Model save
Browse files- README.md +67 -0
- all_results.json +8 -0
- generation_config.json +9 -0
- train_results.json +8 -0
- trainer_state.json +1493 -0
README.md
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
model_name: ER-GRPO-alpha30
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
- trl
|
| 7 |
+
- grpo
|
| 8 |
+
licence: license
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# Model Card for ER-GRPO-alpha30
|
| 12 |
+
|
| 13 |
+
This model is a fine-tuned version of [None](https://huggingface.co/None).
|
| 14 |
+
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 15 |
+
|
| 16 |
+
## Quick start
|
| 17 |
+
|
| 18 |
+
```python
|
| 19 |
+
from transformers import pipeline
|
| 20 |
+
|
| 21 |
+
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?"
|
| 22 |
+
generator = pipeline("text-generation", model="lalalaDa/ER-GRPO-alpha30", device="cuda")
|
| 23 |
+
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
|
| 24 |
+
print(output["generated_text"])
|
| 25 |
+
```
|
| 26 |
+
|
| 27 |
+
## Training procedure
|
| 28 |
+
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
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).
|
| 33 |
+
|
| 34 |
+
### Framework versions
|
| 35 |
+
|
| 36 |
+
- TRL: 0.18.1
|
| 37 |
+
- Transformers: 4.52.4
|
| 38 |
+
- Pytorch: 2.5.1
|
| 39 |
+
- Datasets: 3.6.0
|
| 40 |
+
- Tokenizers: 0.21.1
|
| 41 |
+
|
| 42 |
+
## Citations
|
| 43 |
+
|
| 44 |
+
Cite GRPO as:
|
| 45 |
+
|
| 46 |
+
```bibtex
|
| 47 |
+
@article{zhihong2024deepseekmath,
|
| 48 |
+
title = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
|
| 49 |
+
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},
|
| 50 |
+
year = 2024,
|
| 51 |
+
eprint = {arXiv:2402.03300},
|
| 52 |
+
}
|
| 53 |
+
|
| 54 |
+
```
|
| 55 |
+
|
| 56 |
+
Cite TRL as:
|
| 57 |
+
|
| 58 |
+
```bibtex
|
| 59 |
+
@misc{vonwerra2022trl,
|
| 60 |
+
title = {{TRL: Transformer Reinforcement Learning}},
|
| 61 |
+
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 Gallou{\'e}dec},
|
| 62 |
+
year = 2020,
|
| 63 |
+
journal = {GitHub repository},
|
| 64 |
+
publisher = {GitHub},
|
| 65 |
+
howpublished = {\url{https://github.com/huggingface/trl}}
|
| 66 |
+
}
|
| 67 |
+
```
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 0.0,
|
| 3 |
+
"train_loss": -0.45832799572497607,
|
| 4 |
+
"train_runtime": 4441.2019,
|
| 5 |
+
"train_samples": 7000,
|
| 6 |
+
"train_samples_per_second": 0.54,
|
| 7 |
+
"train_steps_per_second": 0.011
|
| 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": "4.52.4"
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 0.0,
|
| 3 |
+
"train_loss": -0.45832799572497607,
|
| 4 |
+
"train_runtime": 4441.2019,
|
| 5 |
+
"train_samples": 7000,
|
| 6 |
+
"train_samples_per_second": 0.54,
|
| 7 |
+
"train_steps_per_second": 0.011
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1493 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.05714285714285714,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 50,
|
| 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 |
+
"completions/clipped_ratio": 0.5208333333333333,
|
| 19 |
+
"completions/max_length": 3584.0,
|
| 20 |
+
"completions/max_terminated_length": 3128.0,
|
| 21 |
+
"completions/mean_length": 2584.104248046875,
|
| 22 |
+
"completions/mean_terminated_length": 1497.2608642578125,
|
| 23 |
+
"completions/min_length": 557.0,
|
| 24 |
+
"completions/min_terminated_length": 557.0,
|
| 25 |
+
"epoch": 0.001142857142857143,
|
| 26 |
+
"frac_reward_zero_std": 0.0,
|
| 27 |
+
"grad_norm": 1.6251945495605469,
|
| 28 |
+
"kl": 0.0,
|
| 29 |
+
"learning_rate": 0.0,
|
| 30 |
+
"loss": -0.6946,
|
| 31 |
+
"num_tokens": 131153.0,
|
| 32 |
+
"policy_entropy_avg": 8.125,
|
| 33 |
+
"reward": 0.2579294741153717,
|
| 34 |
+
"reward_std": 0.505131185054779,
|
| 35 |
+
"rewards/cosine_scaled_reward/mean": -0.062009382992982864,
|
| 36 |
+
"rewards/cosine_scaled_reward/std": 0.43048128485679626,
|
| 37 |
+
"rewards/format_reward/mean": 0.5208333134651184,
|
| 38 |
+
"rewards/format_reward/std": 0.504852294921875,
|
| 39 |
+
"step": 1
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"clip_ratio/high_max": 0.0,
|
| 43 |
+
"clip_ratio/high_mean": 0.0,
|
| 44 |
+
"clip_ratio/low_mean": 0.0,
|
| 45 |
+
"clip_ratio/low_min": 0.0,
|
| 46 |
+
"clip_ratio/region_mean": 0.0,
|
| 47 |
+
"completions/clipped_ratio": 0.5833333333333333,
|
| 48 |
+
"completions/max_length": 3584.0,
|
| 49 |
+
"completions/max_terminated_length": 3280.0,
|
| 50 |
+
"completions/mean_length": 2761.666748046875,
|
| 51 |
+
"completions/mean_terminated_length": 1610.4000244140625,
|
| 52 |
+
"completions/min_length": 465.0,
|
| 53 |
+
"completions/min_terminated_length": 465.0,
|
| 54 |
+
"epoch": 0.002285714285714286,
|
| 55 |
+
"frac_reward_zero_std": 0.0,
|
| 56 |
+
"grad_norm": 1.6243528127670288,
|
| 57 |
+
"kl": 0.0,
|
| 58 |
+
"learning_rate": 2e-07,
|
| 59 |
+
"loss": -1.3501,
|
| 60 |
+
"num_tokens": 271243.0,
|
| 61 |
+
"policy_entropy_avg": 8.125,
|
| 62 |
+
"reward": 0.26637595891952515,
|
| 63 |
+
"reward_std": 0.5504351854324341,
|
| 64 |
+
"rewards/cosine_scaled_reward/mean": -0.003428752301260829,
|
| 65 |
+
"rewards/cosine_scaled_reward/std": 0.4935320317745209,
|
| 66 |
+
"rewards/format_reward/mean": 0.4166666567325592,
|
| 67 |
+
"rewards/format_reward/std": 0.49822381138801575,
|
| 68 |
+
"step": 2
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"clip_ratio/high_max": 0.0,
|
| 72 |
+
"clip_ratio/high_mean": 0.0,
|
| 73 |
+
"clip_ratio/low_mean": 0.0,
|
| 74 |
+
"clip_ratio/low_min": 0.0,
|
| 75 |
+
"clip_ratio/region_mean": 0.0,
|
| 76 |
+
"completions/clipped_ratio": 0.9166666666666666,
|
| 77 |
+
"completions/max_length": 3584.0,
|
| 78 |
+
"completions/max_terminated_length": 2914.0,
|
| 79 |
+
"completions/mean_length": 3405.541748046875,
|
| 80 |
+
"completions/mean_terminated_length": 1442.5,
|
| 81 |
+
"completions/min_length": 490.0,
|
| 82 |
+
"completions/min_terminated_length": 490.0,
|
| 83 |
+
"epoch": 0.0034285714285714284,
|
| 84 |
+
"frac_reward_zero_std": 0.0,
|
| 85 |
+
"grad_norm": 0.6798118352890015,
|
| 86 |
+
"kl": 0.0006580352783203125,
|
| 87 |
+
"learning_rate": 4e-07,
|
| 88 |
+
"loss": 1.3636,
|
| 89 |
+
"num_tokens": 442563.0,
|
| 90 |
+
"policy_entropy_avg": 8.125,
|
| 91 |
+
"reward": -0.14459262788295746,
|
| 92 |
+
"reward_std": 0.37038296461105347,
|
| 93 |
+
"rewards/cosine_scaled_reward/mean": -0.16330842673778534,
|
| 94 |
+
"rewards/cosine_scaled_reward/std": 0.2756437659263611,
|
| 95 |
+
"rewards/format_reward/mean": 0.1041666641831398,
|
| 96 |
+
"rewards/format_reward/std": 0.3087092638015747,
|
| 97 |
+
"step": 3
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"clip_ratio/high_max": 0.0,
|
| 101 |
+
"clip_ratio/high_mean": 0.0,
|
| 102 |
+
"clip_ratio/low_mean": 0.0,
|
| 103 |
+
"clip_ratio/low_min": 0.0,
|
| 104 |
+
"clip_ratio/region_mean": 0.0,
|
| 105 |
+
"completions/clipped_ratio": 0.45833333333333337,
|
| 106 |
+
"completions/max_length": 3584.0,
|
| 107 |
+
"completions/max_terminated_length": 3581.0,
|
| 108 |
+
"completions/mean_length": 2397.45849609375,
|
| 109 |
+
"completions/mean_terminated_length": 1393.4615478515625,
|
| 110 |
+
"completions/min_length": 416.0,
|
| 111 |
+
"completions/min_terminated_length": 416.0,
|
| 112 |
+
"epoch": 0.004571428571428572,
|
| 113 |
+
"frac_reward_zero_std": 0.0,
|
| 114 |
+
"grad_norm": 1.0801085233688354,
|
| 115 |
+
"kl": 0.0006548563639322916,
|
| 116 |
+
"learning_rate": 6e-07,
|
| 117 |
+
"loss": -0.4683,
|
| 118 |
+
"num_tokens": 564997.0,
|
| 119 |
+
"policy_entropy_avg": 8.125,
|
| 120 |
+
"reward": 0.4537672996520996,
|
| 121 |
+
"reward_std": 0.6287642121315002,
|
| 122 |
+
"rewards/cosine_scaled_reward/mean": 0.0053017884492874146,
|
| 123 |
+
"rewards/cosine_scaled_reward/std": 0.44670990109443665,
|
| 124 |
+
"rewards/format_reward/mean": 0.6875,
|
| 125 |
+
"rewards/format_reward/std": 0.4684174358844757,
|
| 126 |
+
"step": 4
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"clip_ratio/high_max": 0.0,
|
| 130 |
+
"clip_ratio/high_mean": 0.0,
|
| 131 |
+
"clip_ratio/low_mean": 0.0,
|
| 132 |
+
"clip_ratio/low_min": 0.0,
|
| 133 |
+
"clip_ratio/region_mean": 0.0,
|
| 134 |
+
"completions/clipped_ratio": 0.7291666666666667,
|
| 135 |
+
"completions/max_length": 3584.0,
|
| 136 |
+
"completions/max_terminated_length": 3368.0,
|
| 137 |
+
"completions/mean_length": 3252.5,
|
| 138 |
+
"completions/mean_terminated_length": 2360.0,
|
| 139 |
+
"completions/min_length": 598.0,
|
| 140 |
+
"completions/min_terminated_length": 598.0,
|
| 141 |
+
"epoch": 0.005714285714285714,
|
| 142 |
+
"frac_reward_zero_std": 0.0,
|
| 143 |
+
"grad_norm": 0.6357815265655518,
|
| 144 |
+
"kl": 0.0006510416666666666,
|
| 145 |
+
"learning_rate": 8e-07,
|
| 146 |
+
"loss": -0.2046,
|
| 147 |
+
"num_tokens": 729229.0,
|
| 148 |
+
"policy_entropy_avg": 8.125,
|
| 149 |
+
"reward": 0.025602590292692184,
|
| 150 |
+
"reward_std": 0.46193015575408936,
|
| 151 |
+
"rewards/cosine_scaled_reward/mean": -0.15738904476165771,
|
| 152 |
+
"rewards/cosine_scaled_reward/std": 0.38784292340278625,
|
| 153 |
+
"rewards/format_reward/mean": 0.3541666567325592,
|
| 154 |
+
"rewards/format_reward/std": 0.48332110047340393,
|
| 155 |
+
"step": 5
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"clip_ratio/high_max": 0.0,
|
| 159 |
+
"clip_ratio/high_mean": 0.0,
|
| 160 |
+
"clip_ratio/low_mean": 0.0,
|
| 161 |
+
"clip_ratio/low_min": 0.0,
|
| 162 |
+
"clip_ratio/region_mean": 0.0,
|
| 163 |
+
"completions/clipped_ratio": 0.7916666666666666,
|
| 164 |
+
"completions/max_length": 3584.0,
|
| 165 |
+
"completions/max_terminated_length": 3435.0,
|
| 166 |
+
"completions/mean_length": 3184.45849609375,
|
| 167 |
+
"completions/mean_terminated_length": 1666.2000732421875,
|
| 168 |
+
"completions/min_length": 600.0,
|
| 169 |
+
"completions/min_terminated_length": 600.0,
|
| 170 |
+
"epoch": 0.006857142857142857,
|
| 171 |
+
"frac_reward_zero_std": 0.0,
|
| 172 |
+
"grad_norm": 0.611210286617279,
|
| 173 |
+
"kl": 0.0006860097249348959,
|
| 174 |
+
"learning_rate": 1e-06,
|
| 175 |
+
"loss": 0.6384,
|
| 176 |
+
"num_tokens": 890819.0,
|
| 177 |
+
"policy_entropy_avg": 8.125,
|
| 178 |
+
"reward": 0.006859039422124624,
|
| 179 |
+
"reward_std": 0.45094144344329834,
|
| 180 |
+
"rewards/cosine_scaled_reward/mean": -0.14055712521076202,
|
| 181 |
+
"rewards/cosine_scaled_reward/std": 0.32825249433517456,
|
| 182 |
+
"rewards/format_reward/mean": 0.2916666567325592,
|
| 183 |
+
"rewards/format_reward/std": 0.4593396484851837,
|
| 184 |
+
"step": 6
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"clip_ratio/high_max": 0.0,
|
| 188 |
+
"clip_ratio/high_mean": 0.0,
|
| 189 |
+
"clip_ratio/low_mean": 0.0,
|
| 190 |
+
"clip_ratio/low_min": 0.0,
|
| 191 |
+
"clip_ratio/region_mean": 0.0,
|
| 192 |
+
"completions/clipped_ratio": 0.6041666666666667,
|
| 193 |
+
"completions/max_length": 3584.0,
|
| 194 |
+
"completions/max_terminated_length": 3543.0,
|
| 195 |
+
"completions/mean_length": 3090.0,
|
| 196 |
+
"completions/mean_terminated_length": 2336.0,
|
| 197 |
+
"completions/min_length": 952.0,
|
| 198 |
+
"completions/min_terminated_length": 952.0,
|
| 199 |
+
"epoch": 0.008,
|
| 200 |
+
"frac_reward_zero_std": 0.0,
|
| 201 |
+
"grad_norm": 0.6517191529273987,
|
| 202 |
+
"kl": 0.0005308787027994791,
|
| 203 |
+
"learning_rate": 9.989038226169207e-07,
|
| 204 |
+
"loss": -0.3634,
|
| 205 |
+
"num_tokens": 1046945.0,
|
| 206 |
+
"policy_entropy_avg": 8.125,
|
| 207 |
+
"reward": 0.3611307144165039,
|
| 208 |
+
"reward_std": 0.6935849189758301,
|
| 209 |
+
"rewards/cosine_scaled_reward/mean": 0.017376163974404335,
|
| 210 |
+
"rewards/cosine_scaled_reward/std": 0.5143836736679077,
|
| 211 |
+
"rewards/format_reward/mean": 0.5208333134651184,
|
| 212 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 213 |
+
"step": 7
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"clip_ratio/high_max": 0.0,
|
| 217 |
+
"clip_ratio/high_mean": 0.0,
|
| 218 |
+
"clip_ratio/low_mean": 0.0,
|
| 219 |
+
"clip_ratio/low_min": 0.0,
|
| 220 |
+
"clip_ratio/region_mean": 0.0,
|
| 221 |
+
"completions/clipped_ratio": 0.5625,
|
| 222 |
+
"completions/max_length": 3584.0,
|
| 223 |
+
"completions/max_terminated_length": 3421.0,
|
| 224 |
+
"completions/mean_length": 2711.27099609375,
|
| 225 |
+
"completions/mean_terminated_length": 1589.1905517578125,
|
| 226 |
+
"completions/min_length": 375.0,
|
| 227 |
+
"completions/min_terminated_length": 375.0,
|
| 228 |
+
"epoch": 0.009142857142857144,
|
| 229 |
+
"frac_reward_zero_std": 0.0,
|
| 230 |
+
"grad_norm": 1.013113021850586,
|
| 231 |
+
"kl": 0.0005372365315755209,
|
| 232 |
+
"learning_rate": 9.956206309337066e-07,
|
| 233 |
+
"loss": -1.336,
|
| 234 |
+
"num_tokens": 1184898.0,
|
| 235 |
+
"policy_entropy_avg": 8.125,
|
| 236 |
+
"reward": 0.5312491655349731,
|
| 237 |
+
"reward_std": 0.5121049880981445,
|
| 238 |
+
"rewards/cosine_scaled_reward/mean": 0.16906984150409698,
|
| 239 |
+
"rewards/cosine_scaled_reward/std": 0.48645660281181335,
|
| 240 |
+
"rewards/format_reward/mean": 0.4791666567325592,
|
| 241 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 242 |
+
"step": 8
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"clip_ratio/high_max": 0.0,
|
| 246 |
+
"clip_ratio/high_mean": 0.0,
|
| 247 |
+
"clip_ratio/low_mean": 0.0,
|
| 248 |
+
"clip_ratio/low_min": 0.0,
|
| 249 |
+
"clip_ratio/region_mean": 0.0,
|
| 250 |
+
"completions/clipped_ratio": 0.625,
|
| 251 |
+
"completions/max_length": 3584.0,
|
| 252 |
+
"completions/max_terminated_length": 3575.0,
|
| 253 |
+
"completions/mean_length": 2904.875,
|
| 254 |
+
"completions/mean_terminated_length": 1773.0,
|
| 255 |
+
"completions/min_length": 553.0,
|
| 256 |
+
"completions/min_terminated_length": 553.0,
|
| 257 |
+
"epoch": 0.010285714285714285,
|
| 258 |
+
"frac_reward_zero_std": 0.0,
|
| 259 |
+
"grad_norm": 0.7274989485740662,
|
| 260 |
+
"kl": 0.00063323974609375,
|
| 261 |
+
"learning_rate": 9.901664203302124e-07,
|
| 262 |
+
"loss": 0.2595,
|
| 263 |
+
"num_tokens": 1332924.0,
|
| 264 |
+
"policy_entropy_avg": 8.125,
|
| 265 |
+
"reward": 0.15100964903831482,
|
| 266 |
+
"reward_std": 0.5337426066398621,
|
| 267 |
+
"rewards/cosine_scaled_reward/mean": -0.09217208623886108,
|
| 268 |
+
"rewards/cosine_scaled_reward/std": 0.36159586906433105,
|
| 269 |
+
"rewards/format_reward/mean": 0.4166666567325592,
|
| 270 |
+
"rewards/format_reward/std": 0.49822381138801575,
|
| 271 |
+
"step": 9
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"clip_ratio/high_max": 0.0,
|
| 275 |
+
"clip_ratio/high_mean": 0.0,
|
| 276 |
+
"clip_ratio/low_mean": 0.0,
|
| 277 |
+
"clip_ratio/low_min": 0.0,
|
| 278 |
+
"clip_ratio/region_mean": 0.0,
|
| 279 |
+
"completions/clipped_ratio": 0.6875,
|
| 280 |
+
"completions/max_length": 3584.0,
|
| 281 |
+
"completions/max_terminated_length": 3407.0,
|
| 282 |
+
"completions/mean_length": 2811.229248046875,
|
| 283 |
+
"completions/mean_terminated_length": 1111.1334228515625,
|
| 284 |
+
"completions/min_length": 407.0,
|
| 285 |
+
"completions/min_terminated_length": 407.0,
|
| 286 |
+
"epoch": 0.011428571428571429,
|
| 287 |
+
"frac_reward_zero_std": 0.0,
|
| 288 |
+
"grad_norm": 0.9453772306442261,
|
| 289 |
+
"kl": 0.0005699793497721354,
|
| 290 |
+
"learning_rate": 9.825677631722435e-07,
|
| 291 |
+
"loss": 0.1,
|
| 292 |
+
"num_tokens": 1475987.0,
|
| 293 |
+
"policy_entropy_avg": 8.125,
|
| 294 |
+
"reward": 0.001259398995898664,
|
| 295 |
+
"reward_std": 0.425419420003891,
|
| 296 |
+
"rewards/cosine_scaled_reward/mean": -0.17611455917358398,
|
| 297 |
+
"rewards/cosine_scaled_reward/std": 0.3572066128253937,
|
| 298 |
+
"rewards/format_reward/mean": 0.3541666567325592,
|
| 299 |
+
"rewards/format_reward/std": 0.48332110047340393,
|
| 300 |
+
"step": 10
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"clip_ratio/high_max": 0.0,
|
| 304 |
+
"clip_ratio/high_mean": 0.0,
|
| 305 |
+
"clip_ratio/low_mean": 0.0,
|
| 306 |
+
"clip_ratio/low_min": 0.0,
|
| 307 |
+
"clip_ratio/region_mean": 0.0,
|
| 308 |
+
"completions/clipped_ratio": 0.8958333333333334,
|
| 309 |
+
"completions/max_length": 3584.0,
|
| 310 |
+
"completions/max_terminated_length": 3246.0,
|
| 311 |
+
"completions/mean_length": 3474.0,
|
| 312 |
+
"completions/mean_terminated_length": 2528.0,
|
| 313 |
+
"completions/min_length": 1503.0,
|
| 314 |
+
"completions/min_terminated_length": 1503.0,
|
| 315 |
+
"epoch": 0.012571428571428572,
|
| 316 |
+
"frac_reward_zero_std": 0.0,
|
| 317 |
+
"grad_norm": 0.8819871544837952,
|
| 318 |
+
"kl": 0.0006033579508463541,
|
| 319 |
+
"learning_rate": 9.728616793536587e-07,
|
| 320 |
+
"loss": 1.6889,
|
| 321 |
+
"num_tokens": 1651493.0,
|
| 322 |
+
"policy_entropy_avg": 8.125,
|
| 323 |
+
"reward": -0.1606103479862213,
|
| 324 |
+
"reward_std": 0.441455602645874,
|
| 325 |
+
"rewards/cosine_scaled_reward/mean": -0.19646306335926056,
|
| 326 |
+
"rewards/cosine_scaled_reward/std": 0.34247782826423645,
|
| 327 |
+
"rewards/format_reward/mean": 0.1458333283662796,
|
| 328 |
+
"rewards/format_reward/std": 0.3566739559173584,
|
| 329 |
+
"step": 11
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"clip_ratio/high_max": 0.0,
|
| 333 |
+
"clip_ratio/high_mean": 0.0,
|
| 334 |
+
"clip_ratio/low_mean": 0.0,
|
| 335 |
+
"clip_ratio/low_min": 0.0,
|
| 336 |
+
"clip_ratio/region_mean": 0.0,
|
| 337 |
+
"completions/clipped_ratio": 0.45833333333333337,
|
| 338 |
+
"completions/max_length": 3584.0,
|
| 339 |
+
"completions/max_terminated_length": 3479.0,
|
| 340 |
+
"completions/mean_length": 2459.3125,
|
| 341 |
+
"completions/mean_terminated_length": 1507.6539306640625,
|
| 342 |
+
"completions/min_length": 698.0,
|
| 343 |
+
"completions/min_terminated_length": 698.0,
|
| 344 |
+
"epoch": 0.013714285714285714,
|
| 345 |
+
"frac_reward_zero_std": 0.0,
|
| 346 |
+
"grad_norm": 0.9987263083457947,
|
| 347 |
+
"kl": 0.0006459554036458334,
|
| 348 |
+
"learning_rate": 9.610954559391704e-07,
|
| 349 |
+
"loss": -0.7409,
|
| 350 |
+
"num_tokens": 1777748.0,
|
| 351 |
+
"policy_entropy_avg": 8.125,
|
| 352 |
+
"reward": 0.1868935525417328,
|
| 353 |
+
"reward_std": 0.5621633529663086,
|
| 354 |
+
"rewards/cosine_scaled_reward/mean": -0.13748572766780853,
|
| 355 |
+
"rewards/cosine_scaled_reward/std": 0.37971118092536926,
|
| 356 |
+
"rewards/format_reward/mean": 0.5625,
|
| 357 |
+
"rewards/format_reward/std": 0.5013279914855957,
|
| 358 |
+
"step": 12
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"clip_ratio/high_max": 0.0,
|
| 362 |
+
"clip_ratio/high_mean": 0.0,
|
| 363 |
+
"clip_ratio/low_mean": 0.0,
|
| 364 |
+
"clip_ratio/low_min": 0.0,
|
| 365 |
+
"clip_ratio/region_mean": 0.0,
|
| 366 |
+
"completions/clipped_ratio": 0.5625,
|
| 367 |
+
"completions/max_length": 3584.0,
|
| 368 |
+
"completions/max_terminated_length": 3377.0,
|
| 369 |
+
"completions/mean_length": 2769.729248046875,
|
| 370 |
+
"completions/mean_terminated_length": 1722.8095703125,
|
| 371 |
+
"completions/min_length": 706.0,
|
| 372 |
+
"completions/min_terminated_length": 706.0,
|
| 373 |
+
"epoch": 0.014857142857142857,
|
| 374 |
+
"frac_reward_zero_std": 0.0,
|
| 375 |
+
"grad_norm": 0.6482232213020325,
|
| 376 |
+
"kl": 0.0006001790364583334,
|
| 377 |
+
"learning_rate": 9.473264167865171e-07,
|
| 378 |
+
"loss": -0.0518,
|
| 379 |
+
"num_tokens": 1918975.0,
|
| 380 |
+
"policy_entropy_avg": 8.125,
|
| 381 |
+
"reward": 0.15883508324623108,
|
| 382 |
+
"reward_std": 0.4439440965652466,
|
| 383 |
+
"rewards/cosine_scaled_reward/mean": -0.09656915813684464,
|
| 384 |
+
"rewards/cosine_scaled_reward/std": 0.38122376799583435,
|
| 385 |
+
"rewards/format_reward/mean": 0.4375,
|
| 386 |
+
"rewards/format_reward/std": 0.5013279914855957,
|
| 387 |
+
"step": 13
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"clip_ratio/high_max": 0.0,
|
| 391 |
+
"clip_ratio/high_mean": 0.0,
|
| 392 |
+
"clip_ratio/low_mean": 0.0,
|
| 393 |
+
"clip_ratio/low_min": 0.0,
|
| 394 |
+
"clip_ratio/region_mean": 0.0,
|
| 395 |
+
"completions/clipped_ratio": 0.5208333333333333,
|
| 396 |
+
"completions/max_length": 3584.0,
|
| 397 |
+
"completions/max_terminated_length": 3364.0,
|
| 398 |
+
"completions/mean_length": 2708.541748046875,
|
| 399 |
+
"completions/mean_terminated_length": 1756.95654296875,
|
| 400 |
+
"completions/min_length": 661.0,
|
| 401 |
+
"completions/min_terminated_length": 661.0,
|
| 402 |
+
"epoch": 0.016,
|
| 403 |
+
"frac_reward_zero_std": 0.0,
|
| 404 |
+
"grad_norm": 1.1001851558685303,
|
| 405 |
+
"kl": 0.0006243387858072916,
|
| 406 |
+
"learning_rate": 9.316216432703916e-07,
|
| 407 |
+
"loss": -0.6409,
|
| 408 |
+
"num_tokens": 2056941.0,
|
| 409 |
+
"policy_entropy_avg": 8.125,
|
| 410 |
+
"reward": 0.2933271527290344,
|
| 411 |
+
"reward_std": 0.5121263861656189,
|
| 412 |
+
"rewards/cosine_scaled_reward/mean": -0.013947081752121449,
|
| 413 |
+
"rewards/cosine_scaled_reward/std": 0.49587875604629517,
|
| 414 |
+
"rewards/format_reward/mean": 0.4791666567325592,
|
| 415 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 416 |
+
"step": 14
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"clip_ratio/high_max": 0.0,
|
| 420 |
+
"clip_ratio/high_mean": 0.0,
|
| 421 |
+
"clip_ratio/low_mean": 0.0,
|
| 422 |
+
"clip_ratio/low_min": 0.0,
|
| 423 |
+
"clip_ratio/region_mean": 0.0,
|
| 424 |
+
"completions/clipped_ratio": 0.5833333333333333,
|
| 425 |
+
"completions/max_length": 3584.0,
|
| 426 |
+
"completions/max_terminated_length": 3430.0,
|
| 427 |
+
"completions/mean_length": 2640.125,
|
| 428 |
+
"completions/mean_terminated_length": 1318.7000732421875,
|
| 429 |
+
"completions/min_length": 425.0,
|
| 430 |
+
"completions/min_terminated_length": 425.0,
|
| 431 |
+
"epoch": 0.017142857142857144,
|
| 432 |
+
"frac_reward_zero_std": 0.0,
|
| 433 |
+
"grad_norm": 1.0151442289352417,
|
| 434 |
+
"kl": 0.0005944569905598959,
|
| 435 |
+
"learning_rate": 9.140576474687263e-07,
|
| 436 |
+
"loss": -0.6683,
|
| 437 |
+
"num_tokens": 2191857.0,
|
| 438 |
+
"policy_entropy_avg": 8.125,
|
| 439 |
+
"reward": 0.2889822721481323,
|
| 440 |
+
"reward_std": 0.31458932161331177,
|
| 441 |
+
"rewards/cosine_scaled_reward/mean": -0.038122642785310745,
|
| 442 |
+
"rewards/cosine_scaled_reward/std": 0.44473376870155334,
|
| 443 |
+
"rewards/format_reward/mean": 0.5208333134651184,
|
| 444 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 445 |
+
"step": 15
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"clip_ratio/high_max": 0.0,
|
| 449 |
+
"clip_ratio/high_mean": 0.0,
|
| 450 |
+
"clip_ratio/low_mean": 0.0,
|
| 451 |
+
"clip_ratio/low_min": 0.0,
|
| 452 |
+
"clip_ratio/region_mean": 0.0,
|
| 453 |
+
"completions/clipped_ratio": 0.9583333333333334,
|
| 454 |
+
"completions/max_length": 3584.0,
|
| 455 |
+
"completions/max_terminated_length": 3304.0,
|
| 456 |
+
"completions/mean_length": 3522.1875,
|
| 457 |
+
"completions/mean_terminated_length": 2100.5,
|
| 458 |
+
"completions/min_length": 897.0,
|
| 459 |
+
"completions/min_terminated_length": 897.0,
|
| 460 |
+
"epoch": 0.018285714285714287,
|
| 461 |
+
"frac_reward_zero_std": 0.0,
|
| 462 |
+
"grad_norm": 0.9443197250366211,
|
| 463 |
+
"kl": 0.0007228851318359375,
|
| 464 |
+
"learning_rate": 8.9471999940354e-07,
|
| 465 |
+
"loss": 2.2731,
|
| 466 |
+
"num_tokens": 2368752.0,
|
| 467 |
+
"policy_entropy_avg": 8.125,
|
| 468 |
+
"reward": -0.21434035897254944,
|
| 469 |
+
"reward_std": 0.31254154443740845,
|
| 470 |
+
"rewards/cosine_scaled_reward/mean": -0.20654386281967163,
|
| 471 |
+
"rewards/cosine_scaled_reward/std": 0.2568126320838928,
|
| 472 |
+
"rewards/format_reward/mean": 0.0833333358168602,
|
| 473 |
+
"rewards/format_reward/std": 0.2793101966381073,
|
| 474 |
+
"step": 16
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"clip_ratio/high_max": 0.0,
|
| 478 |
+
"clip_ratio/high_mean": 0.0,
|
| 479 |
+
"clip_ratio/low_mean": 0.0,
|
| 480 |
+
"clip_ratio/low_min": 0.0,
|
| 481 |
+
"clip_ratio/region_mean": 0.0,
|
| 482 |
+
"completions/clipped_ratio": 0.45833333333333337,
|
| 483 |
+
"completions/max_length": 3584.0,
|
| 484 |
+
"completions/max_terminated_length": 3076.0,
|
| 485 |
+
"completions/mean_length": 2299.875,
|
| 486 |
+
"completions/mean_terminated_length": 1213.3077392578125,
|
| 487 |
+
"completions/min_length": 364.0,
|
| 488 |
+
"completions/min_terminated_length": 364.0,
|
| 489 |
+
"epoch": 0.019428571428571427,
|
| 490 |
+
"frac_reward_zero_std": 0.0,
|
| 491 |
+
"grad_norm": 0.764594554901123,
|
| 492 |
+
"kl": 0.000736236572265625,
|
| 493 |
+
"learning_rate": 8.737029101523929e-07,
|
| 494 |
+
"loss": -0.0316,
|
| 495 |
+
"num_tokens": 2487036.0,
|
| 496 |
+
"policy_entropy_avg": 8.125,
|
| 497 |
+
"reward": 0.3496870696544647,
|
| 498 |
+
"reward_std": 0.5126497745513916,
|
| 499 |
+
"rewards/cosine_scaled_reward/mean": -0.0018433034420013428,
|
| 500 |
+
"rewards/cosine_scaled_reward/std": 0.4425233006477356,
|
| 501 |
+
"rewards/format_reward/mean": 0.5416666865348816,
|
| 502 |
+
"rewards/format_reward/std": 0.5035336017608643,
|
| 503 |
+
"step": 17
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"clip_ratio/high_max": 0.0,
|
| 507 |
+
"clip_ratio/high_mean": 0.0,
|
| 508 |
+
"clip_ratio/low_mean": 0.0,
|
| 509 |
+
"clip_ratio/low_min": 0.0,
|
| 510 |
+
"clip_ratio/region_mean": 0.0,
|
| 511 |
+
"completions/clipped_ratio": 0.5416666666666667,
|
| 512 |
+
"completions/max_length": 3584.0,
|
| 513 |
+
"completions/max_terminated_length": 3228.0,
|
| 514 |
+
"completions/mean_length": 2700.3125,
|
| 515 |
+
"completions/mean_terminated_length": 1655.95458984375,
|
| 516 |
+
"completions/min_length": 536.0,
|
| 517 |
+
"completions/min_terminated_length": 536.0,
|
| 518 |
+
"epoch": 0.02057142857142857,
|
| 519 |
+
"frac_reward_zero_std": 0.0,
|
| 520 |
+
"grad_norm": 0.8999841213226318,
|
| 521 |
+
"kl": 0.0006821950276692709,
|
| 522 |
+
"learning_rate": 8.511087728614862e-07,
|
| 523 |
+
"loss": -0.6147,
|
| 524 |
+
"num_tokens": 2624433.0,
|
| 525 |
+
"policy_entropy_avg": 8.125,
|
| 526 |
+
"reward": 0.26751697063446045,
|
| 527 |
+
"reward_std": 0.4905094802379608,
|
| 528 |
+
"rewards/cosine_scaled_reward/mean": -0.05463438108563423,
|
| 529 |
+
"rewards/cosine_scaled_reward/std": 0.447592556476593,
|
| 530 |
+
"rewards/format_reward/mean": 0.5208333134651184,
|
| 531 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 532 |
+
"step": 18
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"clip_ratio/high_max": 0.0,
|
| 536 |
+
"clip_ratio/high_mean": 0.0,
|
| 537 |
+
"clip_ratio/low_mean": 0.0,
|
| 538 |
+
"clip_ratio/low_min": 0.0,
|
| 539 |
+
"clip_ratio/region_mean": 0.0,
|
| 540 |
+
"completions/clipped_ratio": 0.6666666666666667,
|
| 541 |
+
"completions/max_length": 3584.0,
|
| 542 |
+
"completions/max_terminated_length": 3405.0,
|
| 543 |
+
"completions/mean_length": 2786.77099609375,
|
| 544 |
+
"completions/mean_terminated_length": 1192.3125,
|
| 545 |
+
"completions/min_length": 397.0,
|
| 546 |
+
"completions/min_terminated_length": 397.0,
|
| 547 |
+
"epoch": 0.021714285714285714,
|
| 548 |
+
"frac_reward_zero_std": 0.0,
|
| 549 |
+
"grad_norm": 1.0238417387008667,
|
| 550 |
+
"kl": 0.0006643931070963541,
|
| 551 |
+
"learning_rate": 8.270476638965461e-07,
|
| 552 |
+
"loss": -0.0253,
|
| 553 |
+
"num_tokens": 2766640.0,
|
| 554 |
+
"policy_entropy_avg": 8.125,
|
| 555 |
+
"reward": 0.3180326819419861,
|
| 556 |
+
"reward_std": 0.5151762366294861,
|
| 557 |
+
"rewards/cosine_scaled_reward/mean": 0.03630712628364563,
|
| 558 |
+
"rewards/cosine_scaled_reward/std": 0.47822093963623047,
|
| 559 |
+
"rewards/format_reward/mean": 0.4166666567325592,
|
| 560 |
+
"rewards/format_reward/std": 0.49822381138801575,
|
| 561 |
+
"step": 19
|
| 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 |
+
"completions/clipped_ratio": 0.375,
|
| 570 |
+
"completions/max_length": 3584.0,
|
| 571 |
+
"completions/max_terminated_length": 3561.0,
|
| 572 |
+
"completions/mean_length": 2221.1875,
|
| 573 |
+
"completions/mean_terminated_length": 1403.5001220703125,
|
| 574 |
+
"completions/min_length": 359.0,
|
| 575 |
+
"completions/min_terminated_length": 359.0,
|
| 576 |
+
"epoch": 0.022857142857142857,
|
| 577 |
+
"frac_reward_zero_std": 0.0,
|
| 578 |
+
"grad_norm": 1.0431965589523315,
|
| 579 |
+
"kl": 0.0006084442138671875,
|
| 580 |
+
"learning_rate": 8.01636806561836e-07,
|
| 581 |
+
"loss": -0.6551,
|
| 582 |
+
"num_tokens": 2881771.0,
|
| 583 |
+
"policy_entropy_avg": 8.125,
|
| 584 |
+
"reward": 0.3996829688549042,
|
| 585 |
+
"reward_std": 0.34862709045410156,
|
| 586 |
+
"rewards/cosine_scaled_reward/mean": -0.06755157560110092,
|
| 587 |
+
"rewards/cosine_scaled_reward/std": 0.39543306827545166,
|
| 588 |
+
"rewards/format_reward/mean": 0.75,
|
| 589 |
+
"rewards/format_reward/std": 0.4375949800014496,
|
| 590 |
+
"step": 20
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"clip_ratio/high_max": 0.0,
|
| 594 |
+
"clip_ratio/high_mean": 0.0,
|
| 595 |
+
"clip_ratio/low_mean": 0.0,
|
| 596 |
+
"clip_ratio/low_min": 0.0,
|
| 597 |
+
"clip_ratio/region_mean": 0.0,
|
| 598 |
+
"completions/clipped_ratio": 0.5208333333333333,
|
| 599 |
+
"completions/max_length": 3584.0,
|
| 600 |
+
"completions/max_terminated_length": 3408.0,
|
| 601 |
+
"completions/mean_length": 2486.52099609375,
|
| 602 |
+
"completions/mean_terminated_length": 1293.6087646484375,
|
| 603 |
+
"completions/min_length": 343.0,
|
| 604 |
+
"completions/min_terminated_length": 343.0,
|
| 605 |
+
"epoch": 0.024,
|
| 606 |
+
"frac_reward_zero_std": 0.0,
|
| 607 |
+
"grad_norm": 1.1584967374801636,
|
| 608 |
+
"kl": 0.0008227030436197916,
|
| 609 |
+
"learning_rate": 7.75e-07,
|
| 610 |
+
"loss": -0.3451,
|
| 611 |
+
"num_tokens": 3009230.0,
|
| 612 |
+
"policy_entropy_avg": 8.125,
|
| 613 |
+
"reward": 0.08958987891674042,
|
| 614 |
+
"reward_std": 0.3577960431575775,
|
| 615 |
+
"rewards/cosine_scaled_reward/mean": -0.17066805064678192,
|
| 616 |
+
"rewards/cosine_scaled_reward/std": 0.31892454624176025,
|
| 617 |
+
"rewards/format_reward/mean": 0.4791666567325592,
|
| 618 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 619 |
+
"step": 21
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"clip_ratio/high_max": 0.0,
|
| 623 |
+
"clip_ratio/high_mean": 0.0,
|
| 624 |
+
"clip_ratio/low_mean": 0.0,
|
| 625 |
+
"clip_ratio/low_min": 0.0,
|
| 626 |
+
"clip_ratio/region_mean": 0.0,
|
| 627 |
+
"completions/clipped_ratio": 0.22916666666666663,
|
| 628 |
+
"completions/max_length": 3584.0,
|
| 629 |
+
"completions/max_terminated_length": 3488.0,
|
| 630 |
+
"completions/mean_length": 1677.9583740234375,
|
| 631 |
+
"completions/mean_terminated_length": 1111.29736328125,
|
| 632 |
+
"completions/min_length": 253.0,
|
| 633 |
+
"completions/min_terminated_length": 253.0,
|
| 634 |
+
"epoch": 0.025142857142857144,
|
| 635 |
+
"frac_reward_zero_std": 0.0,
|
| 636 |
+
"grad_norm": 2.2852933406829834,
|
| 637 |
+
"kl": 0.000949859619140625,
|
| 638 |
+
"learning_rate": 7.472670160550848e-07,
|
| 639 |
+
"loss": -2.4718,
|
| 640 |
+
"num_tokens": 3096534.0,
|
| 641 |
+
"policy_entropy_avg": 8.125,
|
| 642 |
+
"reward": 0.5749191045761108,
|
| 643 |
+
"reward_std": 0.3676217198371887,
|
| 644 |
+
"rewards/cosine_scaled_reward/mean": 0.015162050724029541,
|
| 645 |
+
"rewards/cosine_scaled_reward/std": 0.4574022591114044,
|
| 646 |
+
"rewards/format_reward/mean": 0.8541666865348816,
|
| 647 |
+
"rewards/format_reward/std": 0.3566739857196808,
|
| 648 |
+
"step": 22
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"clip_ratio/high_max": 0.0,
|
| 652 |
+
"clip_ratio/high_mean": 0.0,
|
| 653 |
+
"clip_ratio/low_mean": 0.0,
|
| 654 |
+
"clip_ratio/low_min": 0.0,
|
| 655 |
+
"clip_ratio/region_mean": 0.0,
|
| 656 |
+
"completions/clipped_ratio": 0.4375,
|
| 657 |
+
"completions/max_length": 3584.0,
|
| 658 |
+
"completions/max_terminated_length": 3523.0,
|
| 659 |
+
"completions/mean_length": 2406.979248046875,
|
| 660 |
+
"completions/mean_terminated_length": 1491.5185546875,
|
| 661 |
+
"completions/min_length": 402.0,
|
| 662 |
+
"completions/min_terminated_length": 402.0,
|
| 663 |
+
"epoch": 0.026285714285714287,
|
| 664 |
+
"frac_reward_zero_std": 0.0,
|
| 665 |
+
"grad_norm": 0.7508360147476196,
|
| 666 |
+
"kl": 0.0007603963216145834,
|
| 667 |
+
"learning_rate": 7.185729670371604e-07,
|
| 668 |
+
"loss": 0.5997,
|
| 669 |
+
"num_tokens": 3220163.0,
|
| 670 |
+
"policy_entropy_avg": 8.125,
|
| 671 |
+
"reward": 0.29059678316116333,
|
| 672 |
+
"reward_std": 0.5978084802627563,
|
| 673 |
+
"rewards/cosine_scaled_reward/mean": -0.06813069432973862,
|
| 674 |
+
"rewards/cosine_scaled_reward/std": 0.44070079922676086,
|
| 675 |
+
"rewards/format_reward/mean": 0.5833333134651184,
|
| 676 |
+
"rewards/format_reward/std": 0.49822378158569336,
|
| 677 |
+
"step": 23
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"clip_ratio/high_max": 0.0,
|
| 681 |
+
"clip_ratio/high_mean": 0.0,
|
| 682 |
+
"clip_ratio/low_mean": 0.0,
|
| 683 |
+
"clip_ratio/low_min": 0.0,
|
| 684 |
+
"clip_ratio/region_mean": 0.0,
|
| 685 |
+
"completions/clipped_ratio": 0.5,
|
| 686 |
+
"completions/max_length": 3584.0,
|
| 687 |
+
"completions/max_terminated_length": 3552.0,
|
| 688 |
+
"completions/mean_length": 2770.39599609375,
|
| 689 |
+
"completions/mean_terminated_length": 1956.791748046875,
|
| 690 |
+
"completions/min_length": 381.0,
|
| 691 |
+
"completions/min_terminated_length": 381.0,
|
| 692 |
+
"epoch": 0.027428571428571427,
|
| 693 |
+
"frac_reward_zero_std": 0.0,
|
| 694 |
+
"grad_norm": 0.5464559197425842,
|
| 695 |
+
"kl": 0.0007966359456380209,
|
| 696 |
+
"learning_rate": 6.890576474687263e-07,
|
| 697 |
+
"loss": -0.6853,
|
| 698 |
+
"num_tokens": 3361338.0,
|
| 699 |
+
"policy_entropy_avg": 8.125,
|
| 700 |
+
"reward": 0.2939620018005371,
|
| 701 |
+
"reward_std": 0.60921311378479,
|
| 702 |
+
"rewards/cosine_scaled_reward/mean": -0.034292057156562805,
|
| 703 |
+
"rewards/cosine_scaled_reward/std": 0.44030100107192993,
|
| 704 |
+
"rewards/format_reward/mean": 0.5208333134651184,
|
| 705 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 706 |
+
"step": 24
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"clip_ratio/high_max": 0.0,
|
| 710 |
+
"clip_ratio/high_mean": 0.0,
|
| 711 |
+
"clip_ratio/low_mean": 0.0,
|
| 712 |
+
"clip_ratio/low_min": 0.0,
|
| 713 |
+
"clip_ratio/region_mean": 0.0,
|
| 714 |
+
"completions/clipped_ratio": 0.6041666666666667,
|
| 715 |
+
"completions/max_length": 3584.0,
|
| 716 |
+
"completions/max_terminated_length": 2654.0,
|
| 717 |
+
"completions/mean_length": 2720.8125,
|
| 718 |
+
"completions/mean_terminated_length": 1403.3157958984375,
|
| 719 |
+
"completions/min_length": 509.0,
|
| 720 |
+
"completions/min_terminated_length": 509.0,
|
| 721 |
+
"epoch": 0.02857142857142857,
|
| 722 |
+
"frac_reward_zero_std": 0.0,
|
| 723 |
+
"grad_norm": 0.8707253336906433,
|
| 724 |
+
"kl": 0.000972747802734375,
|
| 725 |
+
"learning_rate": 6.588648530198504e-07,
|
| 726 |
+
"loss": 0.227,
|
| 727 |
+
"num_tokens": 3499659.0,
|
| 728 |
+
"policy_entropy_avg": 8.125,
|
| 729 |
+
"reward": 0.19027817249298096,
|
| 730 |
+
"reward_std": 0.5284578204154968,
|
| 731 |
+
"rewards/cosine_scaled_reward/mean": -0.051548827439546585,
|
| 732 |
+
"rewards/cosine_scaled_reward/std": 0.4388459026813507,
|
| 733 |
+
"rewards/format_reward/mean": 0.3958333432674408,
|
| 734 |
+
"rewards/format_reward/std": 0.49420398473739624,
|
| 735 |
+
"step": 25
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"clip_ratio/high_max": 0.0,
|
| 739 |
+
"clip_ratio/high_mean": 0.0,
|
| 740 |
+
"clip_ratio/low_mean": 0.0,
|
| 741 |
+
"clip_ratio/low_min": 0.0,
|
| 742 |
+
"clip_ratio/region_mean": 0.0,
|
| 743 |
+
"completions/clipped_ratio": 0.6041666666666667,
|
| 744 |
+
"completions/max_length": 3584.0,
|
| 745 |
+
"completions/max_terminated_length": 3077.0,
|
| 746 |
+
"completions/mean_length": 2909.729248046875,
|
| 747 |
+
"completions/mean_terminated_length": 1880.5789794921875,
|
| 748 |
+
"completions/min_length": 943.0,
|
| 749 |
+
"completions/min_terminated_length": 943.0,
|
| 750 |
+
"epoch": 0.029714285714285714,
|
| 751 |
+
"frac_reward_zero_std": 0.0,
|
| 752 |
+
"grad_norm": 0.9067810773849487,
|
| 753 |
+
"kl": 0.0007839202880859375,
|
| 754 |
+
"learning_rate": 6.281416799501187e-07,
|
| 755 |
+
"loss": 0.0152,
|
| 756 |
+
"num_tokens": 3647300.0,
|
| 757 |
+
"policy_entropy_avg": 8.125,
|
| 758 |
+
"reward": 0.1333228200674057,
|
| 759 |
+
"reward_std": 0.3536017835140228,
|
| 760 |
+
"rewards/cosine_scaled_reward/mean": -0.1057773232460022,
|
| 761 |
+
"rewards/cosine_scaled_reward/std": 0.3763800859451294,
|
| 762 |
+
"rewards/format_reward/mean": 0.4166666567325592,
|
| 763 |
+
"rewards/format_reward/std": 0.49822381138801575,
|
| 764 |
+
"step": 26
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"clip_ratio/high_max": 0.0,
|
| 768 |
+
"clip_ratio/high_mean": 0.0,
|
| 769 |
+
"clip_ratio/low_mean": 0.0,
|
| 770 |
+
"clip_ratio/low_min": 0.0,
|
| 771 |
+
"clip_ratio/region_mean": 0.0,
|
| 772 |
+
"completions/clipped_ratio": 0.6458333333333333,
|
| 773 |
+
"completions/max_length": 3584.0,
|
| 774 |
+
"completions/max_terminated_length": 3495.0,
|
| 775 |
+
"completions/mean_length": 2900.666748046875,
|
| 776 |
+
"completions/mean_terminated_length": 1654.5882568359375,
|
| 777 |
+
"completions/min_length": 471.0,
|
| 778 |
+
"completions/min_terminated_length": 471.0,
|
| 779 |
+
"epoch": 0.030857142857142857,
|
| 780 |
+
"frac_reward_zero_std": 0.0,
|
| 781 |
+
"grad_norm": 2.270411729812622,
|
| 782 |
+
"kl": 0.00095367431640625,
|
| 783 |
+
"learning_rate": 5.97037808470444e-07,
|
| 784 |
+
"loss": -0.0823,
|
| 785 |
+
"num_tokens": 3794248.0,
|
| 786 |
+
"policy_entropy_avg": 8.125,
|
| 787 |
+
"reward": 0.1146991103887558,
|
| 788 |
+
"reward_std": 0.4322494864463806,
|
| 789 |
+
"rewards/cosine_scaled_reward/mean": -0.10968658328056335,
|
| 790 |
+
"rewards/cosine_scaled_reward/std": 0.319176584482193,
|
| 791 |
+
"rewards/format_reward/mean": 0.3958333432674408,
|
| 792 |
+
"rewards/format_reward/std": 0.49420398473739624,
|
| 793 |
+
"step": 27
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"clip_ratio/high_max": 0.0,
|
| 797 |
+
"clip_ratio/high_mean": 0.0,
|
| 798 |
+
"clip_ratio/low_mean": 0.0,
|
| 799 |
+
"clip_ratio/low_min": 0.0,
|
| 800 |
+
"clip_ratio/region_mean": 0.0,
|
| 801 |
+
"completions/clipped_ratio": 0.5208333333333333,
|
| 802 |
+
"completions/max_length": 3584.0,
|
| 803 |
+
"completions/max_terminated_length": 3292.0,
|
| 804 |
+
"completions/mean_length": 2623.14599609375,
|
| 805 |
+
"completions/mean_terminated_length": 1578.7391357421875,
|
| 806 |
+
"completions/min_length": 433.0,
|
| 807 |
+
"completions/min_terminated_length": 433.0,
|
| 808 |
+
"epoch": 0.032,
|
| 809 |
+
"frac_reward_zero_std": 0.0,
|
| 810 |
+
"grad_norm": 1.2852171659469604,
|
| 811 |
+
"kl": 0.0007775624593098959,
|
| 812 |
+
"learning_rate": 5.657047735161255e-07,
|
| 813 |
+
"loss": -0.6041,
|
| 814 |
+
"num_tokens": 3927911.0,
|
| 815 |
+
"policy_entropy_avg": 8.125,
|
| 816 |
+
"reward": 0.45130008459091187,
|
| 817 |
+
"reward_std": 0.44697779417037964,
|
| 818 |
+
"rewards/cosine_scaled_reward/mean": 0.08673719316720963,
|
| 819 |
+
"rewards/cosine_scaled_reward/std": 0.49450618028640747,
|
| 820 |
+
"rewards/format_reward/mean": 0.5208333134651184,
|
| 821 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 822 |
+
"step": 28
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"clip_ratio/high_max": 0.0,
|
| 826 |
+
"clip_ratio/high_mean": 0.0,
|
| 827 |
+
"clip_ratio/low_mean": 0.0,
|
| 828 |
+
"clip_ratio/low_min": 0.0,
|
| 829 |
+
"clip_ratio/region_mean": 0.0,
|
| 830 |
+
"completions/clipped_ratio": 0.8125,
|
| 831 |
+
"completions/max_length": 3584.0,
|
| 832 |
+
"completions/max_terminated_length": 3166.0,
|
| 833 |
+
"completions/mean_length": 3315.20849609375,
|
| 834 |
+
"completions/mean_terminated_length": 2150.444580078125,
|
| 835 |
+
"completions/min_length": 597.0,
|
| 836 |
+
"completions/min_terminated_length": 597.0,
|
| 837 |
+
"epoch": 0.03314285714285714,
|
| 838 |
+
"frac_reward_zero_std": 0.0,
|
| 839 |
+
"grad_norm": 0.5690914392471313,
|
| 840 |
+
"kl": 0.0009403228759765625,
|
| 841 |
+
"learning_rate": 5.342952264838747e-07,
|
| 842 |
+
"loss": 0.4887,
|
| 843 |
+
"num_tokens": 4094895.0,
|
| 844 |
+
"policy_entropy_avg": 8.135416666666666,
|
| 845 |
+
"reward": 0.017309244722127914,
|
| 846 |
+
"reward_std": 0.44446587562561035,
|
| 847 |
+
"rewards/cosine_scaled_reward/mean": -0.10126852989196777,
|
| 848 |
+
"rewards/cosine_scaled_reward/std": 0.3542354702949524,
|
| 849 |
+
"rewards/format_reward/mean": 0.2291666716337204,
|
| 850 |
+
"rewards/format_reward/std": 0.4247443675994873,
|
| 851 |
+
"step": 29
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"clip_ratio/high_max": 0.0,
|
| 855 |
+
"clip_ratio/high_mean": 0.0,
|
| 856 |
+
"clip_ratio/low_mean": 0.0,
|
| 857 |
+
"clip_ratio/low_min": 0.0,
|
| 858 |
+
"clip_ratio/region_mean": 0.0,
|
| 859 |
+
"completions/clipped_ratio": 0.5833333333333333,
|
| 860 |
+
"completions/max_length": 3584.0,
|
| 861 |
+
"completions/max_terminated_length": 3493.0,
|
| 862 |
+
"completions/mean_length": 2867.52099609375,
|
| 863 |
+
"completions/mean_terminated_length": 1864.4500732421875,
|
| 864 |
+
"completions/min_length": 716.0,
|
| 865 |
+
"completions/min_terminated_length": 716.0,
|
| 866 |
+
"epoch": 0.03428571428571429,
|
| 867 |
+
"frac_reward_zero_std": 0.0,
|
| 868 |
+
"grad_norm": 1.0161747932434082,
|
| 869 |
+
"kl": 0.0011774698893229167,
|
| 870 |
+
"learning_rate": 5.02962191529556e-07,
|
| 871 |
+
"loss": 0.2395,
|
| 872 |
+
"num_tokens": 4240438.0,
|
| 873 |
+
"policy_entropy_avg": 8.125,
|
| 874 |
+
"reward": 0.1694916933774948,
|
| 875 |
+
"reward_std": 0.6536720395088196,
|
| 876 |
+
"rewards/cosine_scaled_reward/mean": -0.09878844022750854,
|
| 877 |
+
"rewards/cosine_scaled_reward/std": 0.39795705676078796,
|
| 878 |
+
"rewards/format_reward/mean": 0.4583333432674408,
|
| 879 |
+
"rewards/format_reward/std": 0.5035336017608643,
|
| 880 |
+
"step": 30
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"clip_ratio/high_max": 0.0,
|
| 884 |
+
"clip_ratio/high_mean": 0.0,
|
| 885 |
+
"clip_ratio/low_mean": 0.0,
|
| 886 |
+
"clip_ratio/low_min": 0.0,
|
| 887 |
+
"clip_ratio/region_mean": 0.0,
|
| 888 |
+
"completions/clipped_ratio": 0.7291666666666667,
|
| 889 |
+
"completions/max_length": 3584.0,
|
| 890 |
+
"completions/max_terminated_length": 3384.0,
|
| 891 |
+
"completions/mean_length": 2928.875,
|
| 892 |
+
"completions/mean_terminated_length": 1165.0770263671875,
|
| 893 |
+
"completions/min_length": 271.0,
|
| 894 |
+
"completions/min_terminated_length": 271.0,
|
| 895 |
+
"epoch": 0.03542857142857143,
|
| 896 |
+
"frac_reward_zero_std": 0.0,
|
| 897 |
+
"grad_norm": 1.1587620973587036,
|
| 898 |
+
"kl": 0.000823974609375,
|
| 899 |
+
"learning_rate": 4.7185832004988133e-07,
|
| 900 |
+
"loss": 0.1645,
|
| 901 |
+
"num_tokens": 4388896.0,
|
| 902 |
+
"policy_entropy_avg": 8.125,
|
| 903 |
+
"reward": 0.0713512971997261,
|
| 904 |
+
"reward_std": 0.43281668424606323,
|
| 905 |
+
"rewards/cosine_scaled_reward/mean": -0.0909477099776268,
|
| 906 |
+
"rewards/cosine_scaled_reward/std": 0.3611638844013214,
|
| 907 |
+
"rewards/format_reward/mean": 0.2916666567325592,
|
| 908 |
+
"rewards/format_reward/std": 0.4593396782875061,
|
| 909 |
+
"step": 31
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"clip_ratio/high_max": 0.0,
|
| 913 |
+
"clip_ratio/high_mean": 0.0,
|
| 914 |
+
"clip_ratio/low_mean": 0.0,
|
| 915 |
+
"clip_ratio/low_min": 0.0,
|
| 916 |
+
"clip_ratio/region_mean": 0.0,
|
| 917 |
+
"completions/clipped_ratio": 0.6041666666666667,
|
| 918 |
+
"completions/max_length": 3584.0,
|
| 919 |
+
"completions/max_terminated_length": 2794.0,
|
| 920 |
+
"completions/mean_length": 2802.9375,
|
| 921 |
+
"completions/mean_terminated_length": 1610.7894287109375,
|
| 922 |
+
"completions/min_length": 529.0,
|
| 923 |
+
"completions/min_terminated_length": 529.0,
|
| 924 |
+
"epoch": 0.036571428571428574,
|
| 925 |
+
"frac_reward_zero_std": 0.0,
|
| 926 |
+
"grad_norm": 0.7285576462745667,
|
| 927 |
+
"kl": 0.0011965433756510417,
|
| 928 |
+
"learning_rate": 4.4113514698014953e-07,
|
| 929 |
+
"loss": 0.5699,
|
| 930 |
+
"num_tokens": 4531201.0,
|
| 931 |
+
"policy_entropy_avg": 8.125,
|
| 932 |
+
"reward": 0.13415873050689697,
|
| 933 |
+
"reward_std": 0.4497126042842865,
|
| 934 |
+
"rewards/cosine_scaled_reward/mean": -0.09471765905618668,
|
| 935 |
+
"rewards/cosine_scaled_reward/std": 0.3660888373851776,
|
| 936 |
+
"rewards/format_reward/mean": 0.3958333432674408,
|
| 937 |
+
"rewards/format_reward/std": 0.49420398473739624,
|
| 938 |
+
"step": 32
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"clip_ratio/high_max": 0.0,
|
| 942 |
+
"clip_ratio/high_mean": 0.0,
|
| 943 |
+
"clip_ratio/low_mean": 0.0,
|
| 944 |
+
"clip_ratio/low_min": 0.0,
|
| 945 |
+
"clip_ratio/region_mean": 0.0,
|
| 946 |
+
"completions/clipped_ratio": 0.6666666666666667,
|
| 947 |
+
"completions/max_length": 3584.0,
|
| 948 |
+
"completions/max_terminated_length": 3527.0,
|
| 949 |
+
"completions/mean_length": 3081.33349609375,
|
| 950 |
+
"completions/mean_terminated_length": 2076.0,
|
| 951 |
+
"completions/min_length": 765.0,
|
| 952 |
+
"completions/min_terminated_length": 765.0,
|
| 953 |
+
"epoch": 0.037714285714285714,
|
| 954 |
+
"frac_reward_zero_std": 0.0,
|
| 955 |
+
"grad_norm": 0.5806891322135925,
|
| 956 |
+
"kl": 0.0007527669270833334,
|
| 957 |
+
"learning_rate": 4.1094235253127374e-07,
|
| 958 |
+
"loss": -0.155,
|
| 959 |
+
"num_tokens": 4687517.0,
|
| 960 |
+
"policy_entropy_avg": 8.125,
|
| 961 |
+
"reward": 0.052633434534072876,
|
| 962 |
+
"reward_std": 0.5880012512207031,
|
| 963 |
+
"rewards/cosine_scaled_reward/mean": -0.14701275527477264,
|
| 964 |
+
"rewards/cosine_scaled_reward/std": 0.3462415635585785,
|
| 965 |
+
"rewards/format_reward/mean": 0.375,
|
| 966 |
+
"rewards/format_reward/std": 0.48924607038497925,
|
| 967 |
+
"step": 33
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"clip_ratio/high_max": 0.0,
|
| 971 |
+
"clip_ratio/high_mean": 0.0,
|
| 972 |
+
"clip_ratio/low_mean": 0.0,
|
| 973 |
+
"clip_ratio/low_min": 0.0,
|
| 974 |
+
"clip_ratio/region_mean": 0.0,
|
| 975 |
+
"completions/clipped_ratio": 0.41666666666666663,
|
| 976 |
+
"completions/max_length": 3584.0,
|
| 977 |
+
"completions/max_terminated_length": 3110.0,
|
| 978 |
+
"completions/mean_length": 2213.104248046875,
|
| 979 |
+
"completions/mean_terminated_length": 1233.8929443359375,
|
| 980 |
+
"completions/min_length": 582.0,
|
| 981 |
+
"completions/min_terminated_length": 582.0,
|
| 982 |
+
"epoch": 0.038857142857142854,
|
| 983 |
+
"frac_reward_zero_std": 0.0,
|
| 984 |
+
"grad_norm": 1.0047067403793335,
|
| 985 |
+
"kl": 0.0015347798665364583,
|
| 986 |
+
"learning_rate": 3.8142703296283953e-07,
|
| 987 |
+
"loss": 0.6987,
|
| 988 |
+
"num_tokens": 4800910.0,
|
| 989 |
+
"policy_entropy_avg": 8.125,
|
| 990 |
+
"reward": 0.43960699439048767,
|
| 991 |
+
"reward_std": 0.5042138695716858,
|
| 992 |
+
"rewards/cosine_scaled_reward/mean": 0.025659168139100075,
|
| 993 |
+
"rewards/cosine_scaled_reward/std": 0.5110981464385986,
|
| 994 |
+
"rewards/format_reward/mean": 0.625,
|
| 995 |
+
"rewards/format_reward/std": 0.48924607038497925,
|
| 996 |
+
"step": 34
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"clip_ratio/high_max": 0.0,
|
| 1000 |
+
"clip_ratio/high_mean": 0.0,
|
| 1001 |
+
"clip_ratio/low_mean": 0.0,
|
| 1002 |
+
"clip_ratio/low_min": 0.0,
|
| 1003 |
+
"clip_ratio/region_mean": 0.0,
|
| 1004 |
+
"completions/clipped_ratio": 0.5833333333333333,
|
| 1005 |
+
"completions/max_length": 3584.0,
|
| 1006 |
+
"completions/max_terminated_length": 3482.0,
|
| 1007 |
+
"completions/mean_length": 2852.125,
|
| 1008 |
+
"completions/mean_terminated_length": 1827.5,
|
| 1009 |
+
"completions/min_length": 369.0,
|
| 1010 |
+
"completions/min_terminated_length": 369.0,
|
| 1011 |
+
"epoch": 0.04,
|
| 1012 |
+
"frac_reward_zero_std": 0.0,
|
| 1013 |
+
"grad_norm": 0.6956940293312073,
|
| 1014 |
+
"kl": 0.0008710225423177084,
|
| 1015 |
+
"learning_rate": 3.5273298394491515e-07,
|
| 1016 |
+
"loss": 0.023,
|
| 1017 |
+
"num_tokens": 4946434.0,
|
| 1018 |
+
"policy_entropy_avg": 8.125,
|
| 1019 |
+
"reward": 0.27234378457069397,
|
| 1020 |
+
"reward_std": 0.6467978954315186,
|
| 1021 |
+
"rewards/cosine_scaled_reward/mean": -0.030088132247328758,
|
| 1022 |
+
"rewards/cosine_scaled_reward/std": 0.4617981016635895,
|
| 1023 |
+
"rewards/format_reward/mean": 0.4791666567325592,
|
| 1024 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 1025 |
+
"step": 35
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"clip_ratio/high_max": 0.0,
|
| 1029 |
+
"clip_ratio/high_mean": 0.0,
|
| 1030 |
+
"clip_ratio/low_mean": 0.0,
|
| 1031 |
+
"clip_ratio/low_min": 0.0,
|
| 1032 |
+
"clip_ratio/region_mean": 0.0,
|
| 1033 |
+
"completions/clipped_ratio": 0.7916666666666666,
|
| 1034 |
+
"completions/max_length": 3584.0,
|
| 1035 |
+
"completions/max_terminated_length": 3373.0,
|
| 1036 |
+
"completions/mean_length": 3279.95849609375,
|
| 1037 |
+
"completions/mean_terminated_length": 2124.60009765625,
|
| 1038 |
+
"completions/min_length": 920.0,
|
| 1039 |
+
"completions/min_terminated_length": 920.0,
|
| 1040 |
+
"epoch": 0.04114285714285714,
|
| 1041 |
+
"frac_reward_zero_std": 0.0,
|
| 1042 |
+
"grad_norm": 0.6212595105171204,
|
| 1043 |
+
"kl": 0.0012715657552083333,
|
| 1044 |
+
"learning_rate": 3.250000000000001e-07,
|
| 1045 |
+
"loss": 0.6297,
|
| 1046 |
+
"num_tokens": 5112206.0,
|
| 1047 |
+
"policy_entropy_avg": 8.135416666666666,
|
| 1048 |
+
"reward": -0.14957016706466675,
|
| 1049 |
+
"reward_std": 0.34586209058761597,
|
| 1050 |
+
"rewards/cosine_scaled_reward/mean": -0.2400539517402649,
|
| 1051 |
+
"rewards/cosine_scaled_reward/std": 0.1841082125902176,
|
| 1052 |
+
"rewards/format_reward/mean": 0.25,
|
| 1053 |
+
"rewards/format_reward/std": 0.4375949800014496,
|
| 1054 |
+
"step": 36
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"clip_ratio/high_max": 0.0,
|
| 1058 |
+
"clip_ratio/high_mean": 0.0,
|
| 1059 |
+
"clip_ratio/low_mean": 0.0,
|
| 1060 |
+
"clip_ratio/low_min": 0.0,
|
| 1061 |
+
"clip_ratio/region_mean": 0.0,
|
| 1062 |
+
"completions/clipped_ratio": 0.75,
|
| 1063 |
+
"completions/max_length": 3584.0,
|
| 1064 |
+
"completions/max_terminated_length": 3444.0,
|
| 1065 |
+
"completions/mean_length": 3138.104248046875,
|
| 1066 |
+
"completions/mean_terminated_length": 1800.416748046875,
|
| 1067 |
+
"completions/min_length": 474.0,
|
| 1068 |
+
"completions/min_terminated_length": 474.0,
|
| 1069 |
+
"epoch": 0.04228571428571429,
|
| 1070 |
+
"frac_reward_zero_std": 0.0,
|
| 1071 |
+
"grad_norm": 0.7584943771362305,
|
| 1072 |
+
"kl": 0.0009466807047526041,
|
| 1073 |
+
"learning_rate": 2.9836319343816397e-07,
|
| 1074 |
+
"loss": 1.1086,
|
| 1075 |
+
"num_tokens": 5271103.0,
|
| 1076 |
+
"policy_entropy_avg": 8.135416666666666,
|
| 1077 |
+
"reward": -0.1891360729932785,
|
| 1078 |
+
"reward_std": 0.35265272855758667,
|
| 1079 |
+
"rewards/cosine_scaled_reward/mean": -0.27048927545547485,
|
| 1080 |
+
"rewards/cosine_scaled_reward/std": 0.19963285326957703,
|
| 1081 |
+
"rewards/format_reward/mean": 0.25,
|
| 1082 |
+
"rewards/format_reward/std": 0.4375949800014496,
|
| 1083 |
+
"step": 37
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"clip_ratio/high_max": 0.0,
|
| 1087 |
+
"clip_ratio/high_mean": 0.0,
|
| 1088 |
+
"clip_ratio/low_mean": 0.0,
|
| 1089 |
+
"clip_ratio/low_min": 0.0,
|
| 1090 |
+
"clip_ratio/region_mean": 0.0,
|
| 1091 |
+
"completions/clipped_ratio": 0.8333333333333334,
|
| 1092 |
+
"completions/max_length": 3584.0,
|
| 1093 |
+
"completions/max_terminated_length": 2923.0,
|
| 1094 |
+
"completions/mean_length": 3194.5,
|
| 1095 |
+
"completions/mean_terminated_length": 1247.0,
|
| 1096 |
+
"completions/min_length": 478.0,
|
| 1097 |
+
"completions/min_terminated_length": 478.0,
|
| 1098 |
+
"epoch": 0.04342857142857143,
|
| 1099 |
+
"frac_reward_zero_std": 0.0,
|
| 1100 |
+
"grad_norm": 0.9534838795661926,
|
| 1101 |
+
"kl": 0.0009625752766927084,
|
| 1102 |
+
"learning_rate": 2.729523361034538e-07,
|
| 1103 |
+
"loss": 0.6428,
|
| 1104 |
+
"num_tokens": 5433223.0,
|
| 1105 |
+
"policy_entropy_avg": 8.125,
|
| 1106 |
+
"reward": -0.03065665066242218,
|
| 1107 |
+
"reward_std": 0.27790385484695435,
|
| 1108 |
+
"rewards/cosine_scaled_reward/mean": -0.10691537708044052,
|
| 1109 |
+
"rewards/cosine_scaled_reward/std": 0.2967289388179779,
|
| 1110 |
+
"rewards/format_reward/mean": 0.1666666716337204,
|
| 1111 |
+
"rewards/format_reward/std": 0.3766217827796936,
|
| 1112 |
+
"step": 38
|
| 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 |
+
"completions/clipped_ratio": 0.5416666666666667,
|
| 1121 |
+
"completions/max_length": 3584.0,
|
| 1122 |
+
"completions/max_terminated_length": 3191.0,
|
| 1123 |
+
"completions/mean_length": 2574.58349609375,
|
| 1124 |
+
"completions/mean_terminated_length": 1381.6363525390625,
|
| 1125 |
+
"completions/min_length": 326.0,
|
| 1126 |
+
"completions/min_terminated_length": 326.0,
|
| 1127 |
+
"epoch": 0.044571428571428574,
|
| 1128 |
+
"frac_reward_zero_std": 0.0,
|
| 1129 |
+
"grad_norm": 38.30278396606445,
|
| 1130 |
+
"kl": 0.0010172526041666667,
|
| 1131 |
+
"learning_rate": 2.488912271385139e-07,
|
| 1132 |
+
"loss": -21.1377,
|
| 1133 |
+
"num_tokens": 5564891.0,
|
| 1134 |
+
"policy_entropy_avg": 8.125,
|
| 1135 |
+
"reward": 0.24889466166496277,
|
| 1136 |
+
"reward_std": 0.4223267734050751,
|
| 1137 |
+
"rewards/cosine_scaled_reward/mean": -0.03770924732089043,
|
| 1138 |
+
"rewards/cosine_scaled_reward/std": 0.4333648681640625,
|
| 1139 |
+
"rewards/format_reward/mean": 0.4583333432674408,
|
| 1140 |
+
"rewards/format_reward/std": 0.5035336017608643,
|
| 1141 |
+
"step": 39
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"clip_ratio/high_max": 0.0,
|
| 1145 |
+
"clip_ratio/high_mean": 0.0,
|
| 1146 |
+
"clip_ratio/low_mean": 0.0,
|
| 1147 |
+
"clip_ratio/low_min": 0.0,
|
| 1148 |
+
"clip_ratio/region_mean": 0.0,
|
| 1149 |
+
"completions/clipped_ratio": 0.5,
|
| 1150 |
+
"completions/max_length": 3584.0,
|
| 1151 |
+
"completions/max_terminated_length": 3383.0,
|
| 1152 |
+
"completions/mean_length": 2522.6875,
|
| 1153 |
+
"completions/mean_terminated_length": 1461.375,
|
| 1154 |
+
"completions/min_length": 430.0,
|
| 1155 |
+
"completions/min_terminated_length": 430.0,
|
| 1156 |
+
"epoch": 0.045714285714285714,
|
| 1157 |
+
"frac_reward_zero_std": 0.0,
|
| 1158 |
+
"grad_norm": 0.7990902662277222,
|
| 1159 |
+
"kl": 0.0011386871337890625,
|
| 1160 |
+
"learning_rate": 2.2629708984760706e-07,
|
| 1161 |
+
"loss": -0.3002,
|
| 1162 |
+
"num_tokens": 5693534.0,
|
| 1163 |
+
"policy_entropy_avg": 8.125,
|
| 1164 |
+
"reward": 0.23835521936416626,
|
| 1165 |
+
"reward_std": 0.3610273599624634,
|
| 1166 |
+
"rewards/cosine_scaled_reward/mean": -0.07706651836633682,
|
| 1167 |
+
"rewards/cosine_scaled_reward/std": 0.39551421999931335,
|
| 1168 |
+
"rewards/format_reward/mean": 0.5208333134651184,
|
| 1169 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 1170 |
+
"step": 40
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"clip_ratio/high_max": 0.0,
|
| 1174 |
+
"clip_ratio/high_mean": 0.0,
|
| 1175 |
+
"clip_ratio/low_mean": 0.0,
|
| 1176 |
+
"clip_ratio/low_min": 0.0,
|
| 1177 |
+
"clip_ratio/region_mean": 0.0,
|
| 1178 |
+
"completions/clipped_ratio": 0.5,
|
| 1179 |
+
"completions/max_length": 3584.0,
|
| 1180 |
+
"completions/max_terminated_length": 3214.0,
|
| 1181 |
+
"completions/mean_length": 2528.33349609375,
|
| 1182 |
+
"completions/mean_terminated_length": 1472.666748046875,
|
| 1183 |
+
"completions/min_length": 547.0,
|
| 1184 |
+
"completions/min_terminated_length": 547.0,
|
| 1185 |
+
"epoch": 0.046857142857142854,
|
| 1186 |
+
"frac_reward_zero_std": 0.0,
|
| 1187 |
+
"grad_norm": 1.0874838829040527,
|
| 1188 |
+
"kl": 0.001148223876953125,
|
| 1189 |
+
"learning_rate": 2.0528000059645995e-07,
|
| 1190 |
+
"loss": -0.3434,
|
| 1191 |
+
"num_tokens": 5823258.0,
|
| 1192 |
+
"policy_entropy_avg": 8.125,
|
| 1193 |
+
"reward": 0.3574024438858032,
|
| 1194 |
+
"reward_std": 0.6161948442459106,
|
| 1195 |
+
"rewards/cosine_scaled_reward/mean": 0.01450828742235899,
|
| 1196 |
+
"rewards/cosine_scaled_reward/std": 0.46221035718917847,
|
| 1197 |
+
"rewards/format_reward/mean": 0.5208333134651184,
|
| 1198 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 1199 |
+
"step": 41
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"clip_ratio/high_max": 0.0,
|
| 1203 |
+
"clip_ratio/high_mean": 0.0,
|
| 1204 |
+
"clip_ratio/low_mean": 0.0,
|
| 1205 |
+
"clip_ratio/low_min": 0.0,
|
| 1206 |
+
"clip_ratio/region_mean": 0.0,
|
| 1207 |
+
"completions/clipped_ratio": 0.6458333333333333,
|
| 1208 |
+
"completions/max_length": 3584.0,
|
| 1209 |
+
"completions/max_terminated_length": 2705.0,
|
| 1210 |
+
"completions/mean_length": 2656.416748046875,
|
| 1211 |
+
"completions/mean_terminated_length": 964.941162109375,
|
| 1212 |
+
"completions/min_length": 209.0,
|
| 1213 |
+
"completions/min_terminated_length": 209.0,
|
| 1214 |
+
"epoch": 0.048,
|
| 1215 |
+
"frac_reward_zero_std": 0.0,
|
| 1216 |
+
"grad_norm": 1.283677339553833,
|
| 1217 |
+
"kl": 0.0011049906412760417,
|
| 1218 |
+
"learning_rate": 1.8594235253127372e-07,
|
| 1219 |
+
"loss": 0.27,
|
| 1220 |
+
"num_tokens": 5958536.0,
|
| 1221 |
+
"policy_entropy_avg": 8.125,
|
| 1222 |
+
"reward": -0.12885941565036774,
|
| 1223 |
+
"reward_std": 0.28405821323394775,
|
| 1224 |
+
"rewards/cosine_scaled_reward/mean": -0.2866226136684418,
|
| 1225 |
+
"rewards/cosine_scaled_reward/std": 0.1905842125415802,
|
| 1226 |
+
"rewards/format_reward/mean": 0.375,
|
| 1227 |
+
"rewards/format_reward/std": 0.48924607038497925,
|
| 1228 |
+
"step": 42
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"clip_ratio/high_max": 0.0,
|
| 1232 |
+
"clip_ratio/high_mean": 0.0,
|
| 1233 |
+
"clip_ratio/low_mean": 0.0,
|
| 1234 |
+
"clip_ratio/low_min": 0.0,
|
| 1235 |
+
"clip_ratio/region_mean": 0.0,
|
| 1236 |
+
"completions/clipped_ratio": 0.5833333333333333,
|
| 1237 |
+
"completions/max_length": 3584.0,
|
| 1238 |
+
"completions/max_terminated_length": 2993.0,
|
| 1239 |
+
"completions/mean_length": 2602.354248046875,
|
| 1240 |
+
"completions/mean_terminated_length": 1228.050048828125,
|
| 1241 |
+
"completions/min_length": 586.0,
|
| 1242 |
+
"completions/min_terminated_length": 586.0,
|
| 1243 |
+
"epoch": 0.04914285714285714,
|
| 1244 |
+
"frac_reward_zero_std": 0.0,
|
| 1245 |
+
"grad_norm": 0.7168472409248352,
|
| 1246 |
+
"kl": 0.0010967254638671875,
|
| 1247 |
+
"learning_rate": 1.6837835672960831e-07,
|
| 1248 |
+
"loss": -0.0951,
|
| 1249 |
+
"num_tokens": 6092161.0,
|
| 1250 |
+
"policy_entropy_avg": 8.114583333333334,
|
| 1251 |
+
"reward": 0.21493911743164062,
|
| 1252 |
+
"reward_std": 0.4637143015861511,
|
| 1253 |
+
"rewards/cosine_scaled_reward/mean": -0.07424557209014893,
|
| 1254 |
+
"rewards/cosine_scaled_reward/std": 0.43746232986450195,
|
| 1255 |
+
"rewards/format_reward/mean": 0.4791666567325592,
|
| 1256 |
+
"rewards/format_reward/std": 0.5048523545265198,
|
| 1257 |
+
"step": 43
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"clip_ratio/high_max": 0.0,
|
| 1261 |
+
"clip_ratio/high_mean": 0.0,
|
| 1262 |
+
"clip_ratio/low_mean": 0.0,
|
| 1263 |
+
"clip_ratio/low_min": 0.0,
|
| 1264 |
+
"clip_ratio/region_mean": 0.0,
|
| 1265 |
+
"completions/clipped_ratio": 0.5833333333333333,
|
| 1266 |
+
"completions/max_length": 3584.0,
|
| 1267 |
+
"completions/max_terminated_length": 3087.0,
|
| 1268 |
+
"completions/mean_length": 2491.875,
|
| 1269 |
+
"completions/mean_terminated_length": 962.9000244140625,
|
| 1270 |
+
"completions/min_length": 340.0,
|
| 1271 |
+
"completions/min_terminated_length": 340.0,
|
| 1272 |
+
"epoch": 0.05028571428571429,
|
| 1273 |
+
"frac_reward_zero_std": 0.0,
|
| 1274 |
+
"grad_norm": 0.7064040303230286,
|
| 1275 |
+
"kl": 0.0012423197428385417,
|
| 1276 |
+
"learning_rate": 1.5267358321348285e-07,
|
| 1277 |
+
"loss": 0.4683,
|
| 1278 |
+
"num_tokens": 6219793.0,
|
| 1279 |
+
"policy_entropy_avg": 8.125,
|
| 1280 |
+
"reward": 0.2282370775938034,
|
| 1281 |
+
"reward_std": 0.4510888457298279,
|
| 1282 |
+
"rewards/cosine_scaled_reward/mean": -0.04318302869796753,
|
| 1283 |
+
"rewards/cosine_scaled_reward/std": 0.4827073812484741,
|
| 1284 |
+
"rewards/format_reward/mean": 0.4375,
|
| 1285 |
+
"rewards/format_reward/std": 0.5013279914855957,
|
| 1286 |
+
"step": 44
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"clip_ratio/high_max": 0.0,
|
| 1290 |
+
"clip_ratio/high_mean": 0.0,
|
| 1291 |
+
"clip_ratio/low_mean": 0.0,
|
| 1292 |
+
"clip_ratio/low_min": 0.0,
|
| 1293 |
+
"clip_ratio/region_mean": 0.0,
|
| 1294 |
+
"completions/clipped_ratio": 0.8125,
|
| 1295 |
+
"completions/max_length": 3584.0,
|
| 1296 |
+
"completions/max_terminated_length": 3524.0,
|
| 1297 |
+
"completions/mean_length": 3407.64599609375,
|
| 1298 |
+
"completions/mean_terminated_length": 2643.444580078125,
|
| 1299 |
+
"completions/min_length": 1086.0,
|
| 1300 |
+
"completions/min_terminated_length": 1086.0,
|
| 1301 |
+
"epoch": 0.05142857142857143,
|
| 1302 |
+
"frac_reward_zero_std": 0.0,
|
| 1303 |
+
"grad_norm": 0.6878222227096558,
|
| 1304 |
+
"kl": 0.0010045369466145833,
|
| 1305 |
+
"learning_rate": 1.3890454406082956e-07,
|
| 1306 |
+
"loss": 0.5395,
|
| 1307 |
+
"num_tokens": 6392102.0,
|
| 1308 |
+
"policy_entropy_avg": 8.135416666666666,
|
| 1309 |
+
"reward": 0.18636834621429443,
|
| 1310 |
+
"reward_std": 0.656842827796936,
|
| 1311 |
+
"rewards/cosine_scaled_reward/mean": -0.012889747507870197,
|
| 1312 |
+
"rewards/cosine_scaled_reward/std": 0.47439250349998474,
|
| 1313 |
+
"rewards/format_reward/mean": 0.3125,
|
| 1314 |
+
"rewards/format_reward/std": 0.4684174358844757,
|
| 1315 |
+
"step": 45
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"clip_ratio/high_max": 0.0,
|
| 1319 |
+
"clip_ratio/high_mean": 0.0,
|
| 1320 |
+
"clip_ratio/low_mean": 0.0,
|
| 1321 |
+
"clip_ratio/low_min": 0.0,
|
| 1322 |
+
"clip_ratio/region_mean": 0.0,
|
| 1323 |
+
"completions/clipped_ratio": 0.6666666666666667,
|
| 1324 |
+
"completions/max_length": 3584.0,
|
| 1325 |
+
"completions/max_terminated_length": 1561.0,
|
| 1326 |
+
"completions/mean_length": 2694.75,
|
| 1327 |
+
"completions/mean_terminated_length": 916.25,
|
| 1328 |
+
"completions/min_length": 596.0,
|
| 1329 |
+
"completions/min_terminated_length": 596.0,
|
| 1330 |
+
"epoch": 0.052571428571428575,
|
| 1331 |
+
"frac_reward_zero_std": 0.0,
|
| 1332 |
+
"grad_norm": 1.1607235670089722,
|
| 1333 |
+
"kl": 0.0015538533528645833,
|
| 1334 |
+
"learning_rate": 1.2713832064634125e-07,
|
| 1335 |
+
"loss": -0.0955,
|
| 1336 |
+
"num_tokens": 6529826.0,
|
| 1337 |
+
"policy_entropy_avg": 8.125,
|
| 1338 |
+
"reward": -0.13178138434886932,
|
| 1339 |
+
"reward_std": 0.2914193272590637,
|
| 1340 |
+
"rewards/cosine_scaled_reward/mean": -0.26803696155548096,
|
| 1341 |
+
"rewards/cosine_scaled_reward/std": 0.174865260720253,
|
| 1342 |
+
"rewards/format_reward/mean": 0.3333333432674408,
|
| 1343 |
+
"rewards/format_reward/std": 0.47639307379722595,
|
| 1344 |
+
"step": 46
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"clip_ratio/high_max": 0.0,
|
| 1348 |
+
"clip_ratio/high_mean": 0.0,
|
| 1349 |
+
"clip_ratio/low_mean": 0.0,
|
| 1350 |
+
"clip_ratio/low_min": 0.0,
|
| 1351 |
+
"clip_ratio/region_mean": 0.0,
|
| 1352 |
+
"completions/clipped_ratio": 0.45833333333333337,
|
| 1353 |
+
"completions/max_length": 3584.0,
|
| 1354 |
+
"completions/max_terminated_length": 3485.0,
|
| 1355 |
+
"completions/mean_length": 2533.6875,
|
| 1356 |
+
"completions/mean_terminated_length": 1644.9615478515625,
|
| 1357 |
+
"completions/min_length": 296.0,
|
| 1358 |
+
"completions/min_terminated_length": 296.0,
|
| 1359 |
+
"epoch": 0.053714285714285714,
|
| 1360 |
+
"frac_reward_zero_std": 0.0,
|
| 1361 |
+
"grad_norm": 2.012225389480591,
|
| 1362 |
+
"kl": 0.000888824462890625,
|
| 1363 |
+
"learning_rate": 1.1743223682775649e-07,
|
| 1364 |
+
"loss": -1.3532,
|
| 1365 |
+
"num_tokens": 6659243.0,
|
| 1366 |
+
"policy_entropy_avg": 8.125,
|
| 1367 |
+
"reward": 0.5446165800094604,
|
| 1368 |
+
"reward_std": 0.5056490898132324,
|
| 1369 |
+
"rewards/cosine_scaled_reward/mean": 0.11685246229171753,
|
| 1370 |
+
"rewards/cosine_scaled_reward/std": 0.5205204486846924,
|
| 1371 |
+
"rewards/format_reward/mean": 0.6041666865348816,
|
| 1372 |
+
"rewards/format_reward/std": 0.49420401453971863,
|
| 1373 |
+
"step": 47
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"clip_ratio/high_max": 0.0,
|
| 1377 |
+
"clip_ratio/high_mean": 0.0,
|
| 1378 |
+
"clip_ratio/low_mean": 0.0,
|
| 1379 |
+
"clip_ratio/low_min": 0.0,
|
| 1380 |
+
"clip_ratio/region_mean": 0.0,
|
| 1381 |
+
"completions/clipped_ratio": 0.5208333333333333,
|
| 1382 |
+
"completions/max_length": 3584.0,
|
| 1383 |
+
"completions/max_terminated_length": 3350.0,
|
| 1384 |
+
"completions/mean_length": 2581.33349609375,
|
| 1385 |
+
"completions/mean_terminated_length": 1491.478271484375,
|
| 1386 |
+
"completions/min_length": 552.0,
|
| 1387 |
+
"completions/min_terminated_length": 552.0,
|
| 1388 |
+
"epoch": 0.054857142857142854,
|
| 1389 |
+
"frac_reward_zero_std": 0.0,
|
| 1390 |
+
"grad_norm": 0.6057020425796509,
|
| 1391 |
+
"kl": 0.00128936767578125,
|
| 1392 |
+
"learning_rate": 1.0983357966978745e-07,
|
| 1393 |
+
"loss": 0.4498,
|
| 1394 |
+
"num_tokens": 6791055.0,
|
| 1395 |
+
"policy_entropy_avg": 8.125,
|
| 1396 |
+
"reward": 0.21042108535766602,
|
| 1397 |
+
"reward_std": 0.5790350437164307,
|
| 1398 |
+
"rewards/cosine_scaled_reward/mean": -0.10897094756364822,
|
| 1399 |
+
"rewards/cosine_scaled_reward/std": 0.40693506598472595,
|
| 1400 |
+
"rewards/format_reward/mean": 0.5416666865348816,
|
| 1401 |
+
"rewards/format_reward/std": 0.5035336017608643,
|
| 1402 |
+
"step": 48
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"clip_ratio/high_max": 0.0,
|
| 1406 |
+
"clip_ratio/high_mean": 0.0,
|
| 1407 |
+
"clip_ratio/low_mean": 0.0,
|
| 1408 |
+
"clip_ratio/low_min": 0.0,
|
| 1409 |
+
"clip_ratio/region_mean": 0.0,
|
| 1410 |
+
"completions/clipped_ratio": 0.39583333333333337,
|
| 1411 |
+
"completions/max_length": 3584.0,
|
| 1412 |
+
"completions/max_terminated_length": 3352.0,
|
| 1413 |
+
"completions/mean_length": 2207.791748046875,
|
| 1414 |
+
"completions/mean_terminated_length": 1306.137939453125,
|
| 1415 |
+
"completions/min_length": 281.0,
|
| 1416 |
+
"completions/min_terminated_length": 281.0,
|
| 1417 |
+
"epoch": 0.056,
|
| 1418 |
+
"frac_reward_zero_std": 0.0,
|
| 1419 |
+
"grad_norm": 0.9485021233558655,
|
| 1420 |
+
"kl": 0.0012709299723307292,
|
| 1421 |
+
"learning_rate": 1.0437936906629334e-07,
|
| 1422 |
+
"loss": -0.5865,
|
| 1423 |
+
"num_tokens": 6904577.0,
|
| 1424 |
+
"policy_entropy_avg": 8.125,
|
| 1425 |
+
"reward": 0.2760649025440216,
|
| 1426 |
+
"reward_std": 0.5270255208015442,
|
| 1427 |
+
"rewards/cosine_scaled_reward/mean": -0.10014239698648453,
|
| 1428 |
+
"rewards/cosine_scaled_reward/std": 0.42991697788238525,
|
| 1429 |
+
"rewards/format_reward/mean": 0.625,
|
| 1430 |
+
"rewards/format_reward/std": 0.48924607038497925,
|
| 1431 |
+
"step": 49
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"clip_ratio/high_max": 0.0,
|
| 1435 |
+
"clip_ratio/high_mean": 0.0,
|
| 1436 |
+
"clip_ratio/low_mean": 0.0,
|
| 1437 |
+
"clip_ratio/low_min": 0.0,
|
| 1438 |
+
"clip_ratio/region_mean": 0.0,
|
| 1439 |
+
"completions/clipped_ratio": 0.6875,
|
| 1440 |
+
"completions/max_length": 3584.0,
|
| 1441 |
+
"completions/max_terminated_length": 2808.0,
|
| 1442 |
+
"completions/mean_length": 2786.791748046875,
|
| 1443 |
+
"completions/mean_terminated_length": 1032.933349609375,
|
| 1444 |
+
"completions/min_length": 358.0,
|
| 1445 |
+
"completions/min_terminated_length": 358.0,
|
| 1446 |
+
"epoch": 0.05714285714285714,
|
| 1447 |
+
"frac_reward_zero_std": 0.0,
|
| 1448 |
+
"grad_norm": 0.7447382807731628,
|
| 1449 |
+
"kl": 0.001110076904296875,
|
| 1450 |
+
"learning_rate": 1.0109617738307911e-07,
|
| 1451 |
+
"loss": -0.2738,
|
| 1452 |
+
"num_tokens": 7046455.0,
|
| 1453 |
+
"policy_entropy_avg": 8.135416666666666,
|
| 1454 |
+
"reward": 0.2738919258117676,
|
| 1455 |
+
"reward_std": 0.44107958674430847,
|
| 1456 |
+
"rewards/cosine_scaled_reward/mean": 0.023186095058918,
|
| 1457 |
+
"rewards/cosine_scaled_reward/std": 0.3980216383934021,
|
| 1458 |
+
"rewards/format_reward/mean": 0.375,
|
| 1459 |
+
"rewards/format_reward/std": 0.48924607038497925,
|
| 1460 |
+
"step": 50
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 0.05714285714285714,
|
| 1464 |
+
"step": 50,
|
| 1465 |
+
"total_flos": 0.0,
|
| 1466 |
+
"train_loss": -0.45832799572497607,
|
| 1467 |
+
"train_runtime": 4441.2019,
|
| 1468 |
+
"train_samples_per_second": 0.54,
|
| 1469 |
+
"train_steps_per_second": 0.011
|
| 1470 |
+
}
|
| 1471 |
+
],
|
| 1472 |
+
"logging_steps": 1,
|
| 1473 |
+
"max_steps": 50,
|
| 1474 |
+
"num_input_tokens_seen": 7046455,
|
| 1475 |
+
"num_train_epochs": 1,
|
| 1476 |
+
"save_steps": 50,
|
| 1477 |
+
"stateful_callbacks": {
|
| 1478 |
+
"TrainerControl": {
|
| 1479 |
+
"args": {
|
| 1480 |
+
"should_epoch_stop": false,
|
| 1481 |
+
"should_evaluate": false,
|
| 1482 |
+
"should_log": false,
|
| 1483 |
+
"should_save": true,
|
| 1484 |
+
"should_training_stop": true
|
| 1485 |
+
},
|
| 1486 |
+
"attributes": {}
|
| 1487 |
+
}
|
| 1488 |
+
},
|
| 1489 |
+
"total_flos": 0.0,
|
| 1490 |
+
"train_batch_size": 4,
|
| 1491 |
+
"trial_name": null,
|
| 1492 |
+
"trial_params": null
|
| 1493 |
+
}
|