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 +1943 -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: GRPO_0.2_0.2
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- grpo
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for GRPO_0.2_0.2
|
| 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="LLucass/GRPO_0.2_0.2", 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/lavatorywang-nus/Gaussian/runs/bxs2lsz9)
|
| 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: 4.51.3
|
| 39 |
+
- Pytorch: 2.5.1
|
| 40 |
+
- Datasets: 3.6.0
|
| 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": 4.974420491237197e-05,
|
| 4 |
+
"train_runtime": 8330.7391,
|
| 5 |
+
"train_samples": 7000,
|
| 6 |
+
"train_samples_per_second": 1.152,
|
| 7 |
+
"train_steps_per_second": 0.012
|
| 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": "4.51.3"
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 4.974420491237197e-05,
|
| 4 |
+
"train_runtime": 8330.7391,
|
| 5 |
+
"train_samples": 7000,
|
| 6 |
+
"train_samples_per_second": 1.152,
|
| 7 |
+
"train_steps_per_second": 0.012
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1943 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.11428571428571428,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 100,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"clip_ratio/high_max": 0.0,
|
| 14 |
+
"clip_ratio/high_mean": 0.0,
|
| 15 |
+
"clip_ratio/low_mean": 0.0,
|
| 16 |
+
"clip_ratio/low_min": 0.0,
|
| 17 |
+
"clip_ratio/region_mean": 0.0,
|
| 18 |
+
"completion_length": 2700.4271850585938,
|
| 19 |
+
"entropy": 0.36767578125,
|
| 20 |
+
"epoch": 0.001142857142857143,
|
| 21 |
+
"grad_norm": 0.11875967681407928,
|
| 22 |
+
"kl": 0.0,
|
| 23 |
+
"learning_rate": 0.0,
|
| 24 |
+
"loss": 0.0,
|
| 25 |
+
"reward": 0.7604166893288493,
|
| 26 |
+
"reward_std": 0.4268697127699852,
|
| 27 |
+
"rewards/accuracy_reward": 0.25000001303851604,
|
| 28 |
+
"rewards/format_reward": 0.5104166669771075,
|
| 29 |
+
"step": 1
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"clip_ratio/high_max": 0.0,
|
| 33 |
+
"clip_ratio/high_mean": 0.0,
|
| 34 |
+
"clip_ratio/low_mean": 0.0,
|
| 35 |
+
"clip_ratio/low_min": 0.0,
|
| 36 |
+
"clip_ratio/region_mean": 0.0,
|
| 37 |
+
"completion_length": 3127.3958435058594,
|
| 38 |
+
"entropy": 0.353515625,
|
| 39 |
+
"epoch": 0.002285714285714286,
|
| 40 |
+
"grad_norm": 0.14844515919685364,
|
| 41 |
+
"kl": 0.0,
|
| 42 |
+
"learning_rate": 1e-07,
|
| 43 |
+
"loss": 0.0,
|
| 44 |
+
"reward": 0.6458333637565374,
|
| 45 |
+
"reward_std": 0.4249730706214905,
|
| 46 |
+
"rewards/accuracy_reward": 0.2812500102445483,
|
| 47 |
+
"rewards/format_reward": 0.3645833386108279,
|
| 48 |
+
"step": 2
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"clip_ratio/high_max": 0.0,
|
| 52 |
+
"clip_ratio/high_mean": 0.0,
|
| 53 |
+
"clip_ratio/low_mean": 0.0,
|
| 54 |
+
"clip_ratio/low_min": 0.0,
|
| 55 |
+
"clip_ratio/region_mean": 0.0,
|
| 56 |
+
"completion_length": 3672.3229370117188,
|
| 57 |
+
"entropy": 0.43701171875,
|
| 58 |
+
"epoch": 0.0034285714285714284,
|
| 59 |
+
"grad_norm": 0.15467818081378937,
|
| 60 |
+
"kl": 4.011392593383789e-05,
|
| 61 |
+
"learning_rate": 2e-07,
|
| 62 |
+
"loss": 0.0,
|
| 63 |
+
"reward": 0.20833333674818277,
|
| 64 |
+
"reward_std": 0.3576926067471504,
|
| 65 |
+
"rewards/accuracy_reward": 0.05208333395421505,
|
| 66 |
+
"rewards/format_reward": 0.15625000838190317,
|
| 67 |
+
"step": 3
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"clip_ratio/high_max": 0.0,
|
| 71 |
+
"clip_ratio/high_mean": 0.0,
|
| 72 |
+
"clip_ratio/low_mean": 0.0,
|
| 73 |
+
"clip_ratio/low_min": 0.0,
|
| 74 |
+
"clip_ratio/region_mean": 0.0,
|
| 75 |
+
"completion_length": 2434.0000610351562,
|
| 76 |
+
"entropy": 0.40673828125,
|
| 77 |
+
"epoch": 0.004571428571428572,
|
| 78 |
+
"grad_norm": 0.1262386441230774,
|
| 79 |
+
"kl": 3.9577484130859375e-05,
|
| 80 |
+
"learning_rate": 3e-07,
|
| 81 |
+
"loss": 0.0,
|
| 82 |
+
"reward": 0.6666666716337204,
|
| 83 |
+
"reward_std": 0.3760714754462242,
|
| 84 |
+
"rewards/accuracy_reward": 0.07291666697710752,
|
| 85 |
+
"rewards/format_reward": 0.5937500074505806,
|
| 86 |
+
"step": 4
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"clip_ratio/high_max": 0.0,
|
| 90 |
+
"clip_ratio/high_mean": 0.0,
|
| 91 |
+
"clip_ratio/low_mean": 0.0,
|
| 92 |
+
"clip_ratio/low_min": 0.0,
|
| 93 |
+
"clip_ratio/region_mean": 0.0,
|
| 94 |
+
"completion_length": 3514.447998046875,
|
| 95 |
+
"entropy": 0.46142578125,
|
| 96 |
+
"epoch": 0.005714285714285714,
|
| 97 |
+
"grad_norm": 0.11896030604839325,
|
| 98 |
+
"kl": 3.7997961044311523e-05,
|
| 99 |
+
"learning_rate": 4e-07,
|
| 100 |
+
"loss": 0.0,
|
| 101 |
+
"reward": 0.4375000149011612,
|
| 102 |
+
"reward_std": 0.45638682320713997,
|
| 103 |
+
"rewards/accuracy_reward": 0.0729166716337204,
|
| 104 |
+
"rewards/format_reward": 0.3645833432674408,
|
| 105 |
+
"step": 5
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"clip_ratio/high_max": 0.0,
|
| 109 |
+
"clip_ratio/high_mean": 0.0,
|
| 110 |
+
"clip_ratio/low_mean": 0.0,
|
| 111 |
+
"clip_ratio/low_min": 0.0,
|
| 112 |
+
"clip_ratio/region_mean": 0.0,
|
| 113 |
+
"completion_length": 3369.8125610351562,
|
| 114 |
+
"entropy": 0.46142578125,
|
| 115 |
+
"epoch": 0.006857142857142857,
|
| 116 |
+
"grad_norm": 0.11138290911912918,
|
| 117 |
+
"kl": 5.46574592590332e-05,
|
| 118 |
+
"learning_rate": 5e-07,
|
| 119 |
+
"loss": 0.0,
|
| 120 |
+
"reward": 0.3333333395421505,
|
| 121 |
+
"reward_std": 0.3872676342725754,
|
| 122 |
+
"rewards/accuracy_reward": 0.0625,
|
| 123 |
+
"rewards/format_reward": 0.2708333395421505,
|
| 124 |
+
"step": 6
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"clip_ratio/high_max": 0.0,
|
| 128 |
+
"clip_ratio/high_mean": 0.0,
|
| 129 |
+
"clip_ratio/low_mean": 0.0,
|
| 130 |
+
"clip_ratio/low_min": 0.0,
|
| 131 |
+
"clip_ratio/region_mean": 0.0,
|
| 132 |
+
"completion_length": 3402.6146240234375,
|
| 133 |
+
"entropy": 0.400390625,
|
| 134 |
+
"epoch": 0.008,
|
| 135 |
+
"grad_norm": 0.10288004577159882,
|
| 136 |
+
"kl": 2.6792287826538086e-05,
|
| 137 |
+
"learning_rate": 6e-07,
|
| 138 |
+
"loss": 0.0,
|
| 139 |
+
"reward": 0.833333358168602,
|
| 140 |
+
"reward_std": 0.5691196992993355,
|
| 141 |
+
"rewards/accuracy_reward": 0.2500000074505806,
|
| 142 |
+
"rewards/format_reward": 0.5833333358168602,
|
| 143 |
+
"step": 7
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"clip_ratio/high_max": 0.0,
|
| 147 |
+
"clip_ratio/high_mean": 0.0,
|
| 148 |
+
"clip_ratio/low_mean": 0.0,
|
| 149 |
+
"clip_ratio/low_min": 0.0,
|
| 150 |
+
"clip_ratio/region_mean": 0.0,
|
| 151 |
+
"completion_length": 2891.447998046875,
|
| 152 |
+
"entropy": 0.357421875,
|
| 153 |
+
"epoch": 0.009142857142857144,
|
| 154 |
+
"grad_norm": 0.0705084502696991,
|
| 155 |
+
"kl": 2.35140323638916e-05,
|
| 156 |
+
"learning_rate": 7e-07,
|
| 157 |
+
"loss": 0.0,
|
| 158 |
+
"reward": 0.8541666865348816,
|
| 159 |
+
"reward_std": 0.22278132662177086,
|
| 160 |
+
"rewards/accuracy_reward": 0.3750000102445483,
|
| 161 |
+
"rewards/format_reward": 0.479166679084301,
|
| 162 |
+
"step": 8
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"clip_ratio/high_max": 0.0,
|
| 166 |
+
"clip_ratio/high_mean": 0.0,
|
| 167 |
+
"clip_ratio/low_mean": 0.0,
|
| 168 |
+
"clip_ratio/low_min": 0.0,
|
| 169 |
+
"clip_ratio/region_mean": 0.0,
|
| 170 |
+
"completion_length": 3418.947998046875,
|
| 171 |
+
"entropy": 0.45068359375,
|
| 172 |
+
"epoch": 0.010285714285714285,
|
| 173 |
+
"grad_norm": 0.15328024327754974,
|
| 174 |
+
"kl": 4.3720006942749023e-05,
|
| 175 |
+
"learning_rate": 8e-07,
|
| 176 |
+
"loss": 0.0,
|
| 177 |
+
"reward": 0.5000000223517418,
|
| 178 |
+
"reward_std": 0.6385103613138199,
|
| 179 |
+
"rewards/accuracy_reward": 0.15625000279396772,
|
| 180 |
+
"rewards/format_reward": 0.3437500149011612,
|
| 181 |
+
"step": 9
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"clip_ratio/high_max": 0.0,
|
| 185 |
+
"clip_ratio/high_mean": 0.0,
|
| 186 |
+
"clip_ratio/low_mean": 0.0,
|
| 187 |
+
"clip_ratio/low_min": 0.0,
|
| 188 |
+
"clip_ratio/region_mean": 0.0,
|
| 189 |
+
"completion_length": 2938.40625,
|
| 190 |
+
"entropy": 0.3408203125,
|
| 191 |
+
"epoch": 0.011428571428571429,
|
| 192 |
+
"grad_norm": 0.13008344173431396,
|
| 193 |
+
"kl": 2.7120113372802734e-05,
|
| 194 |
+
"learning_rate": 9e-07,
|
| 195 |
+
"loss": 0.0,
|
| 196 |
+
"reward": 0.5625000149011612,
|
| 197 |
+
"reward_std": 0.4825209602713585,
|
| 198 |
+
"rewards/accuracy_reward": 0.15625000558793545,
|
| 199 |
+
"rewards/format_reward": 0.4062500111758709,
|
| 200 |
+
"step": 10
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"clip_ratio/high_max": 0.0,
|
| 204 |
+
"clip_ratio/high_mean": 0.0,
|
| 205 |
+
"clip_ratio/low_mean": 0.0,
|
| 206 |
+
"clip_ratio/low_min": 0.0,
|
| 207 |
+
"clip_ratio/region_mean": 0.0,
|
| 208 |
+
"completion_length": 3608.0625610351562,
|
| 209 |
+
"entropy": 0.37744140625,
|
| 210 |
+
"epoch": 0.012571428571428572,
|
| 211 |
+
"grad_norm": 0.13492900133132935,
|
| 212 |
+
"kl": 2.568960189819336e-05,
|
| 213 |
+
"learning_rate": 1e-06,
|
| 214 |
+
"loss": 0.0,
|
| 215 |
+
"reward": 0.20833334140479565,
|
| 216 |
+
"reward_std": 0.3463892936706543,
|
| 217 |
+
"rewards/accuracy_reward": 0.06250000093132257,
|
| 218 |
+
"rewards/format_reward": 0.14583333488553762,
|
| 219 |
+
"step": 11
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"clip_ratio/high_max": 0.0,
|
| 223 |
+
"clip_ratio/high_mean": 0.0,
|
| 224 |
+
"clip_ratio/low_mean": 0.0,
|
| 225 |
+
"clip_ratio/low_min": 0.0,
|
| 226 |
+
"clip_ratio/region_mean": 0.0,
|
| 227 |
+
"completion_length": 2577.541748046875,
|
| 228 |
+
"entropy": 0.40869140625,
|
| 229 |
+
"epoch": 0.013714285714285714,
|
| 230 |
+
"grad_norm": 0.17103247344493866,
|
| 231 |
+
"kl": 3.412365913391113e-05,
|
| 232 |
+
"learning_rate": 9.997258721585931e-07,
|
| 233 |
+
"loss": 0.0,
|
| 234 |
+
"reward": 0.7812500298023224,
|
| 235 |
+
"reward_std": 0.4395437017083168,
|
| 236 |
+
"rewards/accuracy_reward": 0.17708333861082792,
|
| 237 |
+
"rewards/format_reward": 0.6041666865348816,
|
| 238 |
+
"step": 12
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"clip_ratio/high_max": 0.0,
|
| 242 |
+
"clip_ratio/high_mean": 0.0,
|
| 243 |
+
"clip_ratio/low_mean": 0.0,
|
| 244 |
+
"clip_ratio/low_min": 0.0,
|
| 245 |
+
"clip_ratio/region_mean": 0.0,
|
| 246 |
+
"completion_length": 3150.9168090820312,
|
| 247 |
+
"entropy": 0.39013671875,
|
| 248 |
+
"epoch": 0.014857142857142857,
|
| 249 |
+
"grad_norm": 0.15444810688495636,
|
| 250 |
+
"kl": 2.7447938919067383e-05,
|
| 251 |
+
"learning_rate": 9.989038226169207e-07,
|
| 252 |
+
"loss": 0.0,
|
| 253 |
+
"reward": 0.6145833507180214,
|
| 254 |
+
"reward_std": 0.32639050483703613,
|
| 255 |
+
"rewards/accuracy_reward": 0.18750000279396772,
|
| 256 |
+
"rewards/format_reward": 0.4270833432674408,
|
| 257 |
+
"step": 13
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"clip_ratio/high_max": 0.0,
|
| 261 |
+
"clip_ratio/high_mean": 0.0,
|
| 262 |
+
"clip_ratio/low_mean": 0.0,
|
| 263 |
+
"clip_ratio/low_min": 0.0,
|
| 264 |
+
"clip_ratio/region_mean": 0.0,
|
| 265 |
+
"completion_length": 3099.6771240234375,
|
| 266 |
+
"entropy": 0.3623046875,
|
| 267 |
+
"epoch": 0.016,
|
| 268 |
+
"grad_norm": 0.22930943965911865,
|
| 269 |
+
"kl": 2.720952033996582e-05,
|
| 270 |
+
"learning_rate": 9.975348529157229e-07,
|
| 271 |
+
"loss": 0.0,
|
| 272 |
+
"reward": 0.6041666716337204,
|
| 273 |
+
"reward_std": 0.530846893787384,
|
| 274 |
+
"rewards/accuracy_reward": 0.18750000651925802,
|
| 275 |
+
"rewards/format_reward": 0.416666679084301,
|
| 276 |
+
"step": 14
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"clip_ratio/high_max": 0.0,
|
| 280 |
+
"clip_ratio/high_mean": 0.0,
|
| 281 |
+
"clip_ratio/low_mean": 0.0,
|
| 282 |
+
"clip_ratio/low_min": 0.0,
|
| 283 |
+
"clip_ratio/region_mean": 0.0,
|
| 284 |
+
"completion_length": 2916.572998046875,
|
| 285 |
+
"entropy": 0.3740234375,
|
| 286 |
+
"epoch": 0.017142857142857144,
|
| 287 |
+
"grad_norm": 0.09911036491394043,
|
| 288 |
+
"kl": 2.2709369659423828e-05,
|
| 289 |
+
"learning_rate": 9.956206309337066e-07,
|
| 290 |
+
"loss": 0.0,
|
| 291 |
+
"reward": 0.7187500149011612,
|
| 292 |
+
"reward_std": 0.3208988383412361,
|
| 293 |
+
"rewards/accuracy_reward": 0.2604166679084301,
|
| 294 |
+
"rewards/format_reward": 0.4583333507180214,
|
| 295 |
+
"step": 15
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"clip_ratio/high_max": 0.0,
|
| 299 |
+
"clip_ratio/high_mean": 0.0,
|
| 300 |
+
"clip_ratio/low_mean": 0.0,
|
| 301 |
+
"clip_ratio/low_min": 0.0,
|
| 302 |
+
"clip_ratio/region_mean": 0.0,
|
| 303 |
+
"completion_length": 3797.3541870117188,
|
| 304 |
+
"entropy": 0.45654296875,
|
| 305 |
+
"epoch": 0.018285714285714287,
|
| 306 |
+
"grad_norm": 0.12035630643367767,
|
| 307 |
+
"kl": 2.969801425933838e-05,
|
| 308 |
+
"learning_rate": 9.931634888554935e-07,
|
| 309 |
+
"loss": 0.0,
|
| 310 |
+
"reward": 0.18750000651925802,
|
| 311 |
+
"reward_std": 0.3208543583750725,
|
| 312 |
+
"rewards/accuracy_reward": 0.06250000186264515,
|
| 313 |
+
"rewards/format_reward": 0.12500000465661287,
|
| 314 |
+
"step": 16
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"clip_ratio/high_max": 0.0,
|
| 318 |
+
"clip_ratio/high_mean": 0.0,
|
| 319 |
+
"clip_ratio/low_mean": 0.0,
|
| 320 |
+
"clip_ratio/low_min": 0.0,
|
| 321 |
+
"clip_ratio/region_mean": 0.0,
|
| 322 |
+
"completion_length": 2549.604217529297,
|
| 323 |
+
"entropy": 0.44873046875,
|
| 324 |
+
"epoch": 0.019428571428571427,
|
| 325 |
+
"grad_norm": 0.17906327545642853,
|
| 326 |
+
"kl": 3.573298454284668e-05,
|
| 327 |
+
"learning_rate": 9.901664203302124e-07,
|
| 328 |
+
"loss": 0.0,
|
| 329 |
+
"reward": 0.8437500223517418,
|
| 330 |
+
"reward_std": 0.5051586776971817,
|
| 331 |
+
"rewards/accuracy_reward": 0.27083334885537624,
|
| 332 |
+
"rewards/format_reward": 0.572916679084301,
|
| 333 |
+
"step": 17
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"clip_ratio/high_max": 0.0,
|
| 337 |
+
"clip_ratio/high_mean": 0.0,
|
| 338 |
+
"clip_ratio/low_mean": 0.0,
|
| 339 |
+
"clip_ratio/low_min": 0.0,
|
| 340 |
+
"clip_ratio/region_mean": 0.0,
|
| 341 |
+
"completion_length": 3061.9896240234375,
|
| 342 |
+
"entropy": 0.35693359375,
|
| 343 |
+
"epoch": 0.02057142857142857,
|
| 344 |
+
"grad_norm": 0.10486536473035812,
|
| 345 |
+
"kl": 2.0988285541534424e-05,
|
| 346 |
+
"learning_rate": 9.866330768241983e-07,
|
| 347 |
+
"loss": 0.0,
|
| 348 |
+
"reward": 0.6770833656191826,
|
| 349 |
+
"reward_std": 0.45130809396505356,
|
| 350 |
+
"rewards/accuracy_reward": 0.19791667349636555,
|
| 351 |
+
"rewards/format_reward": 0.4791666828095913,
|
| 352 |
+
"step": 18
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"clip_ratio/high_max": 0.0,
|
| 356 |
+
"clip_ratio/high_mean": 0.0,
|
| 357 |
+
"clip_ratio/low_mean": 0.0,
|
| 358 |
+
"clip_ratio/low_min": 0.0,
|
| 359 |
+
"clip_ratio/region_mean": 0.0,
|
| 360 |
+
"completion_length": 3092.8959350585938,
|
| 361 |
+
"entropy": 0.39306640625,
|
| 362 |
+
"epoch": 0.021714285714285714,
|
| 363 |
+
"grad_norm": 0.1308853030204773,
|
| 364 |
+
"kl": 4.026293754577637e-05,
|
| 365 |
+
"learning_rate": 9.825677631722435e-07,
|
| 366 |
+
"loss": 0.0,
|
| 367 |
+
"reward": 0.864583358168602,
|
| 368 |
+
"reward_std": 0.630811333656311,
|
| 369 |
+
"rewards/accuracy_reward": 0.39583334513008595,
|
| 370 |
+
"rewards/format_reward": 0.4687500149011612,
|
| 371 |
+
"step": 19
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"clip_ratio/high_max": 0.0,
|
| 375 |
+
"clip_ratio/high_mean": 0.0,
|
| 376 |
+
"clip_ratio/low_mean": 0.0,
|
| 377 |
+
"clip_ratio/low_min": 0.0,
|
| 378 |
+
"clip_ratio/region_mean": 0.0,
|
| 379 |
+
"completion_length": 2584.5833740234375,
|
| 380 |
+
"entropy": 0.3388671875,
|
| 381 |
+
"epoch": 0.022857142857142857,
|
| 382 |
+
"grad_norm": 0.16664722561836243,
|
| 383 |
+
"kl": 4.4479966163635254e-05,
|
| 384 |
+
"learning_rate": 9.779754323328192e-07,
|
| 385 |
+
"loss": 0.0,
|
| 386 |
+
"reward": 0.958333358168602,
|
| 387 |
+
"reward_std": 0.539019413292408,
|
| 388 |
+
"rewards/accuracy_reward": 0.2812500037252903,
|
| 389 |
+
"rewards/format_reward": 0.677083358168602,
|
| 390 |
+
"step": 20
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"clip_ratio/high_max": 0.0,
|
| 394 |
+
"clip_ratio/high_mean": 0.0,
|
| 395 |
+
"clip_ratio/low_mean": 0.0,
|
| 396 |
+
"clip_ratio/low_min": 0.0,
|
| 397 |
+
"clip_ratio/region_mean": 0.0,
|
| 398 |
+
"completion_length": 2894.4479370117188,
|
| 399 |
+
"entropy": 0.4404296875,
|
| 400 |
+
"epoch": 0.024,
|
| 401 |
+
"grad_norm": 0.11469059437513351,
|
| 402 |
+
"kl": 0.00022789835929870605,
|
| 403 |
+
"learning_rate": 9.728616793536587e-07,
|
| 404 |
+
"loss": 0.0,
|
| 405 |
+
"reward": 0.5937500055879354,
|
| 406 |
+
"reward_std": 0.3156265318393707,
|
| 407 |
+
"rewards/accuracy_reward": 0.15625000279396772,
|
| 408 |
+
"rewards/format_reward": 0.4375000027939677,
|
| 409 |
+
"step": 21
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"clip_ratio/high_max": 0.0,
|
| 413 |
+
"clip_ratio/high_mean": 0.0,
|
| 414 |
+
"clip_ratio/low_mean": 0.0,
|
| 415 |
+
"clip_ratio/low_min": 0.0,
|
| 416 |
+
"clip_ratio/region_mean": 0.0,
|
| 417 |
+
"completion_length": 1811.4271697998047,
|
| 418 |
+
"entropy": 0.39404296875,
|
| 419 |
+
"epoch": 0.025142857142857144,
|
| 420 |
+
"grad_norm": 0.1692061573266983,
|
| 421 |
+
"kl": 0.00019675493240356445,
|
| 422 |
+
"learning_rate": 9.672327345550543e-07,
|
| 423 |
+
"loss": 0.0,
|
| 424 |
+
"reward": 1.1250000298023224,
|
| 425 |
+
"reward_std": 0.41911277174949646,
|
| 426 |
+
"rewards/accuracy_reward": 0.30208334419876337,
|
| 427 |
+
"rewards/format_reward": 0.8229166865348816,
|
| 428 |
+
"step": 22
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"clip_ratio/high_max": 0.0,
|
| 432 |
+
"clip_ratio/high_mean": 0.0,
|
| 433 |
+
"clip_ratio/low_mean": 0.0,
|
| 434 |
+
"clip_ratio/low_min": 0.0,
|
| 435 |
+
"clip_ratio/region_mean": 0.0,
|
| 436 |
+
"completion_length": 2823.072998046875,
|
| 437 |
+
"entropy": 0.38671875,
|
| 438 |
+
"epoch": 0.026285714285714287,
|
| 439 |
+
"grad_norm": 0.1204013079404831,
|
| 440 |
+
"kl": 7.50422477722168e-05,
|
| 441 |
+
"learning_rate": 9.610954559391704e-07,
|
| 442 |
+
"loss": 0.0,
|
| 443 |
+
"reward": 0.6562500223517418,
|
| 444 |
+
"reward_std": 0.506660170853138,
|
| 445 |
+
"rewards/accuracy_reward": 0.16666667349636555,
|
| 446 |
+
"rewards/format_reward": 0.4895833432674408,
|
| 447 |
+
"step": 23
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"clip_ratio/high_max": 0.0,
|
| 451 |
+
"clip_ratio/high_mean": 0.0,
|
| 452 |
+
"clip_ratio/low_mean": 0.0,
|
| 453 |
+
"clip_ratio/low_min": 0.0,
|
| 454 |
+
"clip_ratio/region_mean": 0.0,
|
| 455 |
+
"completion_length": 2825.416748046875,
|
| 456 |
+
"entropy": 0.38525390625,
|
| 457 |
+
"epoch": 0.027428571428571427,
|
| 458 |
+
"grad_norm": 0.147517591714859,
|
| 459 |
+
"kl": 0.00011518597602844238,
|
| 460 |
+
"learning_rate": 9.54457320834625e-07,
|
| 461 |
+
"loss": 0.0,
|
| 462 |
+
"reward": 0.8229166865348816,
|
| 463 |
+
"reward_std": 0.6497645676136017,
|
| 464 |
+
"rewards/accuracy_reward": 0.2708333358168602,
|
| 465 |
+
"rewards/format_reward": 0.5520833507180214,
|
| 466 |
+
"step": 24
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"clip_ratio/high_max": 0.0,
|
| 470 |
+
"clip_ratio/high_mean": 0.0,
|
| 471 |
+
"clip_ratio/low_mean": 0.0,
|
| 472 |
+
"clip_ratio/low_min": 0.0,
|
| 473 |
+
"clip_ratio/region_mean": 0.0,
|
| 474 |
+
"completion_length": 2880.7813110351562,
|
| 475 |
+
"entropy": 0.43115234375,
|
| 476 |
+
"epoch": 0.02857142857142857,
|
| 477 |
+
"grad_norm": 0.14937955141067505,
|
| 478 |
+
"kl": 0.0002092123031616211,
|
| 479 |
+
"learning_rate": 9.473264167865171e-07,
|
| 480 |
+
"loss": 0.0,
|
| 481 |
+
"reward": 0.7187500102445483,
|
| 482 |
+
"reward_std": 0.3600482754409313,
|
| 483 |
+
"rewards/accuracy_reward": 0.26041667722165585,
|
| 484 |
+
"rewards/format_reward": 0.4583333386108279,
|
| 485 |
+
"step": 25
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"clip_ratio/high_max": 0.0,
|
| 489 |
+
"clip_ratio/high_mean": 0.0,
|
| 490 |
+
"clip_ratio/low_mean": 0.0,
|
| 491 |
+
"clip_ratio/low_min": 0.0,
|
| 492 |
+
"clip_ratio/region_mean": 0.0,
|
| 493 |
+
"completion_length": 3217.5209350585938,
|
| 494 |
+
"entropy": 0.41455078125,
|
| 495 |
+
"epoch": 0.029714285714285714,
|
| 496 |
+
"grad_norm": 0.10219196230173111,
|
| 497 |
+
"kl": 2.9712915420532227e-05,
|
| 498 |
+
"learning_rate": 9.397114317029974e-07,
|
| 499 |
+
"loss": 0.0,
|
| 500 |
+
"reward": 0.6458333507180214,
|
| 501 |
+
"reward_std": 0.3189500719308853,
|
| 502 |
+
"rewards/accuracy_reward": 0.1770833358168602,
|
| 503 |
+
"rewards/format_reward": 0.4687500074505806,
|
| 504 |
+
"step": 26
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"clip_ratio/high_max": 0.0,
|
| 508 |
+
"clip_ratio/high_mean": 0.0,
|
| 509 |
+
"clip_ratio/low_mean": 0.0,
|
| 510 |
+
"clip_ratio/low_min": 0.0,
|
| 511 |
+
"clip_ratio/region_mean": 0.0,
|
| 512 |
+
"completion_length": 3267.0521850585938,
|
| 513 |
+
"entropy": 0.4404296875,
|
| 514 |
+
"epoch": 0.030857142857142857,
|
| 515 |
+
"grad_norm": 0.12435087561607361,
|
| 516 |
+
"kl": 0.0001267567276954651,
|
| 517 |
+
"learning_rate": 9.316216432703916e-07,
|
| 518 |
+
"loss": 0.0,
|
| 519 |
+
"reward": 0.5000000149011612,
|
| 520 |
+
"reward_std": 0.47042689844965935,
|
| 521 |
+
"rewards/accuracy_reward": 0.11458333861082792,
|
| 522 |
+
"rewards/format_reward": 0.3854166716337204,
|
| 523 |
+
"step": 27
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"clip_ratio/high_max": 0.0,
|
| 527 |
+
"clip_ratio/high_mean": 0.0,
|
| 528 |
+
"clip_ratio/low_mean": 0.0,
|
| 529 |
+
"clip_ratio/low_min": 0.0,
|
| 530 |
+
"clip_ratio/region_mean": 0.0,
|
| 531 |
+
"completion_length": 3018.5938110351562,
|
| 532 |
+
"entropy": 0.40185546875,
|
| 533 |
+
"epoch": 0.032,
|
| 534 |
+
"grad_norm": 0.13387200236320496,
|
| 535 |
+
"kl": 0.00012268126010894775,
|
| 536 |
+
"learning_rate": 9.230669076497687e-07,
|
| 537 |
+
"loss": 0.0,
|
| 538 |
+
"reward": 0.7395833432674408,
|
| 539 |
+
"reward_std": 0.40498094633221626,
|
| 540 |
+
"rewards/accuracy_reward": 0.30208334140479565,
|
| 541 |
+
"rewards/format_reward": 0.4375000074505806,
|
| 542 |
+
"step": 28
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"clip_ratio/high_max": 0.0,
|
| 546 |
+
"clip_ratio/high_mean": 0.0,
|
| 547 |
+
"clip_ratio/low_mean": 0.0,
|
| 548 |
+
"clip_ratio/low_min": 0.0,
|
| 549 |
+
"clip_ratio/region_mean": 0.0,
|
| 550 |
+
"completion_length": 3436.9688110351562,
|
| 551 |
+
"entropy": 0.4326171875,
|
| 552 |
+
"epoch": 0.03314285714285714,
|
| 553 |
+
"grad_norm": 0.18476502597332,
|
| 554 |
+
"kl": 0.0002486705780029297,
|
| 555 |
+
"learning_rate": 9.140576474687263e-07,
|
| 556 |
+
"loss": 0.0,
|
| 557 |
+
"reward": 0.3333333460614085,
|
| 558 |
+
"reward_std": 0.3739718273282051,
|
| 559 |
+
"rewards/accuracy_reward": 0.07291666697710752,
|
| 560 |
+
"rewards/format_reward": 0.2604166818782687,
|
| 561 |
+
"step": 29
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"clip_ratio/high_max": 0.0,
|
| 565 |
+
"clip_ratio/high_mean": 0.0,
|
| 566 |
+
"clip_ratio/low_mean": 0.0,
|
| 567 |
+
"clip_ratio/low_min": 0.0,
|
| 568 |
+
"clip_ratio/region_mean": 0.0,
|
| 569 |
+
"completion_length": 3012.510498046875,
|
| 570 |
+
"entropy": 0.38818359375,
|
| 571 |
+
"epoch": 0.03428571428571429,
|
| 572 |
+
"grad_norm": 0.10571455210447311,
|
| 573 |
+
"kl": 0.0007374286651611328,
|
| 574 |
+
"learning_rate": 9.046048391230247e-07,
|
| 575 |
+
"loss": 0.0,
|
| 576 |
+
"reward": 0.7291666716337204,
|
| 577 |
+
"reward_std": 0.5458482205867767,
|
| 578 |
+
"rewards/accuracy_reward": 0.22916667256504297,
|
| 579 |
+
"rewards/format_reward": 0.500000013038516,
|
| 580 |
+
"step": 30
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"clip_ratio/high_max": 0.0,
|
| 584 |
+
"clip_ratio/high_mean": 0.0,
|
| 585 |
+
"clip_ratio/low_mean": 0.0,
|
| 586 |
+
"clip_ratio/low_min": 0.0,
|
| 587 |
+
"clip_ratio/region_mean": 0.0,
|
| 588 |
+
"completion_length": 3214.7083740234375,
|
| 589 |
+
"entropy": 0.38037109375,
|
| 590 |
+
"epoch": 0.03542857142857143,
|
| 591 |
+
"grad_norm": 0.1431698352098465,
|
| 592 |
+
"kl": 0.0004398822784423828,
|
| 593 |
+
"learning_rate": 8.9471999940354e-07,
|
| 594 |
+
"loss": 0.0,
|
| 595 |
+
"reward": 0.5312500298023224,
|
| 596 |
+
"reward_std": 0.5084670260548592,
|
| 597 |
+
"rewards/accuracy_reward": 0.17708334140479565,
|
| 598 |
+
"rewards/format_reward": 0.3541666716337204,
|
| 599 |
+
"step": 31
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"clip_ratio/high_max": 0.0,
|
| 603 |
+
"clip_ratio/high_mean": 0.0,
|
| 604 |
+
"clip_ratio/low_mean": 0.0,
|
| 605 |
+
"clip_ratio/low_min": 0.0,
|
| 606 |
+
"clip_ratio/region_mean": 0.0,
|
| 607 |
+
"completion_length": 3432.5833740234375,
|
| 608 |
+
"entropy": 0.4443359375,
|
| 609 |
+
"epoch": 0.036571428571428574,
|
| 610 |
+
"grad_norm": 0.09518005698919296,
|
| 611 |
+
"kl": 0.0003883838653564453,
|
| 612 |
+
"learning_rate": 8.844151714648274e-07,
|
| 613 |
+
"loss": 0.0,
|
| 614 |
+
"reward": 0.6354166939854622,
|
| 615 |
+
"reward_std": 0.47362349927425385,
|
| 616 |
+
"rewards/accuracy_reward": 0.2708333367481828,
|
| 617 |
+
"rewards/format_reward": 0.36458334140479565,
|
| 618 |
+
"step": 32
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"clip_ratio/high_max": 0.0,
|
| 622 |
+
"clip_ratio/high_mean": 0.0,
|
| 623 |
+
"clip_ratio/low_mean": 0.0,
|
| 624 |
+
"clip_ratio/low_min": 0.0,
|
| 625 |
+
"clip_ratio/region_mean": 0.0,
|
| 626 |
+
"completion_length": 3424.1043090820312,
|
| 627 |
+
"entropy": 0.37109375,
|
| 628 |
+
"epoch": 0.037714285714285714,
|
| 629 |
+
"grad_norm": 0.12133150547742844,
|
| 630 |
+
"kl": 0.0007408857345581055,
|
| 631 |
+
"learning_rate": 8.737029101523929e-07,
|
| 632 |
+
"loss": 0.0,
|
| 633 |
+
"reward": 0.645833358168602,
|
| 634 |
+
"reward_std": 0.5898082852363586,
|
| 635 |
+
"rewards/accuracy_reward": 0.2291666716337204,
|
| 636 |
+
"rewards/format_reward": 0.4166666753590107,
|
| 637 |
+
"step": 33
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"clip_ratio/high_max": 0.0,
|
| 641 |
+
"clip_ratio/high_mean": 0.0,
|
| 642 |
+
"clip_ratio/low_mean": 0.0,
|
| 643 |
+
"clip_ratio/low_min": 0.0,
|
| 644 |
+
"clip_ratio/region_mean": 0.0,
|
| 645 |
+
"completion_length": 2531.0521240234375,
|
| 646 |
+
"entropy": 0.435302734375,
|
| 647 |
+
"epoch": 0.038857142857142854,
|
| 648 |
+
"grad_norm": 0.16009539365768433,
|
| 649 |
+
"kl": 0.0005445480346679688,
|
| 650 |
+
"learning_rate": 8.625962667065487e-07,
|
| 651 |
+
"loss": 0.0,
|
| 652 |
+
"reward": 0.8958333358168602,
|
| 653 |
+
"reward_std": 0.5157222747802734,
|
| 654 |
+
"rewards/accuracy_reward": 0.35416667722165585,
|
| 655 |
+
"rewards/format_reward": 0.5416666744276881,
|
| 656 |
+
"step": 34
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"clip_ratio/high_max": 0.0,
|
| 660 |
+
"clip_ratio/high_mean": 0.0,
|
| 661 |
+
"clip_ratio/low_mean": 0.0,
|
| 662 |
+
"clip_ratio/low_min": 0.0,
|
| 663 |
+
"clip_ratio/region_mean": 0.0,
|
| 664 |
+
"completion_length": 3209.281280517578,
|
| 665 |
+
"entropy": 0.42431640625,
|
| 666 |
+
"epoch": 0.04,
|
| 667 |
+
"grad_norm": 0.14229460060596466,
|
| 668 |
+
"kl": 0.001219034194946289,
|
| 669 |
+
"learning_rate": 8.511087728614862e-07,
|
| 670 |
+
"loss": 0.0,
|
| 671 |
+
"reward": 0.6562500074505806,
|
| 672 |
+
"reward_std": 0.6340432465076447,
|
| 673 |
+
"rewards/accuracy_reward": 0.25000000931322575,
|
| 674 |
+
"rewards/format_reward": 0.4062500111758709,
|
| 675 |
+
"step": 35
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"clip_ratio/high_max": 0.0,
|
| 679 |
+
"clip_ratio/high_mean": 0.0,
|
| 680 |
+
"clip_ratio/low_mean": 0.0,
|
| 681 |
+
"clip_ratio/low_min": 0.0,
|
| 682 |
+
"clip_ratio/region_mean": 0.0,
|
| 683 |
+
"completion_length": 3487.9063720703125,
|
| 684 |
+
"entropy": 0.48828125,
|
| 685 |
+
"epoch": 0.04114285714285714,
|
| 686 |
+
"grad_norm": 0.1350383758544922,
|
| 687 |
+
"kl": 0.0010302066802978516,
|
| 688 |
+
"learning_rate": 8.392544243589427e-07,
|
| 689 |
+
"loss": 0.0,
|
| 690 |
+
"reward": 0.3020833386108279,
|
| 691 |
+
"reward_std": 0.4083498902618885,
|
| 692 |
+
"rewards/accuracy_reward": 0.05208333395421505,
|
| 693 |
+
"rewards/format_reward": 0.2500000027939677,
|
| 694 |
+
"step": 36
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"clip_ratio/high_max": 0.0,
|
| 698 |
+
"clip_ratio/high_mean": 0.0,
|
| 699 |
+
"clip_ratio/low_mean": 0.0,
|
| 700 |
+
"clip_ratio/low_min": 0.0,
|
| 701 |
+
"clip_ratio/region_mean": 0.0,
|
| 702 |
+
"completion_length": 3341.541748046875,
|
| 703 |
+
"entropy": 0.44482421875,
|
| 704 |
+
"epoch": 0.04228571428571429,
|
| 705 |
+
"grad_norm": 0.09916561841964722,
|
| 706 |
+
"kl": 0.0008041858673095703,
|
| 707 |
+
"learning_rate": 8.270476638965461e-07,
|
| 708 |
+
"loss": 0.0,
|
| 709 |
+
"reward": 0.30208334885537624,
|
| 710 |
+
"reward_std": 0.13339675217866898,
|
| 711 |
+
"rewards/accuracy_reward": 0.02083333395421505,
|
| 712 |
+
"rewards/format_reward": 0.28125000558793545,
|
| 713 |
+
"step": 37
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"clip_ratio/high_max": 0.0,
|
| 717 |
+
"clip_ratio/high_mean": 0.0,
|
| 718 |
+
"clip_ratio/low_mean": 0.0,
|
| 719 |
+
"clip_ratio/low_min": 0.0,
|
| 720 |
+
"clip_ratio/region_mean": 0.0,
|
| 721 |
+
"completion_length": 3503.510498046875,
|
| 722 |
+
"entropy": 0.4560546875,
|
| 723 |
+
"epoch": 0.04342857142857143,
|
| 724 |
+
"grad_norm": 0.0784042626619339,
|
| 725 |
+
"kl": 0.00047469139099121094,
|
| 726 |
+
"learning_rate": 8.145033635316128e-07,
|
| 727 |
+
"loss": 0.0,
|
| 728 |
+
"reward": 0.322916679084301,
|
| 729 |
+
"reward_std": 0.22953829169273376,
|
| 730 |
+
"rewards/accuracy_reward": 0.13541666697710752,
|
| 731 |
+
"rewards/format_reward": 0.1875000111758709,
|
| 732 |
+
"step": 38
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"clip_ratio/high_max": 0.0,
|
| 736 |
+
"clip_ratio/high_mean": 0.0,
|
| 737 |
+
"clip_ratio/low_mean": 0.0,
|
| 738 |
+
"clip_ratio/low_min": 0.0,
|
| 739 |
+
"clip_ratio/region_mean": 0.0,
|
| 740 |
+
"completion_length": 3080.7293090820312,
|
| 741 |
+
"entropy": 0.36767578125,
|
| 742 |
+
"epoch": 0.044571428571428574,
|
| 743 |
+
"grad_norm": 0.12402305006980896,
|
| 744 |
+
"kl": 0.0004640817642211914,
|
| 745 |
+
"learning_rate": 8.01636806561836e-07,
|
| 746 |
+
"loss": 0.0,
|
| 747 |
+
"reward": 0.7083333507180214,
|
| 748 |
+
"reward_std": 0.32364466041326523,
|
| 749 |
+
"rewards/accuracy_reward": 0.23958333861082792,
|
| 750 |
+
"rewards/format_reward": 0.4687500074505806,
|
| 751 |
+
"step": 39
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"clip_ratio/high_max": 0.0,
|
| 755 |
+
"clip_ratio/high_mean": 0.0,
|
| 756 |
+
"clip_ratio/low_mean": 0.0,
|
| 757 |
+
"clip_ratio/low_min": 0.0,
|
| 758 |
+
"clip_ratio/region_mean": 0.0,
|
| 759 |
+
"completion_length": 2654.135498046875,
|
| 760 |
+
"entropy": 0.36572265625,
|
| 761 |
+
"epoch": 0.045714285714285714,
|
| 762 |
+
"grad_norm": 0.1607232689857483,
|
| 763 |
+
"kl": 0.0020923614501953125,
|
| 764 |
+
"learning_rate": 7.884636689049422e-07,
|
| 765 |
+
"loss": 0.0001,
|
| 766 |
+
"reward": 0.718750037252903,
|
| 767 |
+
"reward_std": 0.39055147767066956,
|
| 768 |
+
"rewards/accuracy_reward": 0.19791666697710752,
|
| 769 |
+
"rewards/format_reward": 0.520833358168602,
|
| 770 |
+
"step": 40
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"clip_ratio/high_max": 0.0,
|
| 774 |
+
"clip_ratio/high_mean": 0.0,
|
| 775 |
+
"clip_ratio/low_mean": 0.0,
|
| 776 |
+
"clip_ratio/low_min": 0.0,
|
| 777 |
+
"clip_ratio/region_mean": 0.0,
|
| 778 |
+
"completion_length": 3220.4688720703125,
|
| 779 |
+
"entropy": 0.38232421875,
|
| 780 |
+
"epoch": 0.046857142857142854,
|
| 781 |
+
"grad_norm": 0.11851893365383148,
|
| 782 |
+
"kl": 0.0005588531494140625,
|
| 783 |
+
"learning_rate": 7.75e-07,
|
| 784 |
+
"loss": 0.0,
|
| 785 |
+
"reward": 0.5729166865348816,
|
| 786 |
+
"reward_std": 0.5243071913719177,
|
| 787 |
+
"rewards/accuracy_reward": 0.1250000037252903,
|
| 788 |
+
"rewards/format_reward": 0.4479166865348816,
|
| 789 |
+
"step": 41
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"clip_ratio/high_max": 0.0,
|
| 793 |
+
"clip_ratio/high_mean": 0.0,
|
| 794 |
+
"clip_ratio/low_mean": 0.0,
|
| 795 |
+
"clip_ratio/low_min": 0.0,
|
| 796 |
+
"clip_ratio/region_mean": 0.0,
|
| 797 |
+
"completion_length": 3080.8959350585938,
|
| 798 |
+
"entropy": 0.48876953125,
|
| 799 |
+
"epoch": 0.048,
|
| 800 |
+
"grad_norm": 0.057032641023397446,
|
| 801 |
+
"kl": 0.0005335807800292969,
|
| 802 |
+
"learning_rate": 7.612622032536507e-07,
|
| 803 |
+
"loss": 0.0,
|
| 804 |
+
"reward": 0.32291666977107525,
|
| 805 |
+
"reward_std": 0.13653486967086792,
|
| 806 |
+
"rewards/accuracy_reward": 0.010416666977107525,
|
| 807 |
+
"rewards/format_reward": 0.3125000027939677,
|
| 808 |
+
"step": 42
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"clip_ratio/high_max": 0.0,
|
| 812 |
+
"clip_ratio/high_mean": 0.0,
|
| 813 |
+
"clip_ratio/low_mean": 0.0,
|
| 814 |
+
"clip_ratio/low_min": 0.0,
|
| 815 |
+
"clip_ratio/region_mean": 0.0,
|
| 816 |
+
"completion_length": 3166.4584350585938,
|
| 817 |
+
"entropy": 0.40380859375,
|
| 818 |
+
"epoch": 0.04914285714285714,
|
| 819 |
+
"grad_norm": 0.1308462768793106,
|
| 820 |
+
"kl": 0.001255035400390625,
|
| 821 |
+
"learning_rate": 7.472670160550848e-07,
|
| 822 |
+
"loss": 0.0001,
|
| 823 |
+
"reward": 0.3958333432674408,
|
| 824 |
+
"reward_std": 0.26092809066176414,
|
| 825 |
+
"rewards/accuracy_reward": 0.12500000093132257,
|
| 826 |
+
"rewards/format_reward": 0.2708333358168602,
|
| 827 |
+
"step": 43
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"clip_ratio/high_max": 0.0,
|
| 831 |
+
"clip_ratio/high_mean": 0.0,
|
| 832 |
+
"clip_ratio/low_mean": 0.0,
|
| 833 |
+
"clip_ratio/low_min": 0.0,
|
| 834 |
+
"clip_ratio/region_mean": 0.0,
|
| 835 |
+
"completion_length": 2762.635467529297,
|
| 836 |
+
"entropy": 0.3974609375,
|
| 837 |
+
"epoch": 0.05028571428571429,
|
| 838 |
+
"grad_norm": 0.15369124710559845,
|
| 839 |
+
"kl": 0.0009186267852783203,
|
| 840 |
+
"learning_rate": 7.330314893841101e-07,
|
| 841 |
+
"loss": 0.0,
|
| 842 |
+
"reward": 0.854166692122817,
|
| 843 |
+
"reward_std": 0.41533562541007996,
|
| 844 |
+
"rewards/accuracy_reward": 0.3020833507180214,
|
| 845 |
+
"rewards/format_reward": 0.5520833414047956,
|
| 846 |
+
"step": 44
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"clip_ratio/high_max": 0.0,
|
| 850 |
+
"clip_ratio/high_mean": 0.0,
|
| 851 |
+
"clip_ratio/low_mean": 0.0,
|
| 852 |
+
"clip_ratio/low_min": 0.0,
|
| 853 |
+
"clip_ratio/region_mean": 0.0,
|
| 854 |
+
"completion_length": 3691.5521240234375,
|
| 855 |
+
"entropy": 0.4072265625,
|
| 856 |
+
"epoch": 0.05142857142857143,
|
| 857 |
+
"grad_norm": 0.08415117114782333,
|
| 858 |
+
"kl": 0.0013604164123535156,
|
| 859 |
+
"learning_rate": 7.185729670371604e-07,
|
| 860 |
+
"loss": 0.0001,
|
| 861 |
+
"reward": 0.5520833488553762,
|
| 862 |
+
"reward_std": 0.4458491802215576,
|
| 863 |
+
"rewards/accuracy_reward": 0.23958334606140852,
|
| 864 |
+
"rewards/format_reward": 0.3125000102445483,
|
| 865 |
+
"step": 45
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"clip_ratio/high_max": 0.0,
|
| 869 |
+
"clip_ratio/high_mean": 0.0,
|
| 870 |
+
"clip_ratio/low_mean": 0.0,
|
| 871 |
+
"clip_ratio/low_min": 0.0,
|
| 872 |
+
"clip_ratio/region_mean": 0.0,
|
| 873 |
+
"completion_length": 3427.3438720703125,
|
| 874 |
+
"entropy": 0.49755859375,
|
| 875 |
+
"epoch": 0.052571428571428575,
|
| 876 |
+
"grad_norm": 0.13379938900470734,
|
| 877 |
+
"kl": 0.0010638236999511719,
|
| 878 |
+
"learning_rate": 7.039090644965509e-07,
|
| 879 |
+
"loss": 0.0,
|
| 880 |
+
"reward": 0.23958334606140852,
|
| 881 |
+
"reward_std": 0.25305089354515076,
|
| 882 |
+
"rewards/accuracy_reward": 0.031250000931322575,
|
| 883 |
+
"rewards/format_reward": 0.20833333488553762,
|
| 884 |
+
"step": 46
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"clip_ratio/high_max": 0.0,
|
| 888 |
+
"clip_ratio/high_mean": 0.0,
|
| 889 |
+
"clip_ratio/low_mean": 0.0,
|
| 890 |
+
"clip_ratio/low_min": 0.0,
|
| 891 |
+
"clip_ratio/region_mean": 0.0,
|
| 892 |
+
"completion_length": 3061.1250610351562,
|
| 893 |
+
"entropy": 0.4091796875,
|
| 894 |
+
"epoch": 0.053714285714285714,
|
| 895 |
+
"grad_norm": 0.15724660456180573,
|
| 896 |
+
"kl": 0.0008676052093505859,
|
| 897 |
+
"learning_rate": 6.890576474687263e-07,
|
| 898 |
+
"loss": 0.0,
|
| 899 |
+
"reward": 0.6979166939854622,
|
| 900 |
+
"reward_std": 0.6069704815745354,
|
| 901 |
+
"rewards/accuracy_reward": 0.2395833395421505,
|
| 902 |
+
"rewards/format_reward": 0.4583333432674408,
|
| 903 |
+
"step": 47
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"clip_ratio/high_max": 0.0,
|
| 907 |
+
"clip_ratio/high_mean": 0.0,
|
| 908 |
+
"clip_ratio/low_mean": 0.0,
|
| 909 |
+
"clip_ratio/low_min": 0.0,
|
| 910 |
+
"clip_ratio/region_mean": 0.0,
|
| 911 |
+
"completion_length": 2838.947998046875,
|
| 912 |
+
"entropy": 0.38330078125,
|
| 913 |
+
"epoch": 0.054857142857142854,
|
| 914 |
+
"grad_norm": 0.13557472825050354,
|
| 915 |
+
"kl": 0.0023670196533203125,
|
| 916 |
+
"learning_rate": 6.740368101176495e-07,
|
| 917 |
+
"loss": 0.0001,
|
| 918 |
+
"reward": 0.7083333656191826,
|
| 919 |
+
"reward_std": 0.4954182878136635,
|
| 920 |
+
"rewards/accuracy_reward": 0.20833334047347307,
|
| 921 |
+
"rewards/format_reward": 0.5000000111758709,
|
| 922 |
+
"step": 48
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"clip_ratio/high_max": 0.0,
|
| 926 |
+
"clip_ratio/high_mean": 0.0,
|
| 927 |
+
"clip_ratio/low_mean": 0.0,
|
| 928 |
+
"clip_ratio/low_min": 0.0,
|
| 929 |
+
"clip_ratio/region_mean": 0.0,
|
| 930 |
+
"completion_length": 2608.3438110351562,
|
| 931 |
+
"entropy": 0.39111328125,
|
| 932 |
+
"epoch": 0.056,
|
| 933 |
+
"grad_norm": 0.27688586711883545,
|
| 934 |
+
"kl": 0.01327657699584961,
|
| 935 |
+
"learning_rate": 6.588648530198504e-07,
|
| 936 |
+
"loss": 0.0005,
|
| 937 |
+
"reward": 0.7291666865348816,
|
| 938 |
+
"reward_std": 0.3509138487279415,
|
| 939 |
+
"rewards/accuracy_reward": 0.14583333488553762,
|
| 940 |
+
"rewards/format_reward": 0.5833333432674408,
|
| 941 |
+
"step": 49
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"clip_ratio/high_max": 0.0,
|
| 945 |
+
"clip_ratio/high_mean": 0.0,
|
| 946 |
+
"clip_ratio/low_mean": 0.0,
|
| 947 |
+
"clip_ratio/low_min": 0.0,
|
| 948 |
+
"clip_ratio/region_mean": 0.0,
|
| 949 |
+
"completion_length": 3195.7396240234375,
|
| 950 |
+
"entropy": 0.3486328125,
|
| 951 |
+
"epoch": 0.05714285714285714,
|
| 952 |
+
"grad_norm": 0.11028607189655304,
|
| 953 |
+
"kl": 0.0009064674377441406,
|
| 954 |
+
"learning_rate": 6.435602608679916e-07,
|
| 955 |
+
"loss": 0.0,
|
| 956 |
+
"reward": 0.6458333637565374,
|
| 957 |
+
"reward_std": 0.3640809394419193,
|
| 958 |
+
"rewards/accuracy_reward": 0.2812500074505806,
|
| 959 |
+
"rewards/format_reward": 0.36458333395421505,
|
| 960 |
+
"step": 50
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"clip_ratio/high_max": 0.0,
|
| 964 |
+
"clip_ratio/high_mean": 0.0,
|
| 965 |
+
"clip_ratio/low_mean": 0.0,
|
| 966 |
+
"clip_ratio/low_min": 0.0,
|
| 967 |
+
"clip_ratio/region_mean": 0.0,
|
| 968 |
+
"completion_length": 2502.166748046875,
|
| 969 |
+
"entropy": 0.453125,
|
| 970 |
+
"epoch": 0.05828571428571429,
|
| 971 |
+
"grad_norm": 0.13727469742298126,
|
| 972 |
+
"kl": 0.004588127136230469,
|
| 973 |
+
"learning_rate": 6.281416799501187e-07,
|
| 974 |
+
"loss": 0.0002,
|
| 975 |
+
"reward": 0.6354166716337204,
|
| 976 |
+
"reward_std": 0.30697082355618477,
|
| 977 |
+
"rewards/accuracy_reward": 0.1354166679084301,
|
| 978 |
+
"rewards/format_reward": 0.5,
|
| 979 |
+
"step": 51
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"clip_ratio/high_max": 0.0,
|
| 983 |
+
"clip_ratio/high_mean": 0.0,
|
| 984 |
+
"clip_ratio/low_mean": 0.0,
|
| 985 |
+
"clip_ratio/low_min": 0.0,
|
| 986 |
+
"clip_ratio/region_mean": 0.0,
|
| 987 |
+
"completion_length": 3103.4688110351562,
|
| 988 |
+
"entropy": 0.404296875,
|
| 989 |
+
"epoch": 0.05942857142857143,
|
| 990 |
+
"grad_norm": 0.11003529280424118,
|
| 991 |
+
"kl": 0.001992940902709961,
|
| 992 |
+
"learning_rate": 6.126278954320294e-07,
|
| 993 |
+
"loss": 0.0001,
|
| 994 |
+
"reward": 0.7500000102445483,
|
| 995 |
+
"reward_std": 0.47105513513088226,
|
| 996 |
+
"rewards/accuracy_reward": 0.3020833358168602,
|
| 997 |
+
"rewards/format_reward": 0.4479166669771075,
|
| 998 |
+
"step": 52
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"clip_ratio/high_max": 0.0,
|
| 1002 |
+
"clip_ratio/high_mean": 0.0,
|
| 1003 |
+
"clip_ratio/low_mean": 0.0,
|
| 1004 |
+
"clip_ratio/low_min": 0.0,
|
| 1005 |
+
"clip_ratio/region_mean": 0.0,
|
| 1006 |
+
"completion_length": 2895.6146850585938,
|
| 1007 |
+
"entropy": 0.41748046875,
|
| 1008 |
+
"epoch": 0.060571428571428575,
|
| 1009 |
+
"grad_norm": 0.11098165810108185,
|
| 1010 |
+
"kl": 0.0010890960693359375,
|
| 1011 |
+
"learning_rate": 5.97037808470444e-07,
|
| 1012 |
+
"loss": 0.0,
|
| 1013 |
+
"reward": 0.895833358168602,
|
| 1014 |
+
"reward_std": 0.6053251773118973,
|
| 1015 |
+
"rewards/accuracy_reward": 0.3020833469927311,
|
| 1016 |
+
"rewards/format_reward": 0.5937500074505806,
|
| 1017 |
+
"step": 53
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"clip_ratio/high_max": 0.0,
|
| 1021 |
+
"clip_ratio/high_mean": 0.0,
|
| 1022 |
+
"clip_ratio/low_mean": 0.0,
|
| 1023 |
+
"clip_ratio/low_min": 0.0,
|
| 1024 |
+
"clip_ratio/region_mean": 0.0,
|
| 1025 |
+
"completion_length": 2898.0938720703125,
|
| 1026 |
+
"entropy": 0.36865234375,
|
| 1027 |
+
"epoch": 0.061714285714285715,
|
| 1028 |
+
"grad_norm": 0.13266167044639587,
|
| 1029 |
+
"kl": 0.0010752677917480469,
|
| 1030 |
+
"learning_rate": 5.813904131848564e-07,
|
| 1031 |
+
"loss": 0.0,
|
| 1032 |
+
"reward": 0.9791666939854622,
|
| 1033 |
+
"reward_std": 0.6714620888233185,
|
| 1034 |
+
"rewards/accuracy_reward": 0.4062500037252903,
|
| 1035 |
+
"rewards/format_reward": 0.5729166828095913,
|
| 1036 |
+
"step": 54
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"clip_ratio/high_max": 0.0,
|
| 1040 |
+
"clip_ratio/high_mean": 0.0,
|
| 1041 |
+
"clip_ratio/low_mean": 0.0,
|
| 1042 |
+
"clip_ratio/low_min": 0.0,
|
| 1043 |
+
"clip_ratio/region_mean": 0.0,
|
| 1044 |
+
"completion_length": 3296.2084350585938,
|
| 1045 |
+
"entropy": 0.42138671875,
|
| 1046 |
+
"epoch": 0.06285714285714286,
|
| 1047 |
+
"grad_norm": 0.11708013713359833,
|
| 1048 |
+
"kl": 0.0011484622955322266,
|
| 1049 |
+
"learning_rate": 5.657047735161255e-07,
|
| 1050 |
+
"loss": 0.0,
|
| 1051 |
+
"reward": 0.5625000149011612,
|
| 1052 |
+
"reward_std": 0.468790166079998,
|
| 1053 |
+
"rewards/accuracy_reward": 0.18750000465661287,
|
| 1054 |
+
"rewards/format_reward": 0.3750000037252903,
|
| 1055 |
+
"step": 55
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"clip_ratio/high_max": 0.0,
|
| 1059 |
+
"clip_ratio/high_mean": 0.0,
|
| 1060 |
+
"clip_ratio/low_mean": 0.0,
|
| 1061 |
+
"clip_ratio/low_min": 0.0,
|
| 1062 |
+
"clip_ratio/region_mean": 0.0,
|
| 1063 |
+
"completion_length": 3217.760498046875,
|
| 1064 |
+
"entropy": 0.40234375,
|
| 1065 |
+
"epoch": 0.064,
|
| 1066 |
+
"grad_norm": 0.10522608458995819,
|
| 1067 |
+
"kl": 0.0005931854248046875,
|
| 1068 |
+
"learning_rate": 5.5e-07,
|
| 1069 |
+
"loss": 0.0,
|
| 1070 |
+
"reward": 0.5833333414047956,
|
| 1071 |
+
"reward_std": 0.3264950141310692,
|
| 1072 |
+
"rewards/accuracy_reward": 0.1770833432674408,
|
| 1073 |
+
"rewards/format_reward": 0.40625001303851604,
|
| 1074 |
+
"step": 56
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"clip_ratio/high_max": 0.0,
|
| 1078 |
+
"clip_ratio/high_mean": 0.0,
|
| 1079 |
+
"clip_ratio/low_mean": 0.0,
|
| 1080 |
+
"clip_ratio/low_min": 0.0,
|
| 1081 |
+
"clip_ratio/region_mean": 0.0,
|
| 1082 |
+
"completion_length": 3551.9063110351562,
|
| 1083 |
+
"entropy": 0.32470703125,
|
| 1084 |
+
"epoch": 0.06514285714285714,
|
| 1085 |
+
"grad_norm": 0.08617226034402847,
|
| 1086 |
+
"kl": 0.00047588348388671875,
|
| 1087 |
+
"learning_rate": 5.342952264838747e-07,
|
| 1088 |
+
"loss": 0.0,
|
| 1089 |
+
"reward": 0.5312500260770321,
|
| 1090 |
+
"reward_std": 0.5744922757148743,
|
| 1091 |
+
"rewards/accuracy_reward": 0.14583334140479565,
|
| 1092 |
+
"rewards/format_reward": 0.38541668467223644,
|
| 1093 |
+
"step": 57
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"clip_ratio/high_max": 0.0,
|
| 1097 |
+
"clip_ratio/high_mean": 0.0,
|
| 1098 |
+
"clip_ratio/low_mean": 0.0,
|
| 1099 |
+
"clip_ratio/low_min": 0.0,
|
| 1100 |
+
"clip_ratio/region_mean": 0.0,
|
| 1101 |
+
"completion_length": 2254.4375610351562,
|
| 1102 |
+
"entropy": 0.337646484375,
|
| 1103 |
+
"epoch": 0.06628571428571428,
|
| 1104 |
+
"grad_norm": 0.13133276998996735,
|
| 1105 |
+
"kl": 0.0045833587646484375,
|
| 1106 |
+
"learning_rate": 5.186095868151436e-07,
|
| 1107 |
+
"loss": 0.0002,
|
| 1108 |
+
"reward": 1.0208333879709244,
|
| 1109 |
+
"reward_std": 0.42615631967782974,
|
| 1110 |
+
"rewards/accuracy_reward": 0.291666679084301,
|
| 1111 |
+
"rewards/format_reward": 0.7291667014360428,
|
| 1112 |
+
"step": 58
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"clip_ratio/high_max": 0.0,
|
| 1116 |
+
"clip_ratio/high_mean": 0.0,
|
| 1117 |
+
"clip_ratio/low_mean": 0.0,
|
| 1118 |
+
"clip_ratio/low_min": 0.0,
|
| 1119 |
+
"clip_ratio/region_mean": 0.0,
|
| 1120 |
+
"completion_length": 3097.7083740234375,
|
| 1121 |
+
"entropy": 0.3427734375,
|
| 1122 |
+
"epoch": 0.06742857142857143,
|
| 1123 |
+
"grad_norm": 0.12775066494941711,
|
| 1124 |
+
"kl": 0.0004324913024902344,
|
| 1125 |
+
"learning_rate": 5.02962191529556e-07,
|
| 1126 |
+
"loss": 0.0,
|
| 1127 |
+
"reward": 0.5312500009313226,
|
| 1128 |
+
"reward_std": 0.3815914988517761,
|
| 1129 |
+
"rewards/accuracy_reward": 0.1666666716337204,
|
| 1130 |
+
"rewards/format_reward": 0.36458334419876337,
|
| 1131 |
+
"step": 59
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"clip_ratio/high_max": 0.0,
|
| 1135 |
+
"clip_ratio/high_mean": 0.0,
|
| 1136 |
+
"clip_ratio/low_mean": 0.0,
|
| 1137 |
+
"clip_ratio/low_min": 0.0,
|
| 1138 |
+
"clip_ratio/region_mean": 0.0,
|
| 1139 |
+
"completion_length": 3122.7396240234375,
|
| 1140 |
+
"entropy": 0.37744140625,
|
| 1141 |
+
"epoch": 0.06857142857142857,
|
| 1142 |
+
"grad_norm": 0.09946239739656448,
|
| 1143 |
+
"kl": 0.0018048286437988281,
|
| 1144 |
+
"learning_rate": 4.873721045679706e-07,
|
| 1145 |
+
"loss": 0.0001,
|
| 1146 |
+
"reward": 0.5520833432674408,
|
| 1147 |
+
"reward_std": 0.2942040413618088,
|
| 1148 |
+
"rewards/accuracy_reward": 0.13541666977107525,
|
| 1149 |
+
"rewards/format_reward": 0.41666668467223644,
|
| 1150 |
+
"step": 60
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"clip_ratio/high_max": 0.0,
|
| 1154 |
+
"clip_ratio/high_mean": 0.0,
|
| 1155 |
+
"clip_ratio/low_mean": 0.0,
|
| 1156 |
+
"clip_ratio/low_min": 0.0,
|
| 1157 |
+
"clip_ratio/region_mean": 0.0,
|
| 1158 |
+
"completion_length": 3085.6771240234375,
|
| 1159 |
+
"entropy": 0.37646484375,
|
| 1160 |
+
"epoch": 0.06971428571428571,
|
| 1161 |
+
"grad_norm": 0.12444304674863815,
|
| 1162 |
+
"kl": 0.0009784698486328125,
|
| 1163 |
+
"learning_rate": 4.7185832004988133e-07,
|
| 1164 |
+
"loss": 0.0,
|
| 1165 |
+
"reward": 0.8541667088866234,
|
| 1166 |
+
"reward_std": 0.39620716124773026,
|
| 1167 |
+
"rewards/accuracy_reward": 0.2708333432674408,
|
| 1168 |
+
"rewards/format_reward": 0.5833333507180214,
|
| 1169 |
+
"step": 61
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"clip_ratio/high_max": 0.0,
|
| 1173 |
+
"clip_ratio/high_mean": 0.0,
|
| 1174 |
+
"clip_ratio/low_mean": 0.0,
|
| 1175 |
+
"clip_ratio/low_min": 0.0,
|
| 1176 |
+
"clip_ratio/region_mean": 0.0,
|
| 1177 |
+
"completion_length": 2567.8646240234375,
|
| 1178 |
+
"entropy": 0.326171875,
|
| 1179 |
+
"epoch": 0.07085714285714285,
|
| 1180 |
+
"grad_norm": 0.11932501196861267,
|
| 1181 |
+
"kl": 0.001983642578125,
|
| 1182 |
+
"learning_rate": 4.5643973913200837e-07,
|
| 1183 |
+
"loss": 0.0001,
|
| 1184 |
+
"reward": 0.9791666939854622,
|
| 1185 |
+
"reward_std": 0.4785246327519417,
|
| 1186 |
+
"rewards/accuracy_reward": 0.32291666977107525,
|
| 1187 |
+
"rewards/format_reward": 0.6562500149011612,
|
| 1188 |
+
"step": 62
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"clip_ratio/high_max": 0.0,
|
| 1192 |
+
"clip_ratio/high_mean": 0.0,
|
| 1193 |
+
"clip_ratio/low_mean": 0.0,
|
| 1194 |
+
"clip_ratio/low_min": 0.0,
|
| 1195 |
+
"clip_ratio/region_mean": 0.0,
|
| 1196 |
+
"completion_length": 2210.6459045410156,
|
| 1197 |
+
"entropy": 0.3583984375,
|
| 1198 |
+
"epoch": 0.072,
|
| 1199 |
+
"grad_norm": 0.13491779565811157,
|
| 1200 |
+
"kl": 0.0017650127410888672,
|
| 1201 |
+
"learning_rate": 4.4113514698014953e-07,
|
| 1202 |
+
"loss": 0.0001,
|
| 1203 |
+
"reward": 1.0937500447034836,
|
| 1204 |
+
"reward_std": 0.5017538368701935,
|
| 1205 |
+
"rewards/accuracy_reward": 0.3229166716337204,
|
| 1206 |
+
"rewards/format_reward": 0.7708333432674408,
|
| 1207 |
+
"step": 63
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"clip_ratio/high_max": 0.0,
|
| 1211 |
+
"clip_ratio/high_mean": 0.0,
|
| 1212 |
+
"clip_ratio/low_mean": 0.0,
|
| 1213 |
+
"clip_ratio/low_min": 0.0,
|
| 1214 |
+
"clip_ratio/region_mean": 0.0,
|
| 1215 |
+
"completion_length": 2907.500030517578,
|
| 1216 |
+
"entropy": 0.388671875,
|
| 1217 |
+
"epoch": 0.07314285714285715,
|
| 1218 |
+
"grad_norm": 0.11815346777439117,
|
| 1219 |
+
"kl": 0.0008115768432617188,
|
| 1220 |
+
"learning_rate": 4.2596318988235037e-07,
|
| 1221 |
+
"loss": 0.0,
|
| 1222 |
+
"reward": 0.739583358168602,
|
| 1223 |
+
"reward_std": 0.5059329643845558,
|
| 1224 |
+
"rewards/accuracy_reward": 0.2395833358168602,
|
| 1225 |
+
"rewards/format_reward": 0.5000000149011612,
|
| 1226 |
+
"step": 64
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"clip_ratio/high_max": 0.0,
|
| 1230 |
+
"clip_ratio/high_mean": 0.0,
|
| 1231 |
+
"clip_ratio/low_mean": 0.0,
|
| 1232 |
+
"clip_ratio/low_min": 0.0,
|
| 1233 |
+
"clip_ratio/region_mean": 0.0,
|
| 1234 |
+
"completion_length": 2948.166748046875,
|
| 1235 |
+
"entropy": 0.35888671875,
|
| 1236 |
+
"epoch": 0.07428571428571429,
|
| 1237 |
+
"grad_norm": 0.10616273432970047,
|
| 1238 |
+
"kl": 0.0011377334594726562,
|
| 1239 |
+
"learning_rate": 4.1094235253127374e-07,
|
| 1240 |
+
"loss": 0.0,
|
| 1241 |
+
"reward": 0.6458333637565374,
|
| 1242 |
+
"reward_std": 0.3507690578699112,
|
| 1243 |
+
"rewards/accuracy_reward": 0.2083333432674408,
|
| 1244 |
+
"rewards/format_reward": 0.43750001303851604,
|
| 1245 |
+
"step": 65
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"clip_ratio/high_max": 0.0,
|
| 1249 |
+
"clip_ratio/high_mean": 0.0,
|
| 1250 |
+
"clip_ratio/low_mean": 0.0,
|
| 1251 |
+
"clip_ratio/low_min": 0.0,
|
| 1252 |
+
"clip_ratio/region_mean": 0.0,
|
| 1253 |
+
"completion_length": 2217.9792098999023,
|
| 1254 |
+
"entropy": 0.32080078125,
|
| 1255 |
+
"epoch": 0.07542857142857143,
|
| 1256 |
+
"grad_norm": 0.12008016556501389,
|
| 1257 |
+
"kl": 0.0014886856079101562,
|
| 1258 |
+
"learning_rate": 3.9609093550344907e-07,
|
| 1259 |
+
"loss": 0.0001,
|
| 1260 |
+
"reward": 0.8437500204890966,
|
| 1261 |
+
"reward_std": 0.2087036520242691,
|
| 1262 |
+
"rewards/accuracy_reward": 0.33333334885537624,
|
| 1263 |
+
"rewards/format_reward": 0.5104166669771075,
|
| 1264 |
+
"step": 66
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"clip_ratio/high_max": 0.0,
|
| 1268 |
+
"clip_ratio/high_mean": 0.0,
|
| 1269 |
+
"clip_ratio/low_mean": 0.0,
|
| 1270 |
+
"clip_ratio/low_min": 0.0,
|
| 1271 |
+
"clip_ratio/region_mean": 0.0,
|
| 1272 |
+
"completion_length": 3613.9375610351562,
|
| 1273 |
+
"entropy": 0.3525390625,
|
| 1274 |
+
"epoch": 0.07657142857142857,
|
| 1275 |
+
"grad_norm": 0.06719610840082169,
|
| 1276 |
+
"kl": 0.001219034194946289,
|
| 1277 |
+
"learning_rate": 3.8142703296283953e-07,
|
| 1278 |
+
"loss": 0.0,
|
| 1279 |
+
"reward": 0.23958334606140852,
|
| 1280 |
+
"reward_std": 0.245176762342453,
|
| 1281 |
+
"rewards/accuracy_reward": 0.03125,
|
| 1282 |
+
"rewards/format_reward": 0.20833333861082792,
|
| 1283 |
+
"step": 67
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"clip_ratio/high_max": 0.0,
|
| 1287 |
+
"clip_ratio/high_mean": 0.0,
|
| 1288 |
+
"clip_ratio/low_mean": 0.0,
|
| 1289 |
+
"clip_ratio/low_min": 0.0,
|
| 1290 |
+
"clip_ratio/region_mean": 0.0,
|
| 1291 |
+
"completion_length": 2248.8021697998047,
|
| 1292 |
+
"entropy": 0.38134765625,
|
| 1293 |
+
"epoch": 0.07771428571428571,
|
| 1294 |
+
"grad_norm": 0.183990478515625,
|
| 1295 |
+
"kl": 0.00318145751953125,
|
| 1296 |
+
"learning_rate": 3.6696851061588994e-07,
|
| 1297 |
+
"loss": 0.0001,
|
| 1298 |
+
"reward": 0.8437500298023224,
|
| 1299 |
+
"reward_std": 0.49483707919716835,
|
| 1300 |
+
"rewards/accuracy_reward": 0.2500000102445483,
|
| 1301 |
+
"rewards/format_reward": 0.5937500074505806,
|
| 1302 |
+
"step": 68
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"clip_ratio/high_max": 0.0,
|
| 1306 |
+
"clip_ratio/high_mean": 0.0,
|
| 1307 |
+
"clip_ratio/low_mean": 0.0,
|
| 1308 |
+
"clip_ratio/low_min": 0.0,
|
| 1309 |
+
"clip_ratio/region_mean": 0.0,
|
| 1310 |
+
"completion_length": 2372.416748046875,
|
| 1311 |
+
"entropy": 0.40625,
|
| 1312 |
+
"epoch": 0.07885714285714286,
|
| 1313 |
+
"grad_norm": 0.16380108892917633,
|
| 1314 |
+
"kl": 0.0032024383544921875,
|
| 1315 |
+
"learning_rate": 3.5273298394491515e-07,
|
| 1316 |
+
"loss": 0.0001,
|
| 1317 |
+
"reward": 0.7187500298023224,
|
| 1318 |
+
"reward_std": 0.420441672205925,
|
| 1319 |
+
"rewards/accuracy_reward": 0.12500000279396772,
|
| 1320 |
+
"rewards/format_reward": 0.5937500223517418,
|
| 1321 |
+
"step": 69
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"clip_ratio/high_max": 0.0,
|
| 1325 |
+
"clip_ratio/high_mean": 0.0,
|
| 1326 |
+
"clip_ratio/low_mean": 0.0,
|
| 1327 |
+
"clip_ratio/low_min": 0.0,
|
| 1328 |
+
"clip_ratio/region_mean": 0.0,
|
| 1329 |
+
"completion_length": 3206.5521850585938,
|
| 1330 |
+
"entropy": 0.3671875,
|
| 1331 |
+
"epoch": 0.08,
|
| 1332 |
+
"grad_norm": 0.10725324600934982,
|
| 1333 |
+
"kl": 0.0023908615112304688,
|
| 1334 |
+
"learning_rate": 3.387377967463493e-07,
|
| 1335 |
+
"loss": 0.0001,
|
| 1336 |
+
"reward": 0.4895833535119891,
|
| 1337 |
+
"reward_std": 0.3960477001965046,
|
| 1338 |
+
"rewards/accuracy_reward": 0.0833333358168602,
|
| 1339 |
+
"rewards/format_reward": 0.4062500102445483,
|
| 1340 |
+
"step": 70
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"clip_ratio/high_max": 0.0,
|
| 1344 |
+
"clip_ratio/high_mean": 0.0,
|
| 1345 |
+
"clip_ratio/low_mean": 0.0,
|
| 1346 |
+
"clip_ratio/low_min": 0.0,
|
| 1347 |
+
"clip_ratio/region_mean": 0.0,
|
| 1348 |
+
"completion_length": 2797.8958435058594,
|
| 1349 |
+
"entropy": 0.40966796875,
|
| 1350 |
+
"epoch": 0.08114285714285714,
|
| 1351 |
+
"grad_norm": 0.15477971732616425,
|
| 1352 |
+
"kl": 0.00348663330078125,
|
| 1353 |
+
"learning_rate": 3.250000000000001e-07,
|
| 1354 |
+
"loss": 0.0001,
|
| 1355 |
+
"reward": 0.6458333535119891,
|
| 1356 |
+
"reward_std": 0.31787312403321266,
|
| 1357 |
+
"rewards/accuracy_reward": 0.197916679084301,
|
| 1358 |
+
"rewards/format_reward": 0.4479166669771075,
|
| 1359 |
+
"step": 71
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"clip_ratio/high_max": 0.0,
|
| 1363 |
+
"clip_ratio/high_mean": 0.0,
|
| 1364 |
+
"clip_ratio/low_mean": 0.0,
|
| 1365 |
+
"clip_ratio/low_min": 0.0,
|
| 1366 |
+
"clip_ratio/region_mean": 0.0,
|
| 1367 |
+
"completion_length": 2927.3958740234375,
|
| 1368 |
+
"entropy": 0.46826171875,
|
| 1369 |
+
"epoch": 0.08228571428571428,
|
| 1370 |
+
"grad_norm": 0.22557654976844788,
|
| 1371 |
+
"kl": 0.0026330947875976562,
|
| 1372 |
+
"learning_rate": 3.115363310950578e-07,
|
| 1373 |
+
"loss": 0.0001,
|
| 1374 |
+
"reward": 0.510416679084301,
|
| 1375 |
+
"reward_std": 0.47425539791584015,
|
| 1376 |
+
"rewards/accuracy_reward": 0.052083334885537624,
|
| 1377 |
+
"rewards/format_reward": 0.4583333432674408,
|
| 1378 |
+
"step": 72
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"clip_ratio/high_max": 0.0,
|
| 1382 |
+
"clip_ratio/high_mean": 0.0,
|
| 1383 |
+
"clip_ratio/low_mean": 0.0,
|
| 1384 |
+
"clip_ratio/low_min": 0.0,
|
| 1385 |
+
"clip_ratio/region_mean": 0.0,
|
| 1386 |
+
"completion_length": 3653.78125,
|
| 1387 |
+
"entropy": 0.47802734375,
|
| 1388 |
+
"epoch": 0.08342857142857144,
|
| 1389 |
+
"grad_norm": 0.08231060206890106,
|
| 1390 |
+
"kl": 0.0008916854858398438,
|
| 1391 |
+
"learning_rate": 2.9836319343816397e-07,
|
| 1392 |
+
"loss": 0.0,
|
| 1393 |
+
"reward": 0.385416679084301,
|
| 1394 |
+
"reward_std": 0.3866991028189659,
|
| 1395 |
+
"rewards/accuracy_reward": 0.13541666977107525,
|
| 1396 |
+
"rewards/format_reward": 0.25,
|
| 1397 |
+
"step": 73
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"clip_ratio/high_max": 0.0,
|
| 1401 |
+
"clip_ratio/high_mean": 0.0,
|
| 1402 |
+
"clip_ratio/low_mean": 0.0,
|
| 1403 |
+
"clip_ratio/low_min": 0.0,
|
| 1404 |
+
"clip_ratio/region_mean": 0.0,
|
| 1405 |
+
"completion_length": 3289.0834350585938,
|
| 1406 |
+
"entropy": 0.38671875,
|
| 1407 |
+
"epoch": 0.08457142857142858,
|
| 1408 |
+
"grad_norm": 0.11783669143915176,
|
| 1409 |
+
"kl": 0.001605987548828125,
|
| 1410 |
+
"learning_rate": 2.854966364683872e-07,
|
| 1411 |
+
"loss": 0.0001,
|
| 1412 |
+
"reward": 0.6145833432674408,
|
| 1413 |
+
"reward_std": 0.4752888083457947,
|
| 1414 |
+
"rewards/accuracy_reward": 0.2604166716337204,
|
| 1415 |
+
"rewards/format_reward": 0.3541666679084301,
|
| 1416 |
+
"step": 74
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"clip_ratio/high_max": 0.0,
|
| 1420 |
+
"clip_ratio/high_mean": 0.0,
|
| 1421 |
+
"clip_ratio/low_mean": 0.0,
|
| 1422 |
+
"clip_ratio/low_min": 0.0,
|
| 1423 |
+
"clip_ratio/region_mean": 0.0,
|
| 1424 |
+
"completion_length": 3183.7916870117188,
|
| 1425 |
+
"entropy": 0.39501953125,
|
| 1426 |
+
"epoch": 0.08571428571428572,
|
| 1427 |
+
"grad_norm": 0.10990124195814133,
|
| 1428 |
+
"kl": 0.0020608901977539062,
|
| 1429 |
+
"learning_rate": 2.729523361034538e-07,
|
| 1430 |
+
"loss": 0.0001,
|
| 1431 |
+
"reward": 0.5833333432674408,
|
| 1432 |
+
"reward_std": 0.29638050496578217,
|
| 1433 |
+
"rewards/accuracy_reward": 0.1770833358168602,
|
| 1434 |
+
"rewards/format_reward": 0.4062500149011612,
|
| 1435 |
+
"step": 75
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"clip_ratio/high_max": 0.0,
|
| 1439 |
+
"clip_ratio/high_mean": 0.0,
|
| 1440 |
+
"clip_ratio/low_mean": 0.0,
|
| 1441 |
+
"clip_ratio/low_min": 0.0,
|
| 1442 |
+
"clip_ratio/region_mean": 0.0,
|
| 1443 |
+
"completion_length": 2836.3750610351562,
|
| 1444 |
+
"entropy": 0.40185546875,
|
| 1445 |
+
"epoch": 0.08685714285714285,
|
| 1446 |
+
"grad_norm": 0.13064220547676086,
|
| 1447 |
+
"kl": 0.000682830810546875,
|
| 1448 |
+
"learning_rate": 2.6074557564105724e-07,
|
| 1449 |
+
"loss": 0.0,
|
| 1450 |
+
"reward": 0.5729166716337204,
|
| 1451 |
+
"reward_std": 0.33378810063004494,
|
| 1452 |
+
"rewards/accuracy_reward": 0.07291666697710752,
|
| 1453 |
+
"rewards/format_reward": 0.5,
|
| 1454 |
+
"step": 76
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"clip_ratio/high_max": 0.0,
|
| 1458 |
+
"clip_ratio/high_mean": 0.0,
|
| 1459 |
+
"clip_ratio/low_mean": 0.0,
|
| 1460 |
+
"clip_ratio/low_min": 0.0,
|
| 1461 |
+
"clip_ratio/region_mean": 0.0,
|
| 1462 |
+
"completion_length": 3366.4688720703125,
|
| 1463 |
+
"entropy": 0.38916015625,
|
| 1464 |
+
"epoch": 0.088,
|
| 1465 |
+
"grad_norm": 0.08818019181489944,
|
| 1466 |
+
"kl": 0.0007886886596679688,
|
| 1467 |
+
"learning_rate": 2.488912271385139e-07,
|
| 1468 |
+
"loss": 0.0,
|
| 1469 |
+
"reward": 0.5000000009313226,
|
| 1470 |
+
"reward_std": 0.370398610830307,
|
| 1471 |
+
"rewards/accuracy_reward": 0.125,
|
| 1472 |
+
"rewards/format_reward": 0.3750000009313226,
|
| 1473 |
+
"step": 77
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"clip_ratio/high_max": 0.0,
|
| 1477 |
+
"clip_ratio/high_mean": 0.0,
|
| 1478 |
+
"clip_ratio/low_mean": 0.0,
|
| 1479 |
+
"clip_ratio/low_min": 0.0,
|
| 1480 |
+
"clip_ratio/region_mean": 0.0,
|
| 1481 |
+
"completion_length": 3164.4688720703125,
|
| 1482 |
+
"entropy": 0.38525390625,
|
| 1483 |
+
"epoch": 0.08914285714285715,
|
| 1484 |
+
"grad_norm": 0.1075669676065445,
|
| 1485 |
+
"kl": 0.00061798095703125,
|
| 1486 |
+
"learning_rate": 2.374037332934512e-07,
|
| 1487 |
+
"loss": 0.0,
|
| 1488 |
+
"reward": 0.6354166865348816,
|
| 1489 |
+
"reward_std": 0.5042389929294586,
|
| 1490 |
+
"rewards/accuracy_reward": 0.2187500111758709,
|
| 1491 |
+
"rewards/format_reward": 0.416666679084301,
|
| 1492 |
+
"step": 78
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"clip_ratio/high_max": 0.0,
|
| 1496 |
+
"clip_ratio/high_mean": 0.0,
|
| 1497 |
+
"clip_ratio/low_mean": 0.0,
|
| 1498 |
+
"clip_ratio/low_min": 0.0,
|
| 1499 |
+
"clip_ratio/region_mean": 0.0,
|
| 1500 |
+
"completion_length": 2405.572982788086,
|
| 1501 |
+
"entropy": 0.33349609375,
|
| 1502 |
+
"epoch": 0.09028571428571429,
|
| 1503 |
+
"grad_norm": 0.09348474442958832,
|
| 1504 |
+
"kl": 0.0013494491577148438,
|
| 1505 |
+
"learning_rate": 2.2629708984760706e-07,
|
| 1506 |
+
"loss": 0.0001,
|
| 1507 |
+
"reward": 0.9270833730697632,
|
| 1508 |
+
"reward_std": 0.3268127888441086,
|
| 1509 |
+
"rewards/accuracy_reward": 0.2604166744276881,
|
| 1510 |
+
"rewards/format_reward": 0.6666666865348816,
|
| 1511 |
+
"step": 79
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"clip_ratio/high_max": 0.0,
|
| 1515 |
+
"clip_ratio/high_mean": 0.0,
|
| 1516 |
+
"clip_ratio/low_mean": 0.0,
|
| 1517 |
+
"clip_ratio/low_min": 0.0,
|
| 1518 |
+
"clip_ratio/region_mean": 0.0,
|
| 1519 |
+
"completion_length": 3390.2604370117188,
|
| 1520 |
+
"entropy": 0.48388671875,
|
| 1521 |
+
"epoch": 0.09142857142857143,
|
| 1522 |
+
"grad_norm": 0.1434425562620163,
|
| 1523 |
+
"kl": 0.0012698173522949219,
|
| 1524 |
+
"learning_rate": 2.1558482853517253e-07,
|
| 1525 |
+
"loss": 0.0001,
|
| 1526 |
+
"reward": 0.5208333507180214,
|
| 1527 |
+
"reward_std": 0.44375673681497574,
|
| 1528 |
+
"rewards/accuracy_reward": 0.1562500074505806,
|
| 1529 |
+
"rewards/format_reward": 0.3645833432674408,
|
| 1530 |
+
"step": 80
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"clip_ratio/high_max": 0.0,
|
| 1534 |
+
"clip_ratio/high_mean": 0.0,
|
| 1535 |
+
"clip_ratio/low_mean": 0.0,
|
| 1536 |
+
"clip_ratio/low_min": 0.0,
|
| 1537 |
+
"clip_ratio/region_mean": 0.0,
|
| 1538 |
+
"completion_length": 3077.854248046875,
|
| 1539 |
+
"entropy": 0.533203125,
|
| 1540 |
+
"epoch": 0.09257142857142857,
|
| 1541 |
+
"grad_norm": 0.08657144755125046,
|
| 1542 |
+
"kl": 0.00328826904296875,
|
| 1543 |
+
"learning_rate": 2.0528000059645995e-07,
|
| 1544 |
+
"loss": 0.0001,
|
| 1545 |
+
"reward": 0.5208333432674408,
|
| 1546 |
+
"reward_std": 0.27500855177640915,
|
| 1547 |
+
"rewards/accuracy_reward": 0.13541666697710752,
|
| 1548 |
+
"rewards/format_reward": 0.3854166716337204,
|
| 1549 |
+
"step": 81
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"clip_ratio/high_max": 0.0,
|
| 1553 |
+
"clip_ratio/high_mean": 0.0,
|
| 1554 |
+
"clip_ratio/low_mean": 0.0,
|
| 1555 |
+
"clip_ratio/low_min": 0.0,
|
| 1556 |
+
"clip_ratio/region_mean": 0.0,
|
| 1557 |
+
"completion_length": 3044.9480590820312,
|
| 1558 |
+
"entropy": 0.4287109375,
|
| 1559 |
+
"epoch": 0.09371428571428571,
|
| 1560 |
+
"grad_norm": 0.1634312868118286,
|
| 1561 |
+
"kl": 0.001422882080078125,
|
| 1562 |
+
"learning_rate": 1.9539516087697517e-07,
|
| 1563 |
+
"loss": 0.0001,
|
| 1564 |
+
"reward": 0.6666666865348816,
|
| 1565 |
+
"reward_std": 0.4855338782072067,
|
| 1566 |
+
"rewards/accuracy_reward": 0.2291666716337204,
|
| 1567 |
+
"rewards/format_reward": 0.4375000149011612,
|
| 1568 |
+
"step": 82
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"clip_ratio/high_max": 0.0,
|
| 1572 |
+
"clip_ratio/high_mean": 0.0,
|
| 1573 |
+
"clip_ratio/low_mean": 0.0,
|
| 1574 |
+
"clip_ratio/low_min": 0.0,
|
| 1575 |
+
"clip_ratio/region_mean": 0.0,
|
| 1576 |
+
"completion_length": 2948.4896850585938,
|
| 1577 |
+
"entropy": 0.43408203125,
|
| 1578 |
+
"epoch": 0.09485714285714286,
|
| 1579 |
+
"grad_norm": 0.13163359463214874,
|
| 1580 |
+
"kl": 0.0017070770263671875,
|
| 1581 |
+
"learning_rate": 1.8594235253127372e-07,
|
| 1582 |
+
"loss": 0.0001,
|
| 1583 |
+
"reward": 0.5208333544433117,
|
| 1584 |
+
"reward_std": 0.3695184215903282,
|
| 1585 |
+
"rewards/accuracy_reward": 0.16666667256504297,
|
| 1586 |
+
"rewards/format_reward": 0.35416667722165585,
|
| 1587 |
+
"step": 83
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"clip_ratio/high_max": 0.0,
|
| 1591 |
+
"clip_ratio/high_mean": 0.0,
|
| 1592 |
+
"clip_ratio/low_mean": 0.0,
|
| 1593 |
+
"clip_ratio/low_min": 0.0,
|
| 1594 |
+
"clip_ratio/region_mean": 0.0,
|
| 1595 |
+
"completion_length": 3189.625030517578,
|
| 1596 |
+
"entropy": 0.43603515625,
|
| 1597 |
+
"epoch": 0.096,
|
| 1598 |
+
"grad_norm": 0.1409156173467636,
|
| 1599 |
+
"kl": 0.0008630752563476562,
|
| 1600 |
+
"learning_rate": 1.7693309235023127e-07,
|
| 1601 |
+
"loss": 0.0,
|
| 1602 |
+
"reward": 0.677083358168602,
|
| 1603 |
+
"reward_std": 0.49923326447606087,
|
| 1604 |
+
"rewards/accuracy_reward": 0.2395833358168602,
|
| 1605 |
+
"rewards/format_reward": 0.4375000074505806,
|
| 1606 |
+
"step": 84
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"clip_ratio/high_max": 0.0,
|
| 1610 |
+
"clip_ratio/high_mean": 0.0,
|
| 1611 |
+
"clip_ratio/low_mean": 0.0,
|
| 1612 |
+
"clip_ratio/low_min": 0.0,
|
| 1613 |
+
"clip_ratio/region_mean": 0.0,
|
| 1614 |
+
"completion_length": 3121.0938720703125,
|
| 1615 |
+
"entropy": 0.32177734375,
|
| 1616 |
+
"epoch": 0.09714285714285714,
|
| 1617 |
+
"grad_norm": 0.10903389751911163,
|
| 1618 |
+
"kl": 0.0006976127624511719,
|
| 1619 |
+
"learning_rate": 1.6837835672960831e-07,
|
| 1620 |
+
"loss": 0.0,
|
| 1621 |
+
"reward": 0.7187500149011612,
|
| 1622 |
+
"reward_std": 0.6352673023939133,
|
| 1623 |
+
"rewards/accuracy_reward": 0.1875000074505806,
|
| 1624 |
+
"rewards/format_reward": 0.5312500223517418,
|
| 1625 |
+
"step": 85
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"clip_ratio/high_max": 0.0,
|
| 1629 |
+
"clip_ratio/high_mean": 0.0,
|
| 1630 |
+
"clip_ratio/low_mean": 0.0,
|
| 1631 |
+
"clip_ratio/low_min": 0.0,
|
| 1632 |
+
"clip_ratio/region_mean": 0.0,
|
| 1633 |
+
"completion_length": 2938.385498046875,
|
| 1634 |
+
"entropy": 0.4287109375,
|
| 1635 |
+
"epoch": 0.09828571428571428,
|
| 1636 |
+
"grad_norm": 0.14300768077373505,
|
| 1637 |
+
"kl": 0.0020380020141601562,
|
| 1638 |
+
"learning_rate": 1.6028856829700258e-07,
|
| 1639 |
+
"loss": 0.0001,
|
| 1640 |
+
"reward": 0.6041666818782687,
|
| 1641 |
+
"reward_std": 0.26244670152664185,
|
| 1642 |
+
"rewards/accuracy_reward": 0.1562500037252903,
|
| 1643 |
+
"rewards/format_reward": 0.4479166744276881,
|
| 1644 |
+
"step": 86
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"clip_ratio/high_max": 0.0,
|
| 1648 |
+
"clip_ratio/high_mean": 0.0,
|
| 1649 |
+
"clip_ratio/low_mean": 0.0,
|
| 1650 |
+
"clip_ratio/low_min": 0.0,
|
| 1651 |
+
"clip_ratio/region_mean": 0.0,
|
| 1652 |
+
"completion_length": 3037.6563110351562,
|
| 1653 |
+
"entropy": 0.50390625,
|
| 1654 |
+
"epoch": 0.09942857142857142,
|
| 1655 |
+
"grad_norm": 0.16961807012557983,
|
| 1656 |
+
"kl": 0.0024051666259765625,
|
| 1657 |
+
"learning_rate": 1.5267358321348285e-07,
|
| 1658 |
+
"loss": 0.0001,
|
| 1659 |
+
"reward": 0.7395833432674408,
|
| 1660 |
+
"reward_std": 0.5077016204595566,
|
| 1661 |
+
"rewards/accuracy_reward": 0.1979166716337204,
|
| 1662 |
+
"rewards/format_reward": 0.541666679084301,
|
| 1663 |
+
"step": 87
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"clip_ratio/high_max": 0.0,
|
| 1667 |
+
"clip_ratio/high_mean": 0.0,
|
| 1668 |
+
"clip_ratio/low_mean": 0.0,
|
| 1669 |
+
"clip_ratio/low_min": 0.0,
|
| 1670 |
+
"clip_ratio/region_mean": 0.0,
|
| 1671 |
+
"completion_length": 2826.947998046875,
|
| 1672 |
+
"entropy": 0.439453125,
|
| 1673 |
+
"epoch": 0.10057142857142858,
|
| 1674 |
+
"grad_norm": 0.14141874015331268,
|
| 1675 |
+
"kl": 0.0017194747924804688,
|
| 1676 |
+
"learning_rate": 1.4554267916537495e-07,
|
| 1677 |
+
"loss": 0.0001,
|
| 1678 |
+
"reward": 0.9062500298023224,
|
| 1679 |
+
"reward_std": 0.5221241116523743,
|
| 1680 |
+
"rewards/accuracy_reward": 0.3333333358168602,
|
| 1681 |
+
"rewards/format_reward": 0.572916679084301,
|
| 1682 |
+
"step": 88
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"clip_ratio/high_max": 0.0,
|
| 1686 |
+
"clip_ratio/high_mean": 0.0,
|
| 1687 |
+
"clip_ratio/low_mean": 0.0,
|
| 1688 |
+
"clip_ratio/low_min": 0.0,
|
| 1689 |
+
"clip_ratio/region_mean": 0.0,
|
| 1690 |
+
"completion_length": 2963.260498046875,
|
| 1691 |
+
"entropy": 0.36962890625,
|
| 1692 |
+
"epoch": 0.10171428571428572,
|
| 1693 |
+
"grad_norm": 0.16021747887134552,
|
| 1694 |
+
"kl": 0.0021514892578125,
|
| 1695 |
+
"learning_rate": 1.3890454406082956e-07,
|
| 1696 |
+
"loss": 0.0001,
|
| 1697 |
+
"reward": 0.6250000149011612,
|
| 1698 |
+
"reward_std": 0.4761141389608383,
|
| 1699 |
+
"rewards/accuracy_reward": 0.1562500037252903,
|
| 1700 |
+
"rewards/format_reward": 0.4687500149011612,
|
| 1701 |
+
"step": 89
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"clip_ratio/high_max": 0.0,
|
| 1705 |
+
"clip_ratio/high_mean": 0.0,
|
| 1706 |
+
"clip_ratio/low_mean": 0.0,
|
| 1707 |
+
"clip_ratio/low_min": 0.0,
|
| 1708 |
+
"clip_ratio/region_mean": 0.0,
|
| 1709 |
+
"completion_length": 2510.8958740234375,
|
| 1710 |
+
"entropy": 0.5478515625,
|
| 1711 |
+
"epoch": 0.10285714285714286,
|
| 1712 |
+
"grad_norm": 0.08637464791536331,
|
| 1713 |
+
"kl": 0.0036640167236328125,
|
| 1714 |
+
"learning_rate": 1.3276726544494571e-07,
|
| 1715 |
+
"loss": 0.0001,
|
| 1716 |
+
"reward": 0.5520833386108279,
|
| 1717 |
+
"reward_std": 0.18445607274770737,
|
| 1718 |
+
"rewards/accuracy_reward": 0.031250000931322575,
|
| 1719 |
+
"rewards/format_reward": 0.5208333432674408,
|
| 1720 |
+
"step": 90
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"clip_ratio/high_max": 0.0,
|
| 1724 |
+
"clip_ratio/high_mean": 0.0,
|
| 1725 |
+
"clip_ratio/low_mean": 0.0,
|
| 1726 |
+
"clip_ratio/low_min": 0.0,
|
| 1727 |
+
"clip_ratio/region_mean": 0.0,
|
| 1728 |
+
"completion_length": 3139.1771850585938,
|
| 1729 |
+
"entropy": 0.42236328125,
|
| 1730 |
+
"epoch": 0.104,
|
| 1731 |
+
"grad_norm": 0.09102143347263336,
|
| 1732 |
+
"kl": 0.0018377304077148438,
|
| 1733 |
+
"learning_rate": 1.2713832064634125e-07,
|
| 1734 |
+
"loss": 0.0001,
|
| 1735 |
+
"reward": 0.6250000018626451,
|
| 1736 |
+
"reward_std": 0.40364614874124527,
|
| 1737 |
+
"rewards/accuracy_reward": 0.19791667256504297,
|
| 1738 |
+
"rewards/format_reward": 0.42708334419876337,
|
| 1739 |
+
"step": 91
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"clip_ratio/high_max": 0.0,
|
| 1743 |
+
"clip_ratio/high_mean": 0.0,
|
| 1744 |
+
"clip_ratio/low_mean": 0.0,
|
| 1745 |
+
"clip_ratio/low_min": 0.0,
|
| 1746 |
+
"clip_ratio/region_mean": 0.0,
|
| 1747 |
+
"completion_length": 2716.0938415527344,
|
| 1748 |
+
"entropy": 0.354736328125,
|
| 1749 |
+
"epoch": 0.10514285714285715,
|
| 1750 |
+
"grad_norm": 0.12420715391635895,
|
| 1751 |
+
"kl": 0.002556324005126953,
|
| 1752 |
+
"learning_rate": 1.220245676671809e-07,
|
| 1753 |
+
"loss": 0.0001,
|
| 1754 |
+
"reward": 0.6354166939854622,
|
| 1755 |
+
"reward_std": 0.4150289446115494,
|
| 1756 |
+
"rewards/accuracy_reward": 0.1041666716337204,
|
| 1757 |
+
"rewards/format_reward": 0.5312500223517418,
|
| 1758 |
+
"step": 92
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"clip_ratio/high_max": 0.0,
|
| 1762 |
+
"clip_ratio/high_mean": 0.0,
|
| 1763 |
+
"clip_ratio/low_mean": 0.0,
|
| 1764 |
+
"clip_ratio/low_min": 0.0,
|
| 1765 |
+
"clip_ratio/region_mean": 0.0,
|
| 1766 |
+
"completion_length": 3754.8229370117188,
|
| 1767 |
+
"entropy": 0.5556640625,
|
| 1768 |
+
"epoch": 0.10628571428571429,
|
| 1769 |
+
"grad_norm": 0.1010187566280365,
|
| 1770 |
+
"kl": 0.0030183792114257812,
|
| 1771 |
+
"learning_rate": 1.1743223682775649e-07,
|
| 1772 |
+
"loss": 0.0001,
|
| 1773 |
+
"reward": 0.10416667256504297,
|
| 1774 |
+
"reward_std": 0.14910665899515152,
|
| 1775 |
+
"rewards/accuracy_reward": 0.0,
|
| 1776 |
+
"rewards/format_reward": 0.10416667256504297,
|
| 1777 |
+
"step": 93
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"clip_ratio/high_max": 0.0,
|
| 1781 |
+
"clip_ratio/high_mean": 0.0,
|
| 1782 |
+
"clip_ratio/low_mean": 0.0,
|
| 1783 |
+
"clip_ratio/low_min": 0.0,
|
| 1784 |
+
"clip_ratio/region_mean": 0.0,
|
| 1785 |
+
"completion_length": 3241.3333740234375,
|
| 1786 |
+
"entropy": 0.51806640625,
|
| 1787 |
+
"epoch": 0.10742857142857143,
|
| 1788 |
+
"grad_norm": 0.11734016239643097,
|
| 1789 |
+
"kl": 0.0025129318237304688,
|
| 1790 |
+
"learning_rate": 1.1336692317580158e-07,
|
| 1791 |
+
"loss": 0.0001,
|
| 1792 |
+
"reward": 0.5208333432674408,
|
| 1793 |
+
"reward_std": 0.3755134493112564,
|
| 1794 |
+
"rewards/accuracy_reward": 0.16666666697710752,
|
| 1795 |
+
"rewards/format_reward": 0.354166679084301,
|
| 1796 |
+
"step": 94
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"clip_ratio/high_max": 0.0,
|
| 1800 |
+
"clip_ratio/high_mean": 0.0,
|
| 1801 |
+
"clip_ratio/low_mean": 0.0,
|
| 1802 |
+
"clip_ratio/low_min": 0.0,
|
| 1803 |
+
"clip_ratio/region_mean": 0.0,
|
| 1804 |
+
"completion_length": 3691.875,
|
| 1805 |
+
"entropy": 0.4384765625,
|
| 1806 |
+
"epoch": 0.10857142857142857,
|
| 1807 |
+
"grad_norm": 0.10122071951627731,
|
| 1808 |
+
"kl": 0.0010461807250976562,
|
| 1809 |
+
"learning_rate": 1.0983357966978745e-07,
|
| 1810 |
+
"loss": 0.0,
|
| 1811 |
+
"reward": 0.2500000037252903,
|
| 1812 |
+
"reward_std": 0.39341451972723007,
|
| 1813 |
+
"rewards/accuracy_reward": 0.07291666883975267,
|
| 1814 |
+
"rewards/format_reward": 0.17708333767950535,
|
| 1815 |
+
"step": 95
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"clip_ratio/high_max": 0.0,
|
| 1819 |
+
"clip_ratio/high_mean": 0.0,
|
| 1820 |
+
"clip_ratio/low_mean": 0.0,
|
| 1821 |
+
"clip_ratio/low_min": 0.0,
|
| 1822 |
+
"clip_ratio/region_mean": 0.0,
|
| 1823 |
+
"completion_length": 2813.8751220703125,
|
| 1824 |
+
"entropy": 0.3935546875,
|
| 1825 |
+
"epoch": 0.10971428571428571,
|
| 1826 |
+
"grad_norm": 0.1897701621055603,
|
| 1827 |
+
"kl": 0.0020542144775390625,
|
| 1828 |
+
"learning_rate": 1.068365111445064e-07,
|
| 1829 |
+
"loss": 0.0001,
|
| 1830 |
+
"reward": 0.7291666939854622,
|
| 1831 |
+
"reward_std": 0.500264085829258,
|
| 1832 |
+
"rewards/accuracy_reward": 0.2708333432674408,
|
| 1833 |
+
"rewards/format_reward": 0.4583333507180214,
|
| 1834 |
+
"step": 96
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"clip_ratio/high_max": 0.0,
|
| 1838 |
+
"clip_ratio/high_mean": 0.0,
|
| 1839 |
+
"clip_ratio/low_mean": 0.0,
|
| 1840 |
+
"clip_ratio/low_min": 0.0,
|
| 1841 |
+
"clip_ratio/region_mean": 0.0,
|
| 1842 |
+
"completion_length": 3200.7188110351562,
|
| 1843 |
+
"entropy": 0.43798828125,
|
| 1844 |
+
"epoch": 0.11085714285714286,
|
| 1845 |
+
"grad_norm": 0.1710478961467743,
|
| 1846 |
+
"kl": 0.0015239715576171875,
|
| 1847 |
+
"learning_rate": 1.0437936906629334e-07,
|
| 1848 |
+
"loss": 0.0001,
|
| 1849 |
+
"reward": 0.6562500204890966,
|
| 1850 |
+
"reward_std": 0.44586458802223206,
|
| 1851 |
+
"rewards/accuracy_reward": 0.22916667722165585,
|
| 1852 |
+
"rewards/format_reward": 0.4270833386108279,
|
| 1853 |
+
"step": 97
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"clip_ratio/high_max": 0.0,
|
| 1857 |
+
"clip_ratio/high_mean": 0.0,
|
| 1858 |
+
"clip_ratio/low_mean": 0.0,
|
| 1859 |
+
"clip_ratio/low_min": 0.0,
|
| 1860 |
+
"clip_ratio/region_mean": 0.0,
|
| 1861 |
+
"completion_length": 2654.9688110351562,
|
| 1862 |
+
"entropy": 0.374267578125,
|
| 1863 |
+
"epoch": 0.112,
|
| 1864 |
+
"grad_norm": 0.16486391425132751,
|
| 1865 |
+
"kl": 0.0010390281677246094,
|
| 1866 |
+
"learning_rate": 1.0246514708427701e-07,
|
| 1867 |
+
"loss": 0.0,
|
| 1868 |
+
"reward": 0.6979166939854622,
|
| 1869 |
+
"reward_std": 0.26594046503305435,
|
| 1870 |
+
"rewards/accuracy_reward": 0.14583333395421505,
|
| 1871 |
+
"rewards/format_reward": 0.5520833507180214,
|
| 1872 |
+
"step": 98
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"clip_ratio/high_max": 0.0,
|
| 1876 |
+
"clip_ratio/high_mean": 0.0,
|
| 1877 |
+
"clip_ratio/low_mean": 0.0,
|
| 1878 |
+
"clip_ratio/low_min": 0.0,
|
| 1879 |
+
"clip_ratio/region_mean": 0.0,
|
| 1880 |
+
"completion_length": 3058.5833435058594,
|
| 1881 |
+
"entropy": 0.36572265625,
|
| 1882 |
+
"epoch": 0.11314285714285714,
|
| 1883 |
+
"grad_norm": 0.09798528254032135,
|
| 1884 |
+
"kl": 0.0014705657958984375,
|
| 1885 |
+
"learning_rate": 1.0109617738307911e-07,
|
| 1886 |
+
"loss": 0.0001,
|
| 1887 |
+
"reward": 0.5312500149011612,
|
| 1888 |
+
"reward_std": 0.20392338931560516,
|
| 1889 |
+
"rewards/accuracy_reward": 0.197916679084301,
|
| 1890 |
+
"rewards/format_reward": 0.3333333358168602,
|
| 1891 |
+
"step": 99
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"clip_ratio/high_max": 0.0,
|
| 1895 |
+
"clip_ratio/high_mean": 0.0,
|
| 1896 |
+
"clip_ratio/low_mean": 0.0,
|
| 1897 |
+
"clip_ratio/low_min": 0.0,
|
| 1898 |
+
"clip_ratio/region_mean": 0.0,
|
| 1899 |
+
"completion_length": 2583.0521240234375,
|
| 1900 |
+
"entropy": 0.34912109375,
|
| 1901 |
+
"epoch": 0.11428571428571428,
|
| 1902 |
+
"grad_norm": 0.1591912806034088,
|
| 1903 |
+
"kl": 0.0018825531005859375,
|
| 1904 |
+
"learning_rate": 1.002741278414069e-07,
|
| 1905 |
+
"loss": 0.0001,
|
| 1906 |
+
"reward": 0.9375,
|
| 1907 |
+
"reward_std": 0.5781587325036526,
|
| 1908 |
+
"rewards/accuracy_reward": 0.3333333432674408,
|
| 1909 |
+
"rewards/format_reward": 0.6041666865348816,
|
| 1910 |
+
"step": 100
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 0.11428571428571428,
|
| 1914 |
+
"step": 100,
|
| 1915 |
+
"total_flos": 0.0,
|
| 1916 |
+
"train_loss": 4.974420491237197e-05,
|
| 1917 |
+
"train_runtime": 8330.7391,
|
| 1918 |
+
"train_samples_per_second": 1.152,
|
| 1919 |
+
"train_steps_per_second": 0.012
|
| 1920 |
+
}
|
| 1921 |
+
],
|
| 1922 |
+
"logging_steps": 1,
|
| 1923 |
+
"max_steps": 100,
|
| 1924 |
+
"num_input_tokens_seen": 0,
|
| 1925 |
+
"num_train_epochs": 1,
|
| 1926 |
+
"save_steps": 50,
|
| 1927 |
+
"stateful_callbacks": {
|
| 1928 |
+
"TrainerControl": {
|
| 1929 |
+
"args": {
|
| 1930 |
+
"should_epoch_stop": false,
|
| 1931 |
+
"should_evaluate": false,
|
| 1932 |
+
"should_log": false,
|
| 1933 |
+
"should_save": true,
|
| 1934 |
+
"should_training_stop": true
|
| 1935 |
+
},
|
| 1936 |
+
"attributes": {}
|
| 1937 |
+
}
|
| 1938 |
+
},
|
| 1939 |
+
"total_flos": 0.0,
|
| 1940 |
+
"train_batch_size": 6,
|
| 1941 |
+
"trial_name": null,
|
| 1942 |
+
"trial_params": null
|
| 1943 |
+
}
|