0.0217,
|
| 769 |
+
"reward": 0.8645833283662796,
|
| 770 |
+
"reward_std": 0.25043364986777306,
|
| 771 |
+
"rewards/accuracy_reward": 0.4270833358168602,
|
| 772 |
+
"rewards/format_reward": 0.0,
|
| 773 |
+
"rewards/tag_count_reward": 0.4375,
|
| 774 |
+
"step": 51
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"clip_ratio": 0.0,
|
| 778 |
+
"completion_length": 859.1458587646484,
|
| 779 |
+
"epoch": 0.9952153110047847,
|
| 780 |
+
"grad_norm": 0.07899639612660056,
|
| 781 |
+
"kl": 0.004245758056640625,
|
| 782 |
+
"learning_rate": 6.331267252077352e-07,
|
| 783 |
+
"loss": 0.0249,
|
| 784 |
+
"reward": 0.8576388955116272,
|
| 785 |
+
"reward_std": 0.25043364241719246,
|
| 786 |
+
"rewards/accuracy_reward": 0.423611119389534,
|
| 787 |
+
"rewards/format_reward": 0.0,
|
| 788 |
+
"rewards/tag_count_reward": 0.4340277761220932,
|
| 789 |
+
"step": 52
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"clip_ratio": 0.0,
|
| 793 |
+
"completion_length": 818.5764007568359,
|
| 794 |
+
"epoch": 1.0191387559808613,
|
| 795 |
+
"grad_norm": 0.06800368909135272,
|
| 796 |
+
"kl": 0.0054168701171875,
|
| 797 |
+
"learning_rate": 6.181424998770595e-07,
|
| 798 |
+
"loss": 0.0099,
|
| 799 |
+
"reward": 0.9149305522441864,
|
| 800 |
+
"reward_std": 0.25534411147236824,
|
| 801 |
+
"rewards/accuracy_reward": 0.444444440305233,
|
| 802 |
+
"rewards/format_reward": 0.0,
|
| 803 |
+
"rewards/tag_count_reward": 0.4704861044883728,
|
| 804 |
+
"step": 53
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"clip_ratio": 0.0,
|
| 808 |
+
"completion_length": 825.7847290039062,
|
| 809 |
+
"epoch": 1.0382775119617225,
|
| 810 |
+
"grad_norm": 0.0687845544933541,
|
| 811 |
+
"kl": 0.00484466552734375,
|
| 812 |
+
"learning_rate": 6.030805227178808e-07,
|
| 813 |
+
"loss": 0.0266,
|
| 814 |
+
"reward": 0.8932291716337204,
|
| 815 |
+
"reward_std": 0.27130311727523804,
|
| 816 |
+
"rewards/accuracy_reward": 0.4236111119389534,
|
| 817 |
+
"rewards/format_reward": 0.0,
|
| 818 |
+
"rewards/tag_count_reward": 0.469618059694767,
|
| 819 |
+
"step": 54
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"clip_ratio": 0.0,
|
| 823 |
+
"completion_length": 831.6875152587891,
|
| 824 |
+
"epoch": 1.0574162679425838,
|
| 825 |
+
"grad_norm": 0.06231563710215865,
|
| 826 |
+
"kl": 0.0048065185546875,
|
| 827 |
+
"learning_rate": 5.879579797193701e-07,
|
| 828 |
+
"loss": 0.0051,
|
| 829 |
+
"reward": 0.8385416716337204,
|
| 830 |
+
"reward_std": 0.23324701189994812,
|
| 831 |
+
"rewards/accuracy_reward": 0.3958333358168602,
|
| 832 |
+
"rewards/format_reward": 0.0,
|
| 833 |
+
"rewards/tag_count_reward": 0.4427083358168602,
|
| 834 |
+
"step": 55
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"clip_ratio": 0.0,
|
| 838 |
+
"completion_length": 813.4722290039062,
|
| 839 |
+
"epoch": 1.076555023923445,
|
| 840 |
+
"grad_norm": 0.07317349080061789,
|
| 841 |
+
"kl": 0.0054473876953125,
|
| 842 |
+
"learning_rate": 5.727921259774208e-07,
|
| 843 |
+
"loss": 0.0342,
|
| 844 |
+
"reward": 0.8819444626569748,
|
| 845 |
+
"reward_std": 0.25043365359306335,
|
| 846 |
+
"rewards/accuracy_reward": 0.4236111268401146,
|
| 847 |
+
"rewards/format_reward": 0.0,
|
| 848 |
+
"rewards/tag_count_reward": 0.4583333432674408,
|
| 849 |
+
"step": 56
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"clip_ratio": 0.0,
|
| 853 |
+
"completion_length": 802.3784942626953,
|
| 854 |
+
"epoch": 1.0956937799043063,
|
| 855 |
+
"grad_norm": 0.07553172071799115,
|
| 856 |
+
"kl": 0.00548553466796875,
|
| 857 |
+
"learning_rate": 5.57600266006269e-07,
|
| 858 |
+
"loss": 0.0246,
|
| 859 |
+
"reward": 0.9357638955116272,
|
| 860 |
+
"reward_std": 0.2970830574631691,
|
| 861 |
+
"rewards/accuracy_reward": 0.4583333432674408,
|
| 862 |
+
"rewards/format_reward": 0.0,
|
| 863 |
+
"rewards/tag_count_reward": 0.4774305671453476,
|
| 864 |
+
"step": 57
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"clip_ratio": 0.0,
|
| 868 |
+
"completion_length": 809.1076354980469,
|
| 869 |
+
"epoch": 1.1148325358851674,
|
| 870 |
+
"grad_norm": 0.08337759523447522,
|
| 871 |
+
"kl": 0.0051727294921875,
|
| 872 |
+
"learning_rate": 5.423997339937311e-07,
|
| 873 |
+
"loss": 0.0237,
|
| 874 |
+
"reward": 0.8854167014360428,
|
| 875 |
+
"reward_std": 0.3019935190677643,
|
| 876 |
+
"rewards/accuracy_reward": 0.4097222238779068,
|
| 877 |
+
"rewards/format_reward": 0.0,
|
| 878 |
+
"rewards/tag_count_reward": 0.4756944477558136,
|
| 879 |
+
"step": 58
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"clip_ratio": 0.0,
|
| 883 |
+
"completion_length": 813.4965362548828,
|
| 884 |
+
"epoch": 1.1339712918660287,
|
| 885 |
+
"grad_norm": 0.08335533132880214,
|
| 886 |
+
"kl": 0.006591796875,
|
| 887 |
+
"learning_rate": 5.272078740225792e-07,
|
| 888 |
+
"loss": 0.025,
|
| 889 |
+
"reward": 0.8862847238779068,
|
| 890 |
+
"reward_std": 0.26148218661546707,
|
| 891 |
+
"rewards/accuracy_reward": 0.4062500149011612,
|
| 892 |
+
"rewards/format_reward": 0.0,
|
| 893 |
+
"rewards/tag_count_reward": 0.4800347238779068,
|
| 894 |
+
"step": 59
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"clip_ratio": 0.0,
|
| 898 |
+
"completion_length": 799.8507232666016,
|
| 899 |
+
"epoch": 1.1531100478468899,
|
| 900 |
+
"grad_norm": 0.07203309192614081,
|
| 901 |
+
"kl": 0.0055999755859375,
|
| 902 |
+
"learning_rate": 5.1204202028063e-07,
|
| 903 |
+
"loss": 0.0212,
|
| 904 |
+
"reward": 0.9774305373430252,
|
| 905 |
+
"reward_std": 0.2627098001539707,
|
| 906 |
+
"rewards/accuracy_reward": 0.4930555745959282,
|
| 907 |
+
"rewards/format_reward": 0.0,
|
| 908 |
+
"rewards/tag_count_reward": 0.4843750149011612,
|
| 909 |
+
"step": 60
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"clip_ratio": 0.0,
|
| 913 |
+
"completion_length": 791.4027862548828,
|
| 914 |
+
"epoch": 1.1722488038277512,
|
| 915 |
+
"grad_norm": 0.07531690542769269,
|
| 916 |
+
"kl": 0.00594329833984375,
|
| 917 |
+
"learning_rate": 4.969194772821193e-07,
|
| 918 |
+
"loss": 0.0127,
|
| 919 |
+
"reward": 0.981770858168602,
|
| 920 |
+
"reward_std": 0.26884788647294044,
|
| 921 |
+
"rewards/accuracy_reward": 0.486111119389534,
|
| 922 |
+
"rewards/format_reward": 0.0,
|
| 923 |
+
"rewards/tag_count_reward": 0.4956597313284874,
|
| 924 |
+
"step": 61
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"clip_ratio": 0.0,
|
| 928 |
+
"completion_length": 856.4513854980469,
|
| 929 |
+
"epoch": 1.1913875598086126,
|
| 930 |
+
"grad_norm": 0.062172711166539586,
|
| 931 |
+
"kl": 0.00542449951171875,
|
| 932 |
+
"learning_rate": 4.818575001229406e-07,
|
| 933 |
+
"loss": 0.0132,
|
| 934 |
+
"reward": 0.8255208432674408,
|
| 935 |
+
"reward_std": 0.20255662500858307,
|
| 936 |
+
"rewards/accuracy_reward": 0.3819444477558136,
|
| 937 |
+
"rewards/format_reward": 0.0,
|
| 938 |
+
"rewards/tag_count_reward": 0.4435763955116272,
|
| 939 |
+
"step": 62
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"clip_ratio": 0.0,
|
| 943 |
+
"completion_length": 865.0312652587891,
|
| 944 |
+
"epoch": 1.2105263157894737,
|
| 945 |
+
"grad_norm": 0.056253844990592314,
|
| 946 |
+
"kl": 0.00551605224609375,
|
| 947 |
+
"learning_rate": 4.668732747922649e-07,
|
| 948 |
+
"loss": 0.0133,
|
| 949 |
+
"reward": 0.767361119389534,
|
| 950 |
+
"reward_std": 0.19641854986548424,
|
| 951 |
+
"rewards/accuracy_reward": 0.3472222313284874,
|
| 952 |
+
"rewards/format_reward": 0.0,
|
| 953 |
+
"rewards/tag_count_reward": 0.4201389029622078,
|
| 954 |
+
"step": 63
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"clip_ratio": 0.0,
|
| 958 |
+
"completion_length": 828.03125,
|
| 959 |
+
"epoch": 1.229665071770335,
|
| 960 |
+
"grad_norm": 0.06432066728138643,
|
| 961 |
+
"kl": 0.005279541015625,
|
| 962 |
+
"learning_rate": 4.5198389856301677e-07,
|
| 963 |
+
"loss": 0.0216,
|
| 964 |
+
"reward": 0.8559027761220932,
|
| 965 |
+
"reward_std": 0.23324701562523842,
|
| 966 |
+
"rewards/accuracy_reward": 0.4027777835726738,
|
| 967 |
+
"rewards/format_reward": 0.0,
|
| 968 |
+
"rewards/tag_count_reward": 0.4531250074505806,
|
| 969 |
+
"step": 64
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"clip_ratio": 0.0,
|
| 973 |
+
"completion_length": 827.1354217529297,
|
| 974 |
+
"epoch": 1.2488038277511961,
|
| 975 |
+
"grad_norm": 0.06953232592329361,
|
| 976 |
+
"kl": 0.005645751953125,
|
| 977 |
+
"learning_rate": 4.3720636048357573e-07,
|
| 978 |
+
"loss": 0.0207,
|
| 979 |
+
"reward": 0.8515625,
|
| 980 |
+
"reward_std": 0.20992231741547585,
|
| 981 |
+
"rewards/accuracy_reward": 0.399305559694767,
|
| 982 |
+
"rewards/format_reward": 0.0,
|
| 983 |
+
"rewards/tag_count_reward": 0.4522569477558136,
|
| 984 |
+
"step": 65
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"clip_ratio": 0.0,
|
| 988 |
+
"completion_length": 830.3020782470703,
|
| 989 |
+
"epoch": 1.2679425837320575,
|
| 990 |
+
"grad_norm": 0.06833410523166436,
|
| 991 |
+
"kl": 0.00586700439453125,
|
| 992 |
+
"learning_rate": 4.225575219929631e-07,
|
| 993 |
+
"loss": 0.0255,
|
| 994 |
+
"reward": 0.8984375149011612,
|
| 995 |
+
"reward_std": 0.29340019077062607,
|
| 996 |
+
"rewards/accuracy_reward": 0.416666679084301,
|
| 997 |
+
"rewards/format_reward": 0.0,
|
| 998 |
+
"rewards/tag_count_reward": 0.4817708283662796,
|
| 999 |
+
"step": 66
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"clip_ratio": 0.0,
|
| 1003 |
+
"completion_length": 837.3264007568359,
|
| 1004 |
+
"epoch": 1.2870813397129186,
|
| 1005 |
+
"grad_norm": 0.06984826024421593,
|
| 1006 |
+
"kl": 0.00540924072265625,
|
| 1007 |
+
"learning_rate": 4.0805409768163504e-07,
|
| 1008 |
+
"loss": 0.0133,
|
| 1009 |
+
"reward": 0.864583358168602,
|
| 1010 |
+
"reward_std": 0.28726211935281754,
|
| 1011 |
+
"rewards/accuracy_reward": 0.4027777910232544,
|
| 1012 |
+
"rewards/format_reward": 0.0,
|
| 1013 |
+
"rewards/tag_count_reward": 0.4618055522441864,
|
| 1014 |
+
"step": 67
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"clip_ratio": 0.0,
|
| 1018 |
+
"completion_length": 813.1770782470703,
|
| 1019 |
+
"epoch": 1.30622009569378,
|
| 1020 |
+
"grad_norm": 0.07284090789531786,
|
| 1021 |
+
"kl": 0.00634765625,
|
| 1022 |
+
"learning_rate": 3.9371263621983097e-07,
|
| 1023 |
+
"loss": 0.0162,
|
| 1024 |
+
"reward": 0.8993055671453476,
|
| 1025 |
+
"reward_std": 0.27744119614362717,
|
| 1026 |
+
"rewards/accuracy_reward": 0.4270833358168602,
|
| 1027 |
+
"rewards/format_reward": 0.0,
|
| 1028 |
+
"rewards/tag_count_reward": 0.4722222313284874,
|
| 1029 |
+
"step": 68
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"clip_ratio": 0.0,
|
| 1033 |
+
"completion_length": 816.5382080078125,
|
| 1034 |
+
"epoch": 1.325358851674641,
|
| 1035 |
+
"grad_norm": 0.07703787035943734,
|
| 1036 |
+
"kl": 0.0065155029296875,
|
| 1037 |
+
"learning_rate": 3.795495014752408e-07,
|
| 1038 |
+
"loss": 0.0256,
|
| 1039 |
+
"reward": 0.956597238779068,
|
| 1040 |
+
"reward_std": 0.3044487535953522,
|
| 1041 |
+
"rewards/accuracy_reward": 0.461805559694767,
|
| 1042 |
+
"rewards/format_reward": 0.0,
|
| 1043 |
+
"rewards/tag_count_reward": 0.4947916641831398,
|
| 1044 |
+
"step": 69
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"clip_ratio": 0.0,
|
| 1048 |
+
"completion_length": 803.5173645019531,
|
| 1049 |
+
"epoch": 1.3444976076555024,
|
| 1050 |
+
"grad_norm": 0.07147321350786168,
|
| 1051 |
+
"kl": 0.0061187744140625,
|
| 1052 |
+
"learning_rate": 3.655808538415357e-07,
|
| 1053 |
+
"loss": 0.026,
|
| 1054 |
+
"reward": 0.9496528059244156,
|
| 1055 |
+
"reward_std": 0.2700754851102829,
|
| 1056 |
+
"rewards/accuracy_reward": 0.4687500074505806,
|
| 1057 |
+
"rewards/format_reward": 0.0,
|
| 1058 |
+
"rewards/tag_count_reward": 0.4809027761220932,
|
| 1059 |
+
"step": 70
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"clip_ratio": 0.0,
|
| 1063 |
+
"completion_length": 849.3402709960938,
|
| 1064 |
+
"epoch": 1.3636363636363638,
|
| 1065 |
+
"grad_norm": 0.062268739561404315,
|
| 1066 |
+
"kl": 0.00612640380859375,
|
| 1067 |
+
"learning_rate": 3.5182263179906457e-07,
|
| 1068 |
+
"loss": 0.0233,
|
| 1069 |
+
"reward": 0.8515624850988388,
|
| 1070 |
+
"reward_std": 0.23692986741662025,
|
| 1071 |
+
"rewards/accuracy_reward": 0.3993055671453476,
|
| 1072 |
+
"rewards/format_reward": 0.0,
|
| 1073 |
+
"rewards/tag_count_reward": 0.4522569477558136,
|
| 1074 |
+
"step": 71
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"clip_ratio": 0.0,
|
| 1078 |
+
"completion_length": 827.8125,
|
| 1079 |
+
"epoch": 1.3827751196172249,
|
| 1080 |
+
"grad_norm": 0.07644202407601894,
|
| 1081 |
+
"kl": 0.00571441650390625,
|
| 1082 |
+
"learning_rate": 3.3829053372876006e-07,
|
| 1083 |
+
"loss": 0.029,
|
| 1084 |
+
"reward": 0.8593750149011612,
|
| 1085 |
+
"reward_std": 0.2700754962861538,
|
| 1086 |
+
"rewards/accuracy_reward": 0.3888888955116272,
|
| 1087 |
+
"rewards/format_reward": 0.0,
|
| 1088 |
+
"rewards/tag_count_reward": 0.4704861119389534,
|
| 1089 |
+
"step": 72
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"clip_ratio": 0.0,
|
| 1093 |
+
"completion_length": 824.4514007568359,
|
| 1094 |
+
"epoch": 1.401913875598086,
|
| 1095 |
+
"grad_norm": 0.07665948925496174,
|
| 1096 |
+
"kl": 0.0063323974609375,
|
| 1097 |
+
"learning_rate": 3.250000000000001e-07,
|
| 1098 |
+
"loss": 0.025,
|
| 1099 |
+
"reward": 0.9288194477558136,
|
| 1100 |
+
"reward_std": 0.2995382845401764,
|
| 1101 |
+
"rewards/accuracy_reward": 0.4375,
|
| 1102 |
+
"rewards/format_reward": 0.0,
|
| 1103 |
+
"rewards/tag_count_reward": 0.4913194477558136,
|
| 1104 |
+
"step": 73
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"clip_ratio": 0.0,
|
| 1108 |
+
"completion_length": 798.7048645019531,
|
| 1109 |
+
"epoch": 1.4210526315789473,
|
| 1110 |
+
"grad_norm": 0.06940993863431441,
|
| 1111 |
+
"kl": 0.0066375732421875,
|
| 1112 |
+
"learning_rate": 3.1196619535286707e-07,
|
| 1113 |
+
"loss": 0.0166,
|
| 1114 |
+
"reward": 0.9149305522441864,
|
| 1115 |
+
"reward_std": 0.21360516920685768,
|
| 1116 |
+
"rewards/accuracy_reward": 0.423611119389534,
|
| 1117 |
+
"rewards/format_reward": 0.0,
|
| 1118 |
+
"rewards/tag_count_reward": 0.491319440305233,
|
| 1119 |
+
"step": 74
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"clip_ratio": 0.0,
|
| 1123 |
+
"completion_length": 828.0729217529297,
|
| 1124 |
+
"epoch": 1.4401913875598087,
|
| 1125 |
+
"grad_norm": 0.07296376411015054,
|
| 1126 |
+
"kl": 0.006256103515625,
|
| 1127 |
+
"learning_rate": 2.992039915949036e-07,
|
| 1128 |
+
"loss": 0.0193,
|
| 1129 |
+
"reward": 0.9097222238779068,
|
| 1130 |
+
"reward_std": 0.24552317708730698,
|
| 1131 |
+
"rewards/accuracy_reward": 0.430555559694767,
|
| 1132 |
+
"rewards/format_reward": 0.0,
|
| 1133 |
+
"rewards/tag_count_reward": 0.4791666716337204,
|
| 1134 |
+
"step": 75
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"clip_ratio": 0.0,
|
| 1138 |
+
"completion_length": 816.3090362548828,
|
| 1139 |
+
"epoch": 1.4593301435406698,
|
| 1140 |
+
"grad_norm": 0.07617903452488152,
|
| 1141 |
+
"kl": 0.0065765380859375,
|
| 1142 |
+
"learning_rate": 2.867279506321115e-07,
|
| 1143 |
+
"loss": 0.017,
|
| 1144 |
+
"reward": 0.8394097238779068,
|
| 1145 |
+
"reward_std": 0.2835792824625969,
|
| 1146 |
+
"rewards/accuracy_reward": 0.3715277835726738,
|
| 1147 |
+
"rewards/format_reward": 0.0,
|
| 1148 |
+
"rewards/tag_count_reward": 0.4678819477558136,
|
| 1149 |
+
"step": 76
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"clip_ratio": 0.0,
|
| 1153 |
+
"completion_length": 841.1701507568359,
|
| 1154 |
+
"epoch": 1.4784688995215312,
|
| 1155 |
+
"grad_norm": 0.0675783715098431,
|
| 1156 |
+
"kl": 0.0067291259765625,
|
| 1157 |
+
"learning_rate": 2.7455230785355164e-07,
|
| 1158 |
+
"loss": 0.0299,
|
| 1159 |
+
"reward": 0.8255208432674408,
|
| 1160 |
+
"reward_std": 0.20010139793157578,
|
| 1161 |
+
"rewards/accuracy_reward": 0.3750000074505806,
|
| 1162 |
+
"rewards/format_reward": 0.0,
|
| 1163 |
+
"rewards/tag_count_reward": 0.4505208358168602,
|
| 1164 |
+
"step": 77
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"clip_ratio": 0.0,
|
| 1168 |
+
"completion_length": 863.5416717529297,
|
| 1169 |
+
"epoch": 1.4976076555023923,
|
| 1170 |
+
"grad_norm": 0.06878244801121526,
|
| 1171 |
+
"kl": 0.00676727294921875,
|
| 1172 |
+
"learning_rate": 2.6269095588850744e-07,
|
| 1173 |
+
"loss": 0.0208,
|
| 1174 |
+
"reward": 0.8307291865348816,
|
| 1175 |
+
"reward_std": 0.26148219406604767,
|
| 1176 |
+
"rewards/accuracy_reward": 0.378472238779068,
|
| 1177 |
+
"rewards/format_reward": 0.0,
|
| 1178 |
+
"rewards/tag_count_reward": 0.4522569552063942,
|
| 1179 |
+
"step": 78
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"clip_ratio": 0.0,
|
| 1183 |
+
"completion_length": 827.4583435058594,
|
| 1184 |
+
"epoch": 1.5167464114832536,
|
| 1185 |
+
"grad_norm": 0.07251844117312602,
|
| 1186 |
+
"kl": 0.007293701171875,
|
| 1187 |
+
"learning_rate": 2.511574287547423e-07,
|
| 1188 |
+
"loss": 0.0209,
|
| 1189 |
+
"reward": 0.884548619389534,
|
| 1190 |
+
"reward_std": 0.281124047935009,
|
| 1191 |
+
"rewards/accuracy_reward": 0.4131944552063942,
|
| 1192 |
+
"rewards/format_reward": 0.0,
|
| 1193 |
+
"rewards/tag_count_reward": 0.4713541716337204,
|
| 1194 |
+
"step": 79
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"clip_ratio": 0.0,
|
| 1198 |
+
"completion_length": 815.7916870117188,
|
| 1199 |
+
"epoch": 1.535885167464115,
|
| 1200 |
+
"grad_norm": 0.06598452519685939,
|
| 1201 |
+
"kl": 0.00701141357421875,
|
| 1202 |
+
"learning_rate": 2.39964886415941e-07,
|
| 1203 |
+
"loss": 0.0212,
|
| 1204 |
+
"reward": 0.862847238779068,
|
| 1205 |
+
"reward_std": 0.25043364614248276,
|
| 1206 |
+
"rewards/accuracy_reward": 0.399305559694767,
|
| 1207 |
+
"rewards/format_reward": 0.0,
|
| 1208 |
+
"rewards/tag_count_reward": 0.4635416641831398,
|
| 1209 |
+
"step": 80
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"clip_ratio": 0.0,
|
| 1213 |
+
"completion_length": 809.65625,
|
| 1214 |
+
"epoch": 1.555023923444976,
|
| 1215 |
+
"grad_norm": 0.07534072488970928,
|
| 1216 |
+
"kl": 0.0063323974609375,
|
| 1217 |
+
"learning_rate": 2.291260997659509e-07,
|
| 1218 |
+
"loss": 0.0184,
|
| 1219 |
+
"reward": 0.9218750149011612,
|
| 1220 |
+
"reward_std": 0.28480689227581024,
|
| 1221 |
+
"rewards/accuracy_reward": 0.4340277761220932,
|
| 1222 |
+
"rewards/format_reward": 0.0,
|
| 1223 |
+
"rewards/tag_count_reward": 0.4878472164273262,
|
| 1224 |
+
"step": 81
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"clip_ratio": 0.0,
|
| 1228 |
+
"completion_length": 849.3958282470703,
|
| 1229 |
+
"epoch": 1.5741626794258372,
|
| 1230 |
+
"grad_norm": 0.06432192299979189,
|
| 1231 |
+
"kl": 0.0068206787109375,
|
| 1232 |
+
"learning_rate": 2.1865343605696128e-07,
|
| 1233 |
+
"loss": 0.0169,
|
| 1234 |
+
"reward": 0.8359375149011612,
|
| 1235 |
+
"reward_std": 0.21728801727294922,
|
| 1236 |
+
"rewards/accuracy_reward": 0.3645833358168602,
|
| 1237 |
+
"rewards/format_reward": 0.0,
|
| 1238 |
+
"rewards/tag_count_reward": 0.4713541641831398,
|
| 1239 |
+
"step": 82
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"clip_ratio": 0.0,
|
| 1243 |
+
"completion_length": 839.3923492431641,
|
| 1244 |
+
"epoch": 1.5933014354066986,
|
| 1245 |
+
"grad_norm": 0.0750282166543299,
|
| 1246 |
+
"kl": 0.00699615478515625,
|
| 1247 |
+
"learning_rate": 2.085588447882441e-07,
|
| 1248 |
+
"loss": 0.0295,
|
| 1249 |
+
"reward": 0.8680555671453476,
|
| 1250 |
+
"reward_std": 0.30690397322177887,
|
| 1251 |
+
"rewards/accuracy_reward": 0.3923611119389534,
|
| 1252 |
+
"rewards/format_reward": 0.0,
|
| 1253 |
+
"rewards/tag_count_reward": 0.475694440305233,
|
| 1254 |
+
"step": 83
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"clip_ratio": 0.0,
|
| 1258 |
+
"completion_length": 862.53125,
|
| 1259 |
+
"epoch": 1.61244019138756,
|
| 1260 |
+
"grad_norm": 0.0705730634901919,
|
| 1261 |
+
"kl": 0.00653839111328125,
|
| 1262 |
+
"learning_rate": 1.9885384407155872e-07,
|
| 1263 |
+
"loss": 0.0264,
|
| 1264 |
+
"reward": 0.8081597238779068,
|
| 1265 |
+
"reward_std": 0.2737583592534065,
|
| 1266 |
+
"rewards/accuracy_reward": 0.3611111119389534,
|
| 1267 |
+
"rewards/format_reward": 0.0,
|
| 1268 |
+
"rewards/tag_count_reward": 0.4470486119389534,
|
| 1269 |
+
"step": 84
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"clip_ratio": 0.0,
|
| 1273 |
+
"completion_length": 832.2951354980469,
|
| 1274 |
+
"epoch": 1.631578947368421,
|
| 1275 |
+
"grad_norm": 0.07373687083533499,
|
| 1276 |
+
"kl": 0.00716400146484375,
|
| 1277 |
+
"learning_rate": 1.89549507488777e-07,
|
| 1278 |
+
"loss": 0.018,
|
| 1279 |
+
"reward": 0.9184027761220932,
|
| 1280 |
+
"reward_std": 0.22833656147122383,
|
| 1281 |
+
"rewards/accuracy_reward": 0.4305555671453476,
|
| 1282 |
+
"rewards/format_reward": 0.0,
|
| 1283 |
+
"rewards/tag_count_reward": 0.4878472238779068,
|
| 1284 |
+
"step": 85
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"clip_ratio": 0.0,
|
| 1288 |
+
"completion_length": 803.6944580078125,
|
| 1289 |
+
"epoch": 1.6507177033492821,
|
| 1290 |
+
"grad_norm": 0.06861588416484331,
|
| 1291 |
+
"kl": 0.007843017578125,
|
| 1292 |
+
"learning_rate": 1.8065645145672575e-07,
|
| 1293 |
+
"loss": 0.0224,
|
| 1294 |
+
"reward": 0.9375000149011612,
|
| 1295 |
+
"reward_std": 0.24552318453788757,
|
| 1296 |
+
"rewards/accuracy_reward": 0.4513888955116272,
|
| 1297 |
+
"rewards/format_reward": 0.0,
|
| 1298 |
+
"rewards/tag_count_reward": 0.4861111119389534,
|
| 1299 |
+
"step": 86
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"clip_ratio": 0.0,
|
| 1303 |
+
"completion_length": 832.5069580078125,
|
| 1304 |
+
"epoch": 1.6698564593301435,
|
| 1305 |
+
"grad_norm": 0.06855680549432348,
|
| 1306 |
+
"kl": 0.007110595703125,
|
| 1307 |
+
"learning_rate": 1.7218482311366245e-07,
|
| 1308 |
+
"loss": 0.0183,
|
| 1309 |
+
"reward": 0.8836805522441864,
|
| 1310 |
+
"reward_std": 0.2602545693516731,
|
| 1311 |
+
"rewards/accuracy_reward": 0.4131944477558136,
|
| 1312 |
+
"rewards/format_reward": 0.0,
|
| 1313 |
+
"rewards/tag_count_reward": 0.470486119389534,
|
| 1314 |
+
"step": 87
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"clip_ratio": 0.0,
|
| 1318 |
+
"completion_length": 819.6875152587891,
|
| 1319 |
+
"epoch": 1.6889952153110048,
|
| 1320 |
+
"grad_norm": 0.06513799373796432,
|
| 1321 |
+
"kl": 0.00678253173828125,
|
| 1322 |
+
"learning_rate": 1.64144288741207e-07,
|
| 1323 |
+
"loss": 0.027,
|
| 1324 |
+
"reward": 0.9244791865348816,
|
| 1325 |
+
"reward_std": 0.2590269595384598,
|
| 1326 |
+
"rewards/accuracy_reward": 0.447916679084301,
|
| 1327 |
+
"rewards/format_reward": 0.0,
|
| 1328 |
+
"rewards/tag_count_reward": 0.4765625,
|
| 1329 |
+
"step": 88
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"clip_ratio": 0.0,
|
| 1333 |
+
"completion_length": 807.8333435058594,
|
| 1334 |
+
"epoch": 1.7081339712918662,
|
| 1335 |
+
"grad_norm": 0.06838857628608602,
|
| 1336 |
+
"kl": 0.00734710693359375,
|
| 1337 |
+
"learning_rate": 1.5654402273493807e-07,
|
| 1338 |
+
"loss": 0.0204,
|
| 1339 |
+
"reward": 0.9175347238779068,
|
| 1340 |
+
"reward_std": 0.2516612708568573,
|
| 1341 |
+
"rewards/accuracy_reward": 0.4409722238779068,
|
| 1342 |
+
"rewards/format_reward": 0.0,
|
| 1343 |
+
"rewards/tag_count_reward": 0.4765624925494194,
|
| 1344 |
+
"step": 89
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"clip_ratio": 0.0,
|
| 1348 |
+
"completion_length": 830.8576507568359,
|
| 1349 |
+
"epoch": 1.7272727272727273,
|
| 1350 |
+
"grad_norm": 0.06707500592677755,
|
| 1351 |
+
"kl": 0.00732421875,
|
| 1352 |
+
"learning_rate": 1.4939269713624181e-07,
|
| 1353 |
+
"loss": 0.0171,
|
| 1354 |
+
"reward": 0.9131944626569748,
|
| 1355 |
+
"reward_std": 0.22833655402064323,
|
| 1356 |
+
"rewards/accuracy_reward": 0.4409722313284874,
|
| 1357 |
+
"rewards/format_reward": 0.0,
|
| 1358 |
+
"rewards/tag_count_reward": 0.4722222238779068,
|
| 1359 |
+
"step": 90
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"clip_ratio": 0.0,
|
| 1363 |
+
"completion_length": 805.3263854980469,
|
| 1364 |
+
"epoch": 1.7464114832535884,
|
| 1365 |
+
"grad_norm": 0.08133209124657947,
|
| 1366 |
+
"kl": 0.007537841796875,
|
| 1367 |
+
"learning_rate": 1.4269847173735543e-07,
|
| 1368 |
+
"loss": 0.0284,
|
| 1369 |
+
"reward": 0.942708358168602,
|
| 1370 |
+
"reward_std": 0.27989643067121506,
|
| 1371 |
+
"rewards/accuracy_reward": 0.4409722238779068,
|
| 1372 |
+
"rewards/format_reward": 0.0,
|
| 1373 |
+
"rewards/tag_count_reward": 0.5017361044883728,
|
| 1374 |
+
"step": 91
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"clip_ratio": 0.0,
|
| 1378 |
+
"completion_length": 833.0833282470703,
|
| 1379 |
+
"epoch": 1.7655502392344498,
|
| 1380 |
+
"grad_norm": 0.06350369442926171,
|
| 1381 |
+
"kl": 0.00672149658203125,
|
| 1382 |
+
"learning_rate": 1.3646898477089625e-07,
|
| 1383 |
+
"loss": 0.0197,
|
| 1384 |
+
"reward": 0.8750000149011612,
|
| 1385 |
+
"reward_std": 0.24306794628500938,
|
| 1386 |
+
"rewards/accuracy_reward": 0.4131944477558136,
|
| 1387 |
+
"rewards/format_reward": 0.0,
|
| 1388 |
+
"rewards/tag_count_reward": 0.461805559694767,
|
| 1389 |
+
"step": 92
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"clip_ratio": 0.0,
|
| 1393 |
+
"completion_length": 840.3715362548828,
|
| 1394 |
+
"epoch": 1.784688995215311,
|
| 1395 |
+
"grad_norm": 0.07310106942013637,
|
| 1396 |
+
"kl": 0.00698089599609375,
|
| 1397 |
+
"learning_rate": 1.307113441944992e-07,
|
| 1398 |
+
"loss": 0.0204,
|
| 1399 |
+
"reward": 0.8324652761220932,
|
| 1400 |
+
"reward_std": 0.26639265939593315,
|
| 1401 |
+
"rewards/accuracy_reward": 0.3750000074505806,
|
| 1402 |
+
"rewards/format_reward": 0.0,
|
| 1403 |
+
"rewards/tag_count_reward": 0.4574652835726738,
|
| 1404 |
+
"step": 93
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"clip_ratio": 0.0,
|
| 1408 |
+
"completion_length": 798.3472290039062,
|
| 1409 |
+
"epoch": 1.8038277511961722,
|
| 1410 |
+
"grad_norm": 0.07470777769552865,
|
| 1411 |
+
"kl": 0.0069580078125,
|
| 1412 |
+
"learning_rate": 1.2543211958050894e-07,
|
| 1413 |
+
"loss": 0.0225,
|
| 1414 |
+
"reward": 0.9123263955116272,
|
| 1415 |
+
"reward_std": 0.26884787529706955,
|
| 1416 |
+
"rewards/accuracy_reward": 0.4409722238779068,
|
| 1417 |
+
"rewards/format_reward": 0.0,
|
| 1418 |
+
"rewards/tag_count_reward": 0.4713541641831398,
|
| 1419 |
+
"step": 94
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"clip_ratio": 0.0,
|
| 1423 |
+
"completion_length": 809.6944580078125,
|
| 1424 |
+
"epoch": 1.8229665071770333,
|
| 1425 |
+
"grad_norm": 0.06571441072063877,
|
| 1426 |
+
"kl": 0.007171630859375,
|
| 1427 |
+
"learning_rate": 1.2063733461997803e-07,
|
| 1428 |
+
"loss": 0.0208,
|
| 1429 |
+
"reward": 0.9305555671453476,
|
| 1430 |
+
"reward_std": 0.21606039628386497,
|
| 1431 |
+
"rewards/accuracy_reward": 0.447916679084301,
|
| 1432 |
+
"rewards/format_reward": 0.0,
|
| 1433 |
+
"rewards/tag_count_reward": 0.4826388955116272,
|
| 1434 |
+
"step": 95
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"clip_ratio": 0.0,
|
| 1438 |
+
"completion_length": 814.4757080078125,
|
| 1439 |
+
"epoch": 1.8421052631578947,
|
| 1440 |
+
"grad_norm": 0.07474184052328098,
|
| 1441 |
+
"kl": 0.0072021484375,
|
| 1442 |
+
"learning_rate": 1.1633246024952662e-07,
|
| 1443 |
+
"loss": 0.0291,
|
| 1444 |
+
"reward": 0.8836805522441864,
|
| 1445 |
+
"reward_std": 0.2602545730769634,
|
| 1446 |
+
"rewards/accuracy_reward": 0.40625,
|
| 1447 |
+
"rewards/format_reward": 0.0,
|
| 1448 |
+
"rewards/tag_count_reward": 0.477430559694767,
|
| 1449 |
+
"step": 96
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"clip_ratio": 0.0,
|
| 1453 |
+
"completion_length": 817.5972290039062,
|
| 1454 |
+
"epoch": 1.861244019138756,
|
| 1455 |
+
"grad_norm": 0.08343014010656753,
|
| 1456 |
+
"kl": 0.00762939453125,
|
| 1457 |
+
"learning_rate": 1.1252240840890425e-07,
|
| 1458 |
+
"loss": 0.0248,
|
| 1459 |
+
"reward": 0.9192708283662796,
|
| 1460 |
+
"reward_std": 0.27130311727523804,
|
| 1461 |
+
"rewards/accuracy_reward": 0.4375000149011612,
|
| 1462 |
+
"rewards/format_reward": 0.0,
|
| 1463 |
+
"rewards/tag_count_reward": 0.4817708358168602,
|
| 1464 |
+
"step": 97
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"clip_ratio": 0.0,
|
| 1468 |
+
"completion_length": 812.0278015136719,
|
| 1469 |
+
"epoch": 1.8803827751196174,
|
| 1470 |
+
"grad_norm": 0.07184568208672217,
|
| 1471 |
+
"kl": 0.00701904296875,
|
| 1472 |
+
"learning_rate": 1.0921152643637749e-07,
|
| 1473 |
+
"loss": 0.0289,
|
| 1474 |
+
"reward": 0.90625,
|
| 1475 |
+
"reward_std": 0.2725307270884514,
|
| 1476 |
+
"rewards/accuracy_reward": 0.4305555522441864,
|
| 1477 |
+
"rewards/format_reward": 0.0,
|
| 1478 |
+
"rewards/tag_count_reward": 0.4756944477558136,
|
| 1479 |
+
"step": 98
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"clip_ratio": 0.0,
|
| 1483 |
+
"completion_length": 780.3958435058594,
|
| 1484 |
+
"epoch": 1.8995215311004785,
|
| 1485 |
+
"grad_norm": 0.06820176876725473,
|
| 1486 |
+
"kl": 0.0081787109375,
|
| 1487 |
+
"learning_rate": 1.064035921083383e-07,
|
| 1488 |
+
"loss": 0.0215,
|
| 1489 |
+
"reward": 0.9201388955116272,
|
| 1490 |
+
"reward_std": 0.22588132694363594,
|
| 1491 |
+
"rewards/accuracy_reward": 0.430555559694767,
|
| 1492 |
+
"rewards/format_reward": 0.0,
|
| 1493 |
+
"rewards/tag_count_reward": 0.4895833432674408,
|
| 1494 |
+
"step": 99
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"clip_ratio": 0.0,
|
| 1498 |
+
"completion_length": 812.1562652587891,
|
| 1499 |
+
"epoch": 1.9186602870813396,
|
| 1500 |
+
"grad_norm": 0.07452483871871766,
|
| 1501 |
+
"kl": 0.00782012939453125,
|
| 1502 |
+
"learning_rate": 1.0410180932879252e-07,
|
| 1503 |
+
"loss": 0.0291,
|
| 1504 |
+
"reward": 0.9079861044883728,
|
| 1505 |
+
"reward_std": 0.31181444972753525,
|
| 1506 |
+
"rewards/accuracy_reward": 0.4201388880610466,
|
| 1507 |
+
"rewards/format_reward": 0.0,
|
| 1508 |
+
"rewards/tag_count_reward": 0.4878472313284874,
|
| 1509 |
+
"step": 100
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"clip_ratio": 0.0,
|
| 1513 |
+
"completion_length": 827.5972442626953,
|
| 1514 |
+
"epoch": 1.937799043062201,
|
| 1515 |
+
"grad_norm": 0.06920787242163999,
|
| 1516 |
+
"kl": 0.00714874267578125,
|
| 1517 |
+
"learning_rate": 1.0230880447364718e-07,
|
| 1518 |
+
"loss": 0.0181,
|
| 1519 |
+
"reward": 0.8758680671453476,
|
| 1520 |
+
"reward_std": 0.2762135788798332,
|
| 1521 |
+
"rewards/accuracy_reward": 0.4027777835726738,
|
| 1522 |
+
"rewards/format_reward": 0.0,
|
| 1523 |
+
"rewards/tag_count_reward": 0.4730902835726738,
|
| 1524 |
+
"step": 101
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"clip_ratio": 0.0,
|
| 1528 |
+
"completion_length": 848.9201354980469,
|
| 1529 |
+
"epoch": 1.9569377990430623,
|
| 1530 |
+
"grad_norm": 0.06359574498753441,
|
| 1531 |
+
"kl": 0.00762939453125,
|
| 1532 |
+
"learning_rate": 1.0102662339396817e-07,
|
| 1533 |
+
"loss": 0.0237,
|
| 1534 |
+
"reward": 0.8289930522441864,
|
| 1535 |
+
"reward_std": 0.23938510566949844,
|
| 1536 |
+
"rewards/accuracy_reward": 0.3611111119389534,
|
| 1537 |
+
"rewards/format_reward": 0.0,
|
| 1538 |
+
"rewards/tag_count_reward": 0.4678819552063942,
|
| 1539 |
+
"step": 102
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"clip_ratio": 0.0,
|
| 1543 |
+
"completion_length": 827.0173645019531,
|
| 1544 |
+
"epoch": 1.9760765550239234,
|
| 1545 |
+
"grad_norm": 0.06784856725869416,
|
| 1546 |
+
"kl": 0.00757598876953125,
|
| 1547 |
+
"learning_rate": 1.002567290816269e-07,
|
| 1548 |
+
"loss": 0.0233,
|
| 1549 |
+
"reward": 0.9071180671453476,
|
| 1550 |
+
"reward_std": 0.2565717250108719,
|
| 1551 |
+
"rewards/accuracy_reward": 0.4340277761220932,
|
| 1552 |
+
"rewards/format_reward": 0.0,
|
| 1553 |
+
"rewards/tag_count_reward": 0.4730902835726738,
|
| 1554 |
+
"step": 103
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"clip_ratio": 0.0,
|
| 1558 |
+
"completion_length": 824.1666870117188,
|
| 1559 |
+
"epoch": 1.9952153110047846,
|
| 1560 |
+
"grad_norm": 0.07003424025971802,
|
| 1561 |
+
"kl": 0.00720977783203125,
|
| 1562 |
+
"learning_rate": 1e-07,
|
| 1563 |
+
"loss": 0.0232,
|
| 1564 |
+
"reward": 0.8446180522441864,
|
| 1565 |
+
"reward_std": 0.24920602142810822,
|
| 1566 |
+
"rewards/accuracy_reward": 0.3784722313284874,
|
| 1567 |
+
"rewards/format_reward": 0.0,
|
| 1568 |
+
"rewards/tag_count_reward": 0.4661458358168602,
|
| 1569 |
+
"step": 104
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 1.9952153110047846,
|
| 1573 |
+
"step": 104,
|
| 1574 |
+
"total_flos": 0.0,
|
| 1575 |
+
"train_loss": 0.02088056951134394,
|
| 1576 |
+
"train_runtime": 4550.0212,
|
| 1577 |
+
"train_samples_per_second": 3.297,
|
| 1578 |
+
"train_steps_per_second": 0.023
|
| 1579 |
+
}
|
| 1580 |
+
],
|
| 1581 |
+
"logging_steps": 1,
|
| 1582 |
+
"max_steps": 104,
|
| 1583 |
+
"num_input_tokens_seen": 0,
|
| 1584 |
+
"num_train_epochs": 2,
|
| 1585 |
+
"save_steps": 10,
|
| 1586 |
+
"stateful_callbacks": {
|
| 1587 |
+
"TrainerControl": {
|
| 1588 |
+
"args": {
|
| 1589 |
+
"should_epoch_stop": false,
|
| 1590 |
+
"should_evaluate": false,
|
| 1591 |
+
"should_log": false,
|
| 1592 |
+
"should_save": true,
|
| 1593 |
+
"should_training_stop": true
|
| 1594 |
+
},
|
| 1595 |
+
"attributes": {}
|
| 1596 |
+
}
|
| 1597 |
+
},
|
| 1598 |
+
"total_flos": 0.0,
|
| 1599 |
+
"train_batch_size": 24,
|
| 1600 |
+
"trial_name": null,
|
| 1601 |
+
"trial_params": null
|
| 1602 |
+
}
|