Model save
Browse files- README.md +68 -0
- all_results.json +8 -0
- config.json +1 -1
- generation_config.json +6 -0
- train_results.json +8 -0
- trainer_state.json +1667 -0
README.md
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: Qwen/Qwen2.5-Coder-3B
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: Qwen2.5-Coder-3B
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- grpo
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for Qwen2.5-Coder-3B
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-3B](https://huggingface.co/Qwen/Qwen2.5-Coder-3B).
|
| 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="zerozeroz/Qwen2.5-Coder-3B", 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 |
+
|
| 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: 0.14.0
|
| 38 |
+
- Transformers: 4.48.1
|
| 39 |
+
- Pytorch: 2.5.1+cu121
|
| 40 |
+
- Datasets: 3.1.0
|
| 41 |
+
- Tokenizers: 0.21.0
|
| 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": 1.9367338650191358e-05,
|
| 4 |
+
"train_runtime": 3648.0047,
|
| 5 |
+
"train_samples": 374,
|
| 6 |
+
"train_samples_per_second": 0.206,
|
| 7 |
+
"train_steps_per_second": 0.034
|
| 8 |
+
}
|
config.json
CHANGED
|
@@ -23,7 +23,7 @@
|
|
| 23 |
"tie_word_embeddings": true,
|
| 24 |
"torch_dtype": "bfloat16",
|
| 25 |
"transformers_version": "4.48.1",
|
| 26 |
-
"use_cache":
|
| 27 |
"use_sliding_window": false,
|
| 28 |
"vocab_size": 151936
|
| 29 |
}
|
|
|
|
| 23 |
"tie_word_embeddings": true,
|
| 24 |
"torch_dtype": "bfloat16",
|
| 25 |
"transformers_version": "4.48.1",
|
| 26 |
+
"use_cache": true,
|
| 27 |
"use_sliding_window": false,
|
| 28 |
"vocab_size": 151936
|
| 29 |
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"eos_token_id": 151643,
|
| 4 |
+
"max_new_tokens": 2048,
|
| 5 |
+
"transformers_version": "4.48.1"
|
| 6 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 1.9367338650191358e-05,
|
| 4 |
+
"train_runtime": 3648.0047,
|
| 5 |
+
"train_samples": 374,
|
| 6 |
+
"train_samples_per_second": 0.206,
|
| 7 |
+
"train_steps_per_second": 0.034
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1667 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.992,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 125,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"completion_length": 140.9583396911621,
|
| 13 |
+
"epoch": 0.016,
|
| 14 |
+
"grad_norm": 1.4258953228320013,
|
| 15 |
+
"kl": 0.0,
|
| 16 |
+
"learning_rate": 1.25e-07,
|
| 17 |
+
"loss": 0.0,
|
| 18 |
+
"reward": 0.5152640044689178,
|
| 19 |
+
"reward_std": 0.5508254170417786,
|
| 20 |
+
"rewards/correct_code_reward_func": 0.2291666716337204,
|
| 21 |
+
"rewards/len_reward_func": 0.28609737753868103,
|
| 22 |
+
"step": 1
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"completion_length": 131.50000762939453,
|
| 26 |
+
"epoch": 0.032,
|
| 27 |
+
"grad_norm": 1.1519258687122351,
|
| 28 |
+
"kl": 0.0,
|
| 29 |
+
"learning_rate": 2.5e-07,
|
| 30 |
+
"loss": 0.0,
|
| 31 |
+
"reward": 0.541226252913475,
|
| 32 |
+
"reward_std": 0.5189632624387741,
|
| 33 |
+
"rewards/correct_code_reward_func": 0.2500000111758709,
|
| 34 |
+
"rewards/len_reward_func": 0.29122625291347504,
|
| 35 |
+
"step": 2
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"completion_length": 108.83333587646484,
|
| 39 |
+
"epoch": 0.048,
|
| 40 |
+
"grad_norm": 1.467116667329371,
|
| 41 |
+
"kl": 0.00013637542724609375,
|
| 42 |
+
"learning_rate": 3.75e-07,
|
| 43 |
+
"loss": 0.0,
|
| 44 |
+
"reward": 0.7587994039058685,
|
| 45 |
+
"reward_std": 0.5140225142240524,
|
| 46 |
+
"rewards/correct_code_reward_func": 0.5416666865348816,
|
| 47 |
+
"rewards/len_reward_func": 0.21713273972272873,
|
| 48 |
+
"step": 3
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"completion_length": 159.81250762939453,
|
| 52 |
+
"epoch": 0.064,
|
| 53 |
+
"grad_norm": 1.2682485669137435,
|
| 54 |
+
"kl": 0.00018215179443359375,
|
| 55 |
+
"learning_rate": 5e-07,
|
| 56 |
+
"loss": 0.0,
|
| 57 |
+
"reward": 0.5178248882293701,
|
| 58 |
+
"reward_std": 0.4526914358139038,
|
| 59 |
+
"rewards/correct_code_reward_func": 0.1666666716337204,
|
| 60 |
+
"rewards/len_reward_func": 0.3511582016944885,
|
| 61 |
+
"step": 4
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"completion_length": 176.56250762939453,
|
| 65 |
+
"epoch": 0.08,
|
| 66 |
+
"grad_norm": 1.2033440109589688,
|
| 67 |
+
"kl": 0.00014066696166992188,
|
| 68 |
+
"learning_rate": 4.999157413258781e-07,
|
| 69 |
+
"loss": 0.0,
|
| 70 |
+
"reward": 0.32241350412368774,
|
| 71 |
+
"reward_std": 0.32281263172626495,
|
| 72 |
+
"rewards/correct_code_reward_func": 0.02083333395421505,
|
| 73 |
+
"rewards/len_reward_func": 0.30158019065856934,
|
| 74 |
+
"step": 5
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"completion_length": 124.87500762939453,
|
| 78 |
+
"epoch": 0.096,
|
| 79 |
+
"grad_norm": 1.5120707071506325,
|
| 80 |
+
"kl": 0.00016808509826660156,
|
| 81 |
+
"learning_rate": 4.996630220997057e-07,
|
| 82 |
+
"loss": 0.0,
|
| 83 |
+
"reward": 0.746085911989212,
|
| 84 |
+
"reward_std": 0.5452268123626709,
|
| 85 |
+
"rewards/correct_code_reward_func": 0.4583333432674408,
|
| 86 |
+
"rewards/len_reward_func": 0.28775252401828766,
|
| 87 |
+
"step": 6
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"completion_length": 169.9166717529297,
|
| 91 |
+
"epoch": 0.112,
|
| 92 |
+
"grad_norm": 0.9079518632617903,
|
| 93 |
+
"kl": 0.00011348724365234375,
|
| 94 |
+
"learning_rate": 4.992420126717784e-07,
|
| 95 |
+
"loss": 0.0,
|
| 96 |
+
"reward": 0.36989694088697433,
|
| 97 |
+
"reward_std": 0.45903605222702026,
|
| 98 |
+
"rewards/correct_code_reward_func": 0.125,
|
| 99 |
+
"rewards/len_reward_func": 0.24489693343639374,
|
| 100 |
+
"step": 7
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"completion_length": 219.43750762939453,
|
| 104 |
+
"epoch": 0.128,
|
| 105 |
+
"grad_norm": 1.2633142753352289,
|
| 106 |
+
"kl": 0.0002155303955078125,
|
| 107 |
+
"learning_rate": 4.986529968316653e-07,
|
| 108 |
+
"loss": 0.0,
|
| 109 |
+
"reward": 0.44794920086860657,
|
| 110 |
+
"reward_std": 0.385338693857193,
|
| 111 |
+
"rewards/correct_code_reward_func": 0.1250000037252903,
|
| 112 |
+
"rewards/len_reward_func": 0.3229491859674454,
|
| 113 |
+
"step": 8
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"completion_length": 227.91667938232422,
|
| 117 |
+
"epoch": 0.144,
|
| 118 |
+
"grad_norm": 1.0211344567101885,
|
| 119 |
+
"kl": 0.00011777877807617188,
|
| 120 |
+
"learning_rate": 4.978963716169165e-07,
|
| 121 |
+
"loss": 0.0,
|
| 122 |
+
"reward": 0.6235890090465546,
|
| 123 |
+
"reward_std": 0.5187947303056717,
|
| 124 |
+
"rewards/correct_code_reward_func": 0.3125,
|
| 125 |
+
"rewards/len_reward_func": 0.31108900904655457,
|
| 126 |
+
"step": 9
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"completion_length": 188.25000762939453,
|
| 130 |
+
"epoch": 0.16,
|
| 131 |
+
"grad_norm": 1.0353822839723037,
|
| 132 |
+
"kl": 0.00011730194091796875,
|
| 133 |
+
"learning_rate": 4.969726470454313e-07,
|
| 134 |
+
"loss": 0.0,
|
| 135 |
+
"reward": 0.6911160051822662,
|
| 136 |
+
"reward_std": 0.5456923246383667,
|
| 137 |
+
"rewards/correct_code_reward_func": 0.4166666865348816,
|
| 138 |
+
"rewards/len_reward_func": 0.27444930374622345,
|
| 139 |
+
"step": 10
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"completion_length": 168.27083587646484,
|
| 143 |
+
"epoch": 0.176,
|
| 144 |
+
"grad_norm": 1.7856755608823207,
|
| 145 |
+
"kl": 0.00018310546875,
|
| 146 |
+
"learning_rate": 4.958824457716706e-07,
|
| 147 |
+
"loss": 0.0,
|
| 148 |
+
"reward": 0.4588584154844284,
|
| 149 |
+
"reward_std": 0.40716809034347534,
|
| 150 |
+
"rewards/correct_code_reward_func": 0.1875,
|
| 151 |
+
"rewards/len_reward_func": 0.271358385682106,
|
| 152 |
+
"step": 11
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"completion_length": 203.08333587646484,
|
| 156 |
+
"epoch": 0.192,
|
| 157 |
+
"grad_norm": 0.9296992149271633,
|
| 158 |
+
"kl": 0.00016641616821289062,
|
| 159 |
+
"learning_rate": 4.946265026669454e-07,
|
| 160 |
+
"loss": 0.0,
|
| 161 |
+
"reward": 0.3501324951648712,
|
| 162 |
+
"reward_std": 0.49003708362579346,
|
| 163 |
+
"rewards/correct_code_reward_func": 0.1041666679084301,
|
| 164 |
+
"rewards/len_reward_func": 0.245965838432312,
|
| 165 |
+
"step": 12
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"completion_length": 115.66666793823242,
|
| 169 |
+
"epoch": 0.208,
|
| 170 |
+
"grad_norm": 1.4335533212366607,
|
| 171 |
+
"kl": 0.00016570091247558594,
|
| 172 |
+
"learning_rate": 4.932056643240618e-07,
|
| 173 |
+
"loss": 0.0,
|
| 174 |
+
"reward": 0.7853705883026123,
|
| 175 |
+
"reward_std": 0.46111349761486053,
|
| 176 |
+
"rewards/correct_code_reward_func": 0.5000000149011612,
|
| 177 |
+
"rewards/len_reward_func": 0.2853705883026123,
|
| 178 |
+
"step": 13
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"completion_length": 169.95833587646484,
|
| 182 |
+
"epoch": 0.224,
|
| 183 |
+
"grad_norm": 1.2723280538596287,
|
| 184 |
+
"kl": 0.00021076202392578125,
|
| 185 |
+
"learning_rate": 4.916208884866592e-07,
|
| 186 |
+
"loss": 0.0,
|
| 187 |
+
"reward": 0.5324039310216904,
|
| 188 |
+
"reward_std": 0.5338821411132812,
|
| 189 |
+
"rewards/correct_code_reward_func": 0.2708333432674408,
|
| 190 |
+
"rewards/len_reward_func": 0.26157061755657196,
|
| 191 |
+
"step": 14
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"completion_length": 154.58333587646484,
|
| 195 |
+
"epoch": 0.24,
|
| 196 |
+
"grad_norm": 1.2578666329332273,
|
| 197 |
+
"kl": 0.00019168853759765625,
|
| 198 |
+
"learning_rate": 4.898732434036243e-07,
|
| 199 |
+
"loss": 0.0,
|
| 200 |
+
"reward": 0.5949100255966187,
|
| 201 |
+
"reward_std": 0.5048613250255585,
|
| 202 |
+
"rewards/correct_code_reward_func": 0.3125000149011612,
|
| 203 |
+
"rewards/len_reward_func": 0.28241002559661865,
|
| 204 |
+
"step": 15
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"completion_length": 173.1875114440918,
|
| 208 |
+
"epoch": 0.256,
|
| 209 |
+
"grad_norm": 1.1230347862341579,
|
| 210 |
+
"kl": 0.00029277801513671875,
|
| 211 |
+
"learning_rate": 4.879639071090173e-07,
|
| 212 |
+
"loss": 0.0,
|
| 213 |
+
"reward": 0.4564344882965088,
|
| 214 |
+
"reward_std": 0.4671656936407089,
|
| 215 |
+
"rewards/correct_code_reward_func": 0.1666666679084301,
|
| 216 |
+
"rewards/len_reward_func": 0.2897678166627884,
|
| 217 |
+
"step": 16
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"completion_length": 169.375,
|
| 221 |
+
"epoch": 0.272,
|
| 222 |
+
"grad_norm": 1.3041956300758726,
|
| 223 |
+
"kl": 0.0002574920654296875,
|
| 224 |
+
"learning_rate": 4.858941666279955e-07,
|
| 225 |
+
"loss": 0.0,
|
| 226 |
+
"reward": 0.6347246468067169,
|
| 227 |
+
"reward_std": 0.5289804339408875,
|
| 228 |
+
"rewards/correct_code_reward_func": 0.3541666716337204,
|
| 229 |
+
"rewards/len_reward_func": 0.2805579602718353,
|
| 230 |
+
"step": 17
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"completion_length": 133.25000762939453,
|
| 234 |
+
"epoch": 0.288,
|
| 235 |
+
"grad_norm": 1.354822217310785,
|
| 236 |
+
"kl": 0.0002689361572265625,
|
| 237 |
+
"learning_rate": 4.836654171092682e-07,
|
| 238 |
+
"loss": 0.0,
|
| 239 |
+
"reward": 0.5779364109039307,
|
| 240 |
+
"reward_std": 0.4782462567090988,
|
| 241 |
+
"rewards/correct_code_reward_func": 0.2916666716337204,
|
| 242 |
+
"rewards/len_reward_func": 0.2862697243690491,
|
| 243 |
+
"step": 18
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"completion_length": 99.41667175292969,
|
| 247 |
+
"epoch": 0.304,
|
| 248 |
+
"grad_norm": 1.4087777232916079,
|
| 249 |
+
"kl": 0.00031757354736328125,
|
| 250 |
+
"learning_rate": 4.812791608846709e-07,
|
| 251 |
+
"loss": 0.0,
|
| 252 |
+
"reward": 0.5035808980464935,
|
| 253 |
+
"reward_std": 0.46289560198783875,
|
| 254 |
+
"rewards/correct_code_reward_func": 0.229166679084301,
|
| 255 |
+
"rewards/len_reward_func": 0.27441420406103134,
|
| 256 |
+
"step": 19
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"completion_length": 170.7291717529297,
|
| 260 |
+
"epoch": 0.32,
|
| 261 |
+
"grad_norm": 0.9923230664440412,
|
| 262 |
+
"kl": 0.00028705596923828125,
|
| 263 |
+
"learning_rate": 4.787370064564882e-07,
|
| 264 |
+
"loss": 0.0,
|
| 265 |
+
"reward": 0.5567075908184052,
|
| 266 |
+
"reward_std": 0.44439028203487396,
|
| 267 |
+
"rewards/correct_code_reward_func": 0.2083333432674408,
|
| 268 |
+
"rewards/len_reward_func": 0.34837424755096436,
|
| 269 |
+
"step": 20
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"completion_length": 124.72917175292969,
|
| 273 |
+
"epoch": 0.336,
|
| 274 |
+
"grad_norm": 1.2245791922735345,
|
| 275 |
+
"kl": 0.00035572052001953125,
|
| 276 |
+
"learning_rate": 4.7604066741321253e-07,
|
| 277 |
+
"loss": 0.0,
|
| 278 |
+
"reward": 0.8560027182102203,
|
| 279 |
+
"reward_std": 0.6356588900089264,
|
| 280 |
+
"rewards/correct_code_reward_func": 0.5416666865348816,
|
| 281 |
+
"rewards/len_reward_func": 0.31433598697185516,
|
| 282 |
+
"step": 21
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"completion_length": 123.64583969116211,
|
| 286 |
+
"epoch": 0.352,
|
| 287 |
+
"grad_norm": 1.2080469812565267,
|
| 288 |
+
"kl": 0.00035858154296875,
|
| 289 |
+
"learning_rate": 4.731919612744659e-07,
|
| 290 |
+
"loss": 0.0,
|
| 291 |
+
"reward": 0.7242447733879089,
|
| 292 |
+
"reward_std": 0.4742405414581299,
|
| 293 |
+
"rewards/correct_code_reward_func": 0.3958333432674408,
|
| 294 |
+
"rewards/len_reward_func": 0.32841143012046814,
|
| 295 |
+
"step": 22
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"completion_length": 146.2916717529297,
|
| 299 |
+
"epoch": 0.368,
|
| 300 |
+
"grad_norm": 1.2440640880474592,
|
| 301 |
+
"kl": 0.00040721893310546875,
|
| 302 |
+
"learning_rate": 4.7019280826586604e-07,
|
| 303 |
+
"loss": 0.0,
|
| 304 |
+
"reward": 0.5270938575267792,
|
| 305 |
+
"reward_std": 0.4260385036468506,
|
| 306 |
+
"rewards/correct_code_reward_func": 0.2291666679084301,
|
| 307 |
+
"rewards/len_reward_func": 0.2979271858930588,
|
| 308 |
+
"step": 23
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"completion_length": 141.9166717529297,
|
| 312 |
+
"epoch": 0.384,
|
| 313 |
+
"grad_norm": 1.455943571941334,
|
| 314 |
+
"kl": 0.0006427764892578125,
|
| 315 |
+
"learning_rate": 4.6704523002466094e-07,
|
| 316 |
+
"loss": 0.0,
|
| 317 |
+
"reward": 0.5917265266180038,
|
| 318 |
+
"reward_std": 0.47722122073173523,
|
| 319 |
+
"rewards/correct_code_reward_func": 0.3333333358168602,
|
| 320 |
+
"rewards/len_reward_func": 0.25839313119649887,
|
| 321 |
+
"step": 24
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"completion_length": 240.85417938232422,
|
| 325 |
+
"epoch": 0.4,
|
| 326 |
+
"grad_norm": 0.8411889507435418,
|
| 327 |
+
"kl": 0.0003604888916015625,
|
| 328 |
+
"learning_rate": 4.6375134823700503e-07,
|
| 329 |
+
"loss": 0.0,
|
| 330 |
+
"reward": 0.3353981524705887,
|
| 331 |
+
"reward_std": 0.351834774017334,
|
| 332 |
+
"rewards/correct_code_reward_func": 0.0833333358168602,
|
| 333 |
+
"rewards/len_reward_func": 0.2520648390054703,
|
| 334 |
+
"step": 25
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"completion_length": 97.31250381469727,
|
| 338 |
+
"epoch": 0.416,
|
| 339 |
+
"grad_norm": 1.374585753278975,
|
| 340 |
+
"kl": 0.0008258819580078125,
|
| 341 |
+
"learning_rate": 4.603133832077953e-07,
|
| 342 |
+
"loss": 0.0,
|
| 343 |
+
"reward": 0.6881800889968872,
|
| 344 |
+
"reward_std": 0.5626422464847565,
|
| 345 |
+
"rewards/correct_code_reward_func": 0.4375,
|
| 346 |
+
"rewards/len_reward_func": 0.2506800442934036,
|
| 347 |
+
"step": 26
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"completion_length": 131.08333587646484,
|
| 351 |
+
"epoch": 0.432,
|
| 352 |
+
"grad_norm": 1.5040369557196518,
|
| 353 |
+
"kl": 0.0006847381591796875,
|
| 354 |
+
"learning_rate": 4.5673365236403216e-07,
|
| 355 |
+
"loss": 0.0,
|
| 356 |
+
"reward": 0.6470239758491516,
|
| 357 |
+
"reward_std": 0.39606642723083496,
|
| 358 |
+
"rewards/correct_code_reward_func": 0.4375,
|
| 359 |
+
"rewards/len_reward_func": 0.20952393114566803,
|
| 360 |
+
"step": 27
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"completion_length": 198.06250762939453,
|
| 364 |
+
"epoch": 0.448,
|
| 365 |
+
"grad_norm": 1.1110007536297855,
|
| 366 |
+
"kl": 0.00054168701171875,
|
| 367 |
+
"learning_rate": 4.530145686927125e-07,
|
| 368 |
+
"loss": 0.0,
|
| 369 |
+
"reward": 0.5166794955730438,
|
| 370 |
+
"reward_std": 0.504486620426178,
|
| 371 |
+
"rewards/correct_code_reward_func": 0.2500000149011612,
|
| 372 |
+
"rewards/len_reward_func": 0.2666794955730438,
|
| 373 |
+
"step": 28
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"completion_length": 152.52083587646484,
|
| 377 |
+
"epoch": 0.464,
|
| 378 |
+
"grad_norm": 1.134262039216797,
|
| 379 |
+
"kl": 0.00078582763671875,
|
| 380 |
+
"learning_rate": 4.4915863911430897e-07,
|
| 381 |
+
"loss": 0.0,
|
| 382 |
+
"reward": 0.5144253522157669,
|
| 383 |
+
"reward_std": 0.4733017832040787,
|
| 384 |
+
"rewards/correct_code_reward_func": 0.1875000111758709,
|
| 385 |
+
"rewards/len_reward_func": 0.3269253224134445,
|
| 386 |
+
"step": 29
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"completion_length": 139.7916717529297,
|
| 390 |
+
"epoch": 0.48,
|
| 391 |
+
"grad_norm": 1.010573889887009,
|
| 392 |
+
"kl": 0.0007152557373046875,
|
| 393 |
+
"learning_rate": 4.45168462792932e-07,
|
| 394 |
+
"loss": 0.0,
|
| 395 |
+
"reward": 0.5882390439510345,
|
| 396 |
+
"reward_std": 0.43310636281967163,
|
| 397 |
+
"rewards/correct_code_reward_func": 0.2500000074505806,
|
| 398 |
+
"rewards/len_reward_func": 0.33823904395103455,
|
| 399 |
+
"step": 30
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"completion_length": 87.41666793823242,
|
| 403 |
+
"epoch": 0.496,
|
| 404 |
+
"grad_norm": 1.540244950569226,
|
| 405 |
+
"kl": 0.0012340545654296875,
|
| 406 |
+
"learning_rate": 4.4104672938431223e-07,
|
| 407 |
+
"loss": 0.0,
|
| 408 |
+
"reward": 0.7711681425571442,
|
| 409 |
+
"reward_std": 0.4805651605129242,
|
| 410 |
+
"rewards/correct_code_reward_func": 0.5833333432674408,
|
| 411 |
+
"rewards/len_reward_func": 0.18783476203680038,
|
| 412 |
+
"step": 31
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"completion_length": 101.43750381469727,
|
| 416 |
+
"epoch": 0.512,
|
| 417 |
+
"grad_norm": 2.3673085026520297,
|
| 418 |
+
"kl": 0.0012607574462890625,
|
| 419 |
+
"learning_rate": 4.367962172227866e-07,
|
| 420 |
+
"loss": 0.0,
|
| 421 |
+
"reward": 0.7279457449913025,
|
| 422 |
+
"reward_std": 0.4627054035663605,
|
| 423 |
+
"rewards/correct_code_reward_func": 0.4583333432674408,
|
| 424 |
+
"rewards/len_reward_func": 0.2696124166250229,
|
| 425 |
+
"step": 32
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"completion_length": 155.2291717529297,
|
| 429 |
+
"epoch": 0.528,
|
| 430 |
+
"grad_norm": 1.2624598609488873,
|
| 431 |
+
"kl": 0.00139617919921875,
|
| 432 |
+
"learning_rate": 4.324197914485075e-07,
|
| 433 |
+
"loss": 0.0,
|
| 434 |
+
"reward": 0.6401492655277252,
|
| 435 |
+
"reward_std": 0.515736848115921,
|
| 436 |
+
"rewards/correct_code_reward_func": 0.375,
|
| 437 |
+
"rewards/len_reward_func": 0.26514923572540283,
|
| 438 |
+
"step": 33
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"completion_length": 252.91667938232422,
|
| 442 |
+
"epoch": 0.544,
|
| 443 |
+
"grad_norm": 1.043728438493038,
|
| 444 |
+
"kl": 0.0008392333984375,
|
| 445 |
+
"learning_rate": 4.2792040207614e-07,
|
| 446 |
+
"loss": 0.0,
|
| 447 |
+
"reward": 0.6339870393276215,
|
| 448 |
+
"reward_std": 0.5688490867614746,
|
| 449 |
+
"rewards/correct_code_reward_func": 0.3333333432674408,
|
| 450 |
+
"rewards/len_reward_func": 0.30065372586250305,
|
| 451 |
+
"step": 34
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"completion_length": 178.25,
|
| 455 |
+
"epoch": 0.56,
|
| 456 |
+
"grad_norm": 1.2442169258805433,
|
| 457 |
+
"kl": 0.00205230712890625,
|
| 458 |
+
"learning_rate": 4.2330108200634723e-07,
|
| 459 |
+
"loss": 0.0,
|
| 460 |
+
"reward": 0.43357332795858383,
|
| 461 |
+
"reward_std": 0.3690243661403656,
|
| 462 |
+
"rewards/correct_code_reward_func": 0.16666667722165585,
|
| 463 |
+
"rewards/len_reward_func": 0.26690666377544403,
|
| 464 |
+
"step": 35
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"completion_length": 150.1666717529297,
|
| 468 |
+
"epoch": 0.576,
|
| 469 |
+
"grad_norm": 1.0937981889230137,
|
| 470 |
+
"kl": 0.0016021728515625,
|
| 471 |
+
"learning_rate": 4.185649449814045e-07,
|
| 472 |
+
"loss": 0.0,
|
| 473 |
+
"reward": 0.8725252151489258,
|
| 474 |
+
"reward_std": 0.5368492603302002,
|
| 475 |
+
"rewards/correct_code_reward_func": 0.5416666865348816,
|
| 476 |
+
"rewards/len_reward_func": 0.3308584541082382,
|
| 477 |
+
"step": 36
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"completion_length": 74.41666793823242,
|
| 481 |
+
"epoch": 0.592,
|
| 482 |
+
"grad_norm": 1.4560552034278569,
|
| 483 |
+
"kl": 0.0020904541015625,
|
| 484 |
+
"learning_rate": 4.137151834863213e-07,
|
| 485 |
+
"loss": 0.0,
|
| 486 |
+
"reward": 0.7634576857089996,
|
| 487 |
+
"reward_std": 0.5292592346668243,
|
| 488 |
+
"rewards/correct_code_reward_func": 0.5416666716337204,
|
| 489 |
+
"rewards/len_reward_func": 0.22179099917411804,
|
| 490 |
+
"step": 37
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"completion_length": 111.77083587646484,
|
| 494 |
+
"epoch": 0.608,
|
| 495 |
+
"grad_norm": 1.6125607277054597,
|
| 496 |
+
"kl": 0.002716064453125,
|
| 497 |
+
"learning_rate": 4.087550665968846e-07,
|
| 498 |
+
"loss": 0.0,
|
| 499 |
+
"reward": 0.6047167330980301,
|
| 500 |
+
"reward_std": 0.4415762424468994,
|
| 501 |
+
"rewards/correct_code_reward_func": 0.2916666865348816,
|
| 502 |
+
"rewards/len_reward_func": 0.3130500763654709,
|
| 503 |
+
"step": 38
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"completion_length": 87.0625,
|
| 507 |
+
"epoch": 0.624,
|
| 508 |
+
"grad_norm": 2.0747921723056026,
|
| 509 |
+
"kl": 0.0023193359375,
|
| 510 |
+
"learning_rate": 4.036879377760752e-07,
|
| 511 |
+
"loss": 0.0,
|
| 512 |
+
"reward": 0.7261738479137421,
|
| 513 |
+
"reward_std": 0.6433705389499664,
|
| 514 |
+
"rewards/correct_code_reward_func": 0.520833358168602,
|
| 515 |
+
"rewards/len_reward_func": 0.20534051209688187,
|
| 516 |
+
"step": 39
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"completion_length": 128.0833396911621,
|
| 520 |
+
"epoch": 0.64,
|
| 521 |
+
"grad_norm": 1.352520841789316,
|
| 522 |
+
"kl": 0.00229644775390625,
|
| 523 |
+
"learning_rate": 3.9851721262034157e-07,
|
| 524 |
+
"loss": 0.0,
|
| 525 |
+
"reward": 0.49166351556777954,
|
| 526 |
+
"reward_std": 0.4290030002593994,
|
| 527 |
+
"rewards/correct_code_reward_func": 0.18750000558793545,
|
| 528 |
+
"rewards/len_reward_func": 0.30416350066661835,
|
| 529 |
+
"step": 40
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"completion_length": 117.33333587646484,
|
| 533 |
+
"epoch": 0.656,
|
| 534 |
+
"grad_norm": 1.5281074207353524,
|
| 535 |
+
"kl": 0.003509521484375,
|
| 536 |
+
"learning_rate": 3.932463765572505e-07,
|
| 537 |
+
"loss": 0.0,
|
| 538 |
+
"reward": 0.5800679922103882,
|
| 539 |
+
"reward_std": 0.5416670143604279,
|
| 540 |
+
"rewards/correct_code_reward_func": 0.3125000149011612,
|
| 541 |
+
"rewards/len_reward_func": 0.2675679475069046,
|
| 542 |
+
"step": 41
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"completion_length": 112.43750381469727,
|
| 546 |
+
"epoch": 0.672,
|
| 547 |
+
"grad_norm": 1.2084984435618142,
|
| 548 |
+
"kl": 0.00252532958984375,
|
| 549 |
+
"learning_rate": 3.8787898249606767e-07,
|
| 550 |
+
"loss": 0.0,
|
| 551 |
+
"reward": 0.42490366101264954,
|
| 552 |
+
"reward_std": 0.46323399245738983,
|
| 553 |
+
"rewards/correct_code_reward_func": 0.14583333395421505,
|
| 554 |
+
"rewards/len_reward_func": 0.27907034754753113,
|
| 555 |
+
"step": 42
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"completion_length": 56.85416793823242,
|
| 559 |
+
"epoch": 0.688,
|
| 560 |
+
"grad_norm": 1.8756323954488632,
|
| 561 |
+
"kl": 0.00452423095703125,
|
| 562 |
+
"learning_rate": 3.8241864843284964e-07,
|
| 563 |
+
"loss": 0.0,
|
| 564 |
+
"reward": 0.7274035811424255,
|
| 565 |
+
"reward_std": 0.5209662765264511,
|
| 566 |
+
"rewards/correct_code_reward_func": 0.5000000149011612,
|
| 567 |
+
"rewards/len_reward_func": 0.22740358859300613,
|
| 568 |
+
"step": 43
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"completion_length": 153.68750762939453,
|
| 572 |
+
"epoch": 0.704,
|
| 573 |
+
"grad_norm": 1.785627080388602,
|
| 574 |
+
"kl": 0.0055084228515625,
|
| 575 |
+
"learning_rate": 3.768690550116639e-07,
|
| 576 |
+
"loss": 0.0,
|
| 577 |
+
"reward": 0.49254634976387024,
|
| 578 |
+
"reward_std": 0.4052678644657135,
|
| 579 |
+
"rewards/correct_code_reward_func": 0.1666666716337204,
|
| 580 |
+
"rewards/len_reward_func": 0.32587967813014984,
|
| 581 |
+
"step": 44
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"completion_length": 170.1041717529297,
|
| 585 |
+
"epoch": 0.72,
|
| 586 |
+
"grad_norm": 1.2057879792669277,
|
| 587 |
+
"kl": 0.0038299560546875,
|
| 588 |
+
"learning_rate": 3.712339430435792e-07,
|
| 589 |
+
"loss": 0.0,
|
| 590 |
+
"reward": 0.5373264253139496,
|
| 591 |
+
"reward_std": 0.4612013250589371,
|
| 592 |
+
"rewards/correct_code_reward_func": 0.2708333432674408,
|
| 593 |
+
"rewards/len_reward_func": 0.2664930745959282,
|
| 594 |
+
"step": 45
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"completion_length": 122.79167175292969,
|
| 598 |
+
"epoch": 0.736,
|
| 599 |
+
"grad_norm": 1.23844328247912,
|
| 600 |
+
"kl": 0.00384521484375,
|
| 601 |
+
"learning_rate": 3.65517110985099e-07,
|
| 602 |
+
"loss": 0.0,
|
| 603 |
+
"reward": 0.6534424722194672,
|
| 604 |
+
"reward_std": 0.5896010398864746,
|
| 605 |
+
"rewards/correct_code_reward_func": 0.354166679084301,
|
| 606 |
+
"rewards/len_reward_func": 0.29927581548690796,
|
| 607 |
+
"step": 46
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"completion_length": 73.39583396911621,
|
| 611 |
+
"epoch": 0.752,
|
| 612 |
+
"grad_norm": 2.222315006145743,
|
| 613 |
+
"kl": 0.0058135986328125,
|
| 614 |
+
"learning_rate": 3.597224123777389e-07,
|
| 615 |
+
"loss": 0.0,
|
| 616 |
+
"reward": 0.7357015609741211,
|
| 617 |
+
"reward_std": 0.5119403451681137,
|
| 618 |
+
"rewards/correct_code_reward_func": 0.4583333432674408,
|
| 619 |
+
"rewards/len_reward_func": 0.2773682177066803,
|
| 620 |
+
"step": 47
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"completion_length": 75.54166793823242,
|
| 624 |
+
"epoch": 0.768,
|
| 625 |
+
"grad_norm": 1.9981519435567456,
|
| 626 |
+
"kl": 0.0053863525390625,
|
| 627 |
+
"learning_rate": 3.5385375325047163e-07,
|
| 628 |
+
"loss": 0.0,
|
| 629 |
+
"reward": 0.6428782939910889,
|
| 630 |
+
"reward_std": 0.6202229559421539,
|
| 631 |
+
"rewards/correct_code_reward_func": 0.3958333432674408,
|
| 632 |
+
"rewards/len_reward_func": 0.24704494327306747,
|
| 633 |
+
"step": 48
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"completion_length": 73.27083587646484,
|
| 637 |
+
"epoch": 0.784,
|
| 638 |
+
"grad_norm": 2.073070842958071,
|
| 639 |
+
"kl": 0.00554656982421875,
|
| 640 |
+
"learning_rate": 3.479150894867926e-07,
|
| 641 |
+
"loss": 0.0,
|
| 642 |
+
"reward": 0.8005061745643616,
|
| 643 |
+
"reward_std": 0.5489170849323273,
|
| 644 |
+
"rewards/correct_code_reward_func": 0.5416666865348816,
|
| 645 |
+
"rewards/len_reward_func": 0.25883948802948,
|
| 646 |
+
"step": 49
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"completion_length": 93.62500381469727,
|
| 650 |
+
"epoch": 0.8,
|
| 651 |
+
"grad_norm": 1.7280406240103203,
|
| 652 |
+
"kl": 0.0070953369140625,
|
| 653 |
+
"learning_rate": 3.4191042415818e-07,
|
| 654 |
+
"loss": 0.0,
|
| 655 |
+
"reward": 0.6382943987846375,
|
| 656 |
+
"reward_std": 0.4014574736356735,
|
| 657 |
+
"rewards/correct_code_reward_func": 0.3750000149011612,
|
| 658 |
+
"rewards/len_reward_func": 0.26329439133405685,
|
| 659 |
+
"step": 50
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"completion_length": 110.31250381469727,
|
| 663 |
+
"epoch": 0.816,
|
| 664 |
+
"grad_norm": 1.5732703630042588,
|
| 665 |
+
"kl": 0.008453369140625,
|
| 666 |
+
"learning_rate": 3.3584380482574717e-07,
|
| 667 |
+
"loss": 0.0,
|
| 668 |
+
"reward": 0.8389279842376709,
|
| 669 |
+
"reward_std": 0.6495693922042847,
|
| 670 |
+
"rewards/correct_code_reward_func": 0.5208333432674408,
|
| 671 |
+
"rewards/len_reward_func": 0.31809471547603607,
|
| 672 |
+
"step": 51
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"completion_length": 81.4375,
|
| 676 |
+
"epoch": 0.832,
|
| 677 |
+
"grad_norm": 1.3555162901411408,
|
| 678 |
+
"kl": 0.0072479248046875,
|
| 679 |
+
"learning_rate": 3.297193208119047e-07,
|
| 680 |
+
"loss": 0.0,
|
| 681 |
+
"reward": 0.7050519585609436,
|
| 682 |
+
"reward_std": 0.522288054227829,
|
| 683 |
+
"rewards/correct_code_reward_func": 0.4375000298023224,
|
| 684 |
+
"rewards/len_reward_func": 0.2675519585609436,
|
| 685 |
+
"step": 52
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"completion_length": 145.2291717529297,
|
| 689 |
+
"epoch": 0.848,
|
| 690 |
+
"grad_norm": 1.2256688073258564,
|
| 691 |
+
"kl": 0.00726318359375,
|
| 692 |
+
"learning_rate": 3.235411004438741e-07,
|
| 693 |
+
"loss": 0.0,
|
| 694 |
+
"reward": 0.6400169730186462,
|
| 695 |
+
"reward_std": 0.5816708207130432,
|
| 696 |
+
"rewards/correct_code_reward_func": 0.3541666716337204,
|
| 697 |
+
"rewards/len_reward_func": 0.28585030883550644,
|
| 698 |
+
"step": 53
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"completion_length": 120.20833587646484,
|
| 702 |
+
"epoch": 0.864,
|
| 703 |
+
"grad_norm": 1.8462631631415796,
|
| 704 |
+
"kl": 0.0084991455078125,
|
| 705 |
+
"learning_rate": 3.173133082709086e-07,
|
| 706 |
+
"loss": 0.0,
|
| 707 |
+
"reward": 0.643402487039566,
|
| 708 |
+
"reward_std": 0.3417808264493942,
|
| 709 |
+
"rewards/correct_code_reward_func": 0.3333333432674408,
|
| 710 |
+
"rewards/len_reward_func": 0.31006917357444763,
|
| 711 |
+
"step": 54
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"completion_length": 55.56250190734863,
|
| 715 |
+
"epoch": 0.88,
|
| 716 |
+
"grad_norm": 1.7370166581779802,
|
| 717 |
+
"kl": 0.01177978515625,
|
| 718 |
+
"learning_rate": 3.1104014225709784e-07,
|
| 719 |
+
"loss": 0.0,
|
| 720 |
+
"reward": 0.9137917459011078,
|
| 721 |
+
"reward_std": 0.5003669559955597,
|
| 722 |
+
"rewards/correct_code_reward_func": 0.583333358168602,
|
| 723 |
+
"rewards/len_reward_func": 0.3304583728313446,
|
| 724 |
+
"step": 55
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"completion_length": 189.25000762939453,
|
| 728 |
+
"epoch": 0.896,
|
| 729 |
+
"grad_norm": 1.2196760565152192,
|
| 730 |
+
"kl": 0.0058441162109375,
|
| 731 |
+
"learning_rate": 3.0472583095164873e-07,
|
| 732 |
+
"loss": 0.0,
|
| 733 |
+
"reward": 0.4673280417919159,
|
| 734 |
+
"reward_std": 0.4577627182006836,
|
| 735 |
+
"rewards/correct_code_reward_func": 0.1666666716337204,
|
| 736 |
+
"rewards/len_reward_func": 0.3006613999605179,
|
| 737 |
+
"step": 56
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"completion_length": 57.37500190734863,
|
| 741 |
+
"epoch": 0.912,
|
| 742 |
+
"grad_norm": 2.0919947468048976,
|
| 743 |
+
"kl": 0.010162353515625,
|
| 744 |
+
"learning_rate": 2.983746306385499e-07,
|
| 745 |
+
"loss": 0.0,
|
| 746 |
+
"reward": 0.6931174695491791,
|
| 747 |
+
"reward_std": 0.5172313153743744,
|
| 748 |
+
"rewards/correct_code_reward_func": 0.4791666865348816,
|
| 749 |
+
"rewards/len_reward_func": 0.21395081281661987,
|
| 750 |
+
"step": 57
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"completion_length": 86.00000190734863,
|
| 754 |
+
"epoch": 0.928,
|
| 755 |
+
"grad_norm": 1.5907089477428527,
|
| 756 |
+
"kl": 0.0113677978515625,
|
| 757 |
+
"learning_rate": 2.919908224675412e-07,
|
| 758 |
+
"loss": 0.0,
|
| 759 |
+
"reward": 0.5865814685821533,
|
| 760 |
+
"reward_std": 0.5177368223667145,
|
| 761 |
+
"rewards/correct_code_reward_func": 0.3125000149011612,
|
| 762 |
+
"rewards/len_reward_func": 0.27408143877983093,
|
| 763 |
+
"step": 58
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"completion_length": 90.72916793823242,
|
| 767 |
+
"epoch": 0.944,
|
| 768 |
+
"grad_norm": 1.1269292807249032,
|
| 769 |
+
"kl": 0.00830078125,
|
| 770 |
+
"learning_rate": 2.8557870956832133e-07,
|
| 771 |
+
"loss": 0.0,
|
| 772 |
+
"reward": 0.4935041069984436,
|
| 773 |
+
"reward_std": 0.41843119263648987,
|
| 774 |
+
"rewards/correct_code_reward_func": 0.2083333432674408,
|
| 775 |
+
"rewards/len_reward_func": 0.285170778632164,
|
| 776 |
+
"step": 59
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"completion_length": 85.60416793823242,
|
| 780 |
+
"epoch": 0.96,
|
| 781 |
+
"grad_norm": 2.320388470663489,
|
| 782 |
+
"kl": 0.014678955078125,
|
| 783 |
+
"learning_rate": 2.7914261414993976e-07,
|
| 784 |
+
"loss": 0.0,
|
| 785 |
+
"reward": 0.7554058134555817,
|
| 786 |
+
"reward_std": 0.5069911777973175,
|
| 787 |
+
"rewards/correct_code_reward_func": 0.4166666716337204,
|
| 788 |
+
"rewards/len_reward_func": 0.3387391269207001,
|
| 789 |
+
"step": 60
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"completion_length": 63.375,
|
| 793 |
+
"epoch": 0.976,
|
| 794 |
+
"grad_norm": 1.7319214973496064,
|
| 795 |
+
"kl": 0.02532958984375,
|
| 796 |
+
"learning_rate": 2.726868745873286e-07,
|
| 797 |
+
"loss": 0.0,
|
| 798 |
+
"reward": 0.7839343547821045,
|
| 799 |
+
"reward_std": 0.6209487617015839,
|
| 800 |
+
"rewards/correct_code_reward_func": 0.4791666716337204,
|
| 801 |
+
"rewards/len_reward_func": 0.3047676384449005,
|
| 802 |
+
"step": 61
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"completion_length": 87.14583587646484,
|
| 806 |
+
"epoch": 0.992,
|
| 807 |
+
"grad_norm": 1.8272498546531741,
|
| 808 |
+
"kl": 0.0134735107421875,
|
| 809 |
+
"learning_rate": 2.662158424969357e-07,
|
| 810 |
+
"loss": 0.0,
|
| 811 |
+
"reward": 0.8219521045684814,
|
| 812 |
+
"reward_std": 0.6945097148418427,
|
| 813 |
+
"rewards/correct_code_reward_func": 0.5416666865348816,
|
| 814 |
+
"rewards/len_reward_func": 0.28028544783592224,
|
| 815 |
+
"step": 62
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"completion_length": 55.66666793823242,
|
| 819 |
+
"epoch": 1.0,
|
| 820 |
+
"grad_norm": 1.8272498546531741,
|
| 821 |
+
"kl": 0.02587890625,
|
| 822 |
+
"learning_rate": 2.597338798034344e-07,
|
| 823 |
+
"loss": 0.0,
|
| 824 |
+
"reward": 0.713922381401062,
|
| 825 |
+
"reward_std": 0.519837498664856,
|
| 826 |
+
"rewards/correct_code_reward_func": 0.4166666865348816,
|
| 827 |
+
"rewards/len_reward_func": 0.29725566506385803,
|
| 828 |
+
"step": 63
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"completion_length": 88.75000381469727,
|
| 832 |
+
"epoch": 1.016,
|
| 833 |
+
"grad_norm": 1.6950346991160663,
|
| 834 |
+
"kl": 0.0108642578125,
|
| 835 |
+
"learning_rate": 2.532453557994827e-07,
|
| 836 |
+
"loss": 0.0,
|
| 837 |
+
"reward": 0.5927524715662003,
|
| 838 |
+
"reward_std": 0.39128445088863373,
|
| 839 |
+
"rewards/correct_code_reward_func": 0.3750000149011612,
|
| 840 |
+
"rewards/len_reward_func": 0.21775247156620026,
|
| 841 |
+
"step": 64
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"completion_length": 151.7291717529297,
|
| 845 |
+
"epoch": 1.032,
|
| 846 |
+
"grad_norm": 1.6408461481438466,
|
| 847 |
+
"kl": 0.011138916015625,
|
| 848 |
+
"learning_rate": 2.467546442005173e-07,
|
| 849 |
+
"loss": 0.0,
|
| 850 |
+
"reward": 0.6122622489929199,
|
| 851 |
+
"reward_std": 0.5165137350559235,
|
| 852 |
+
"rewards/correct_code_reward_func": 0.3125000149011612,
|
| 853 |
+
"rewards/len_reward_func": 0.2997622489929199,
|
| 854 |
+
"step": 65
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"completion_length": 104.85417175292969,
|
| 858 |
+
"epoch": 1.048,
|
| 859 |
+
"grad_norm": 1.1573620161491798,
|
| 860 |
+
"kl": 0.01092529296875,
|
| 861 |
+
"learning_rate": 2.4026612019656556e-07,
|
| 862 |
+
"loss": 0.0,
|
| 863 |
+
"reward": 0.8486100733280182,
|
| 864 |
+
"reward_std": 0.3942585438489914,
|
| 865 |
+
"rewards/correct_code_reward_func": 0.5,
|
| 866 |
+
"rewards/len_reward_func": 0.348610058426857,
|
| 867 |
+
"step": 66
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"completion_length": 62.47916793823242,
|
| 871 |
+
"epoch": 1.064,
|
| 872 |
+
"grad_norm": 2.1966023559129266,
|
| 873 |
+
"kl": 0.018798828125,
|
| 874 |
+
"learning_rate": 2.337841575030642e-07,
|
| 875 |
+
"loss": 0.0,
|
| 876 |
+
"reward": 0.8105108737945557,
|
| 877 |
+
"reward_std": 0.4338831454515457,
|
| 878 |
+
"rewards/correct_code_reward_func": 0.4583333432674408,
|
| 879 |
+
"rewards/len_reward_func": 0.35217756032943726,
|
| 880 |
+
"step": 67
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"completion_length": 74.95833587646484,
|
| 884 |
+
"epoch": 1.08,
|
| 885 |
+
"grad_norm": 1.796160832910341,
|
| 886 |
+
"kl": 0.02294921875,
|
| 887 |
+
"learning_rate": 2.2731312541267143e-07,
|
| 888 |
+
"loss": 0.0,
|
| 889 |
+
"reward": 0.549996554851532,
|
| 890 |
+
"reward_std": 0.3687018007040024,
|
| 891 |
+
"rewards/correct_code_reward_func": 0.2083333358168602,
|
| 892 |
+
"rewards/len_reward_func": 0.3416632413864136,
|
| 893 |
+
"step": 68
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"completion_length": 80.14583587646484,
|
| 897 |
+
"epoch": 1.096,
|
| 898 |
+
"grad_norm": 2.1344146728324653,
|
| 899 |
+
"kl": 0.02447509765625,
|
| 900 |
+
"learning_rate": 2.2085738585006021e-07,
|
| 901 |
+
"loss": 0.0,
|
| 902 |
+
"reward": 0.8650955259799957,
|
| 903 |
+
"reward_std": 0.4139704555273056,
|
| 904 |
+
"rewards/correct_code_reward_func": 0.5208333432674408,
|
| 905 |
+
"rewards/len_reward_func": 0.34426216781139374,
|
| 906 |
+
"step": 69
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"completion_length": 60.958335876464844,
|
| 910 |
+
"epoch": 1.112,
|
| 911 |
+
"grad_norm": 1.6686676921157912,
|
| 912 |
+
"kl": 0.025634765625,
|
| 913 |
+
"learning_rate": 2.1442129043167873e-07,
|
| 914 |
+
"loss": 0.0,
|
| 915 |
+
"reward": 0.6947443187236786,
|
| 916 |
+
"reward_std": 0.5725615322589874,
|
| 917 |
+
"rewards/correct_code_reward_func": 0.375,
|
| 918 |
+
"rewards/len_reward_func": 0.319744348526001,
|
| 919 |
+
"step": 70
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"completion_length": 108.1875,
|
| 923 |
+
"epoch": 1.1280000000000001,
|
| 924 |
+
"grad_norm": 1.7272596794076989,
|
| 925 |
+
"kl": 0.0130615234375,
|
| 926 |
+
"learning_rate": 2.0800917753245875e-07,
|
| 927 |
+
"loss": 0.0,
|
| 928 |
+
"reward": 0.7587291896343231,
|
| 929 |
+
"reward_std": 0.5232284665107727,
|
| 930 |
+
"rewards/correct_code_reward_func": 0.4166666865348816,
|
| 931 |
+
"rewards/len_reward_func": 0.3420625329017639,
|
| 932 |
+
"step": 71
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"completion_length": 108.04167175292969,
|
| 936 |
+
"epoch": 1.144,
|
| 937 |
+
"grad_norm": 1.6272563745253346,
|
| 938 |
+
"kl": 0.01654052734375,
|
| 939 |
+
"learning_rate": 2.0162536936145008e-07,
|
| 940 |
+
"loss": 0.0,
|
| 941 |
+
"reward": 0.5046872794628143,
|
| 942 |
+
"reward_std": 0.3378771096467972,
|
| 943 |
+
"rewards/correct_code_reward_func": 0.1666666679084301,
|
| 944 |
+
"rewards/len_reward_func": 0.33802059292793274,
|
| 945 |
+
"step": 72
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"completion_length": 54.02083396911621,
|
| 949 |
+
"epoch": 1.16,
|
| 950 |
+
"grad_norm": 1.9418689539056528,
|
| 951 |
+
"kl": 0.0308837890625,
|
| 952 |
+
"learning_rate": 1.9527416904835132e-07,
|
| 953 |
+
"loss": 0.0,
|
| 954 |
+
"reward": 0.9055829644203186,
|
| 955 |
+
"reward_std": 0.3730238378047943,
|
| 956 |
+
"rewards/correct_code_reward_func": 0.5,
|
| 957 |
+
"rewards/len_reward_func": 0.405582919716835,
|
| 958 |
+
"step": 73
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"completion_length": 94.31250381469727,
|
| 962 |
+
"epoch": 1.176,
|
| 963 |
+
"grad_norm": 1.5576616620611914,
|
| 964 |
+
"kl": 0.02215576171875,
|
| 965 |
+
"learning_rate": 1.889598577429022e-07,
|
| 966 |
+
"loss": 0.0,
|
| 967 |
+
"reward": 0.9071804285049438,
|
| 968 |
+
"reward_std": 0.44920457899570465,
|
| 969 |
+
"rewards/correct_code_reward_func": 0.5000000298023224,
|
| 970 |
+
"rewards/len_reward_func": 0.40718045830726624,
|
| 971 |
+
"step": 74
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"completion_length": 53.79166793823242,
|
| 975 |
+
"epoch": 1.192,
|
| 976 |
+
"grad_norm": 2.3725141345867544,
|
| 977 |
+
"kl": 0.03057861328125,
|
| 978 |
+
"learning_rate": 1.8268669172909136e-07,
|
| 979 |
+
"loss": 0.0,
|
| 980 |
+
"reward": 0.9221459329128265,
|
| 981 |
+
"reward_std": 0.4697086811065674,
|
| 982 |
+
"rewards/correct_code_reward_func": 0.5000000298023224,
|
| 983 |
+
"rewards/len_reward_func": 0.42214588820934296,
|
| 984 |
+
"step": 75
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"completion_length": 89.79167175292969,
|
| 988 |
+
"epoch": 1.208,
|
| 989 |
+
"grad_norm": 2.003223060045919,
|
| 990 |
+
"kl": 0.03094482421875,
|
| 991 |
+
"learning_rate": 1.7645889955612592e-07,
|
| 992 |
+
"loss": 0.0,
|
| 993 |
+
"reward": 1.0163878798484802,
|
| 994 |
+
"reward_std": 0.43504565954208374,
|
| 995 |
+
"rewards/correct_code_reward_func": 0.6250000298023224,
|
| 996 |
+
"rewards/len_reward_func": 0.3913878947496414,
|
| 997 |
+
"step": 76
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"completion_length": 68.79166984558105,
|
| 1001 |
+
"epoch": 1.224,
|
| 1002 |
+
"grad_norm": 2.361523245499291,
|
| 1003 |
+
"kl": 0.0457763671875,
|
| 1004 |
+
"learning_rate": 1.7028067918809535e-07,
|
| 1005 |
+
"loss": 0.0,
|
| 1006 |
+
"reward": 0.7535229325294495,
|
| 1007 |
+
"reward_std": 0.47849828004837036,
|
| 1008 |
+
"rewards/correct_code_reward_func": 0.375,
|
| 1009 |
+
"rewards/len_reward_func": 0.3785228729248047,
|
| 1010 |
+
"step": 77
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"completion_length": 54.14583396911621,
|
| 1014 |
+
"epoch": 1.24,
|
| 1015 |
+
"grad_norm": 2.120116927446423,
|
| 1016 |
+
"kl": 0.0394287109375,
|
| 1017 |
+
"learning_rate": 1.6415619517425294e-07,
|
| 1018 |
+
"loss": 0.0,
|
| 1019 |
+
"reward": 0.8538325130939484,
|
| 1020 |
+
"reward_std": 0.44848716259002686,
|
| 1021 |
+
"rewards/correct_code_reward_func": 0.4791666865348816,
|
| 1022 |
+
"rewards/len_reward_func": 0.3746658265590668,
|
| 1023 |
+
"step": 78
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"completion_length": 89.0,
|
| 1027 |
+
"epoch": 1.256,
|
| 1028 |
+
"grad_norm": 1.2055136830985975,
|
| 1029 |
+
"kl": 0.0272216796875,
|
| 1030 |
+
"learning_rate": 1.5808957584181994e-07,
|
| 1031 |
+
"loss": 0.0,
|
| 1032 |
+
"reward": 0.755169004201889,
|
| 1033 |
+
"reward_std": 0.4014817923307419,
|
| 1034 |
+
"rewards/correct_code_reward_func": 0.3541666716337204,
|
| 1035 |
+
"rewards/len_reward_func": 0.40100236237049103,
|
| 1036 |
+
"step": 79
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"completion_length": 99.39583969116211,
|
| 1040 |
+
"epoch": 1.272,
|
| 1041 |
+
"grad_norm": 1.84690544945913,
|
| 1042 |
+
"kl": 0.024322509765625,
|
| 1043 |
+
"learning_rate": 1.5208491051320744e-07,
|
| 1044 |
+
"loss": 0.0,
|
| 1045 |
+
"reward": 0.7356246709823608,
|
| 1046 |
+
"reward_std": 0.47616493701934814,
|
| 1047 |
+
"rewards/correct_code_reward_func": 0.3958333432674408,
|
| 1048 |
+
"rewards/len_reward_func": 0.33979131281375885,
|
| 1049 |
+
"step": 80
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"completion_length": 73.04166793823242,
|
| 1053 |
+
"epoch": 1.288,
|
| 1054 |
+
"grad_norm": 1.7278725529442787,
|
| 1055 |
+
"kl": 0.0439453125,
|
| 1056 |
+
"learning_rate": 1.461462467495284e-07,
|
| 1057 |
+
"loss": 0.0,
|
| 1058 |
+
"reward": 0.7051982879638672,
|
| 1059 |
+
"reward_std": 0.48877203464508057,
|
| 1060 |
+
"rewards/correct_code_reward_func": 0.3125,
|
| 1061 |
+
"rewards/len_reward_func": 0.3926983177661896,
|
| 1062 |
+
"step": 81
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"completion_length": 59.354169845581055,
|
| 1066 |
+
"epoch": 1.304,
|
| 1067 |
+
"grad_norm": 2.077567652472909,
|
| 1068 |
+
"kl": 0.0345458984375,
|
| 1069 |
+
"learning_rate": 1.4027758762226107e-07,
|
| 1070 |
+
"loss": 0.0,
|
| 1071 |
+
"reward": 0.816185712814331,
|
| 1072 |
+
"reward_std": 0.4705541431903839,
|
| 1073 |
+
"rewards/correct_code_reward_func": 0.4791666865348816,
|
| 1074 |
+
"rewards/len_reward_func": 0.3370189964771271,
|
| 1075 |
+
"step": 82
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"completion_length": 81.58333587646484,
|
| 1079 |
+
"epoch": 1.32,
|
| 1080 |
+
"grad_norm": 1.609719907980881,
|
| 1081 |
+
"kl": 0.0234375,
|
| 1082 |
+
"learning_rate": 1.3448288901490092e-07,
|
| 1083 |
+
"loss": 0.0,
|
| 1084 |
+
"reward": 0.7908000648021698,
|
| 1085 |
+
"reward_std": 0.45585089921951294,
|
| 1086 |
+
"rewards/correct_code_reward_func": 0.4166666716337204,
|
| 1087 |
+
"rewards/len_reward_func": 0.374133437871933,
|
| 1088 |
+
"step": 83
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"completion_length": 87.33333587646484,
|
| 1092 |
+
"epoch": 1.336,
|
| 1093 |
+
"grad_norm": 1.6587537084233746,
|
| 1094 |
+
"kl": 0.02667236328125,
|
| 1095 |
+
"learning_rate": 1.2876605695642084e-07,
|
| 1096 |
+
"loss": 0.0,
|
| 1097 |
+
"reward": 0.6749401688575745,
|
| 1098 |
+
"reward_std": 0.42905712127685547,
|
| 1099 |
+
"rewards/correct_code_reward_func": 0.3541666716337204,
|
| 1100 |
+
"rewards/len_reward_func": 0.3207734525203705,
|
| 1101 |
+
"step": 84
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"completion_length": 95.20833587646484,
|
| 1105 |
+
"epoch": 1.3519999999999999,
|
| 1106 |
+
"grad_norm": 2.538472018686139,
|
| 1107 |
+
"kl": 0.02581787109375,
|
| 1108 |
+
"learning_rate": 1.231309449883361e-07,
|
| 1109 |
+
"loss": 0.0,
|
| 1110 |
+
"reward": 0.7594759464263916,
|
| 1111 |
+
"reward_std": 0.5746750831604004,
|
| 1112 |
+
"rewards/correct_code_reward_func": 0.3750000149011612,
|
| 1113 |
+
"rewards/len_reward_func": 0.3844759315252304,
|
| 1114 |
+
"step": 85
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"completion_length": 55.43750190734863,
|
| 1118 |
+
"epoch": 1.3679999999999999,
|
| 1119 |
+
"grad_norm": 1.797373425635401,
|
| 1120 |
+
"kl": 0.03289794921875,
|
| 1121 |
+
"learning_rate": 1.1758135156715041e-07,
|
| 1122 |
+
"loss": 0.0,
|
| 1123 |
+
"reward": 0.9961144328117371,
|
| 1124 |
+
"reward_std": 0.5648430436849594,
|
| 1125 |
+
"rewards/correct_code_reward_func": 0.6250000298023224,
|
| 1126 |
+
"rewards/len_reward_func": 0.37111443281173706,
|
| 1127 |
+
"step": 86
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"completion_length": 121.25000762939453,
|
| 1131 |
+
"epoch": 1.384,
|
| 1132 |
+
"grad_norm": 1.7119982491506713,
|
| 1133 |
+
"kl": 0.0286865234375,
|
| 1134 |
+
"learning_rate": 1.1212101750393235e-07,
|
| 1135 |
+
"loss": 0.0,
|
| 1136 |
+
"reward": 0.7243427634239197,
|
| 1137 |
+
"reward_std": 0.3805614560842514,
|
| 1138 |
+
"rewards/correct_code_reward_func": 0.3333333358168602,
|
| 1139 |
+
"rewards/len_reward_func": 0.39100944995880127,
|
| 1140 |
+
"step": 87
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"completion_length": 57.35416793823242,
|
| 1144 |
+
"epoch": 1.4,
|
| 1145 |
+
"grad_norm": 1.7713124187158098,
|
| 1146 |
+
"kl": 0.034912109375,
|
| 1147 |
+
"learning_rate": 1.0675362344274952e-07,
|
| 1148 |
+
"loss": 0.0,
|
| 1149 |
+
"reward": 0.7016758322715759,
|
| 1150 |
+
"reward_std": 0.5317542552947998,
|
| 1151 |
+
"rewards/correct_code_reward_func": 0.3541666865348816,
|
| 1152 |
+
"rewards/len_reward_func": 0.34750914573669434,
|
| 1153 |
+
"step": 88
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"completion_length": 59.0625,
|
| 1157 |
+
"epoch": 1.416,
|
| 1158 |
+
"grad_norm": 1.6492634665708499,
|
| 1159 |
+
"kl": 0.034423828125,
|
| 1160 |
+
"learning_rate": 1.0148278737965844e-07,
|
| 1161 |
+
"loss": 0.0,
|
| 1162 |
+
"reward": 0.7394144237041473,
|
| 1163 |
+
"reward_std": 0.4491709917783737,
|
| 1164 |
+
"rewards/correct_code_reward_func": 0.3541666716337204,
|
| 1165 |
+
"rewards/len_reward_func": 0.38524775207042694,
|
| 1166 |
+
"step": 89
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"completion_length": 48.6875,
|
| 1170 |
+
"epoch": 1.432,
|
| 1171 |
+
"grad_norm": 1.9432473699712165,
|
| 1172 |
+
"kl": 0.06494140625,
|
| 1173 |
+
"learning_rate": 9.631206222392479e-08,
|
| 1174 |
+
"loss": 0.0001,
|
| 1175 |
+
"reward": 0.8676341474056244,
|
| 1176 |
+
"reward_std": 0.3966159522533417,
|
| 1177 |
+
"rewards/correct_code_reward_func": 0.4791666865348816,
|
| 1178 |
+
"rewards/len_reward_func": 0.388467475771904,
|
| 1179 |
+
"step": 90
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"completion_length": 91.62500381469727,
|
| 1183 |
+
"epoch": 1.448,
|
| 1184 |
+
"grad_norm": 1.9189293687085252,
|
| 1185 |
+
"kl": 0.13482666015625,
|
| 1186 |
+
"learning_rate": 9.124493340311537e-08,
|
| 1187 |
+
"loss": 0.0001,
|
| 1188 |
+
"reward": 0.7231810688972473,
|
| 1189 |
+
"reward_std": 0.4981995224952698,
|
| 1190 |
+
"rewards/correct_code_reward_func": 0.3333333432674408,
|
| 1191 |
+
"rewards/len_reward_func": 0.3898477256298065,
|
| 1192 |
+
"step": 91
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"completion_length": 60.729169845581055,
|
| 1196 |
+
"epoch": 1.464,
|
| 1197 |
+
"grad_norm": 1.9825880271843388,
|
| 1198 |
+
"kl": 0.03424072265625,
|
| 1199 |
+
"learning_rate": 8.628481651367875e-08,
|
| 1200 |
+
"loss": 0.0,
|
| 1201 |
+
"reward": 0.8303024768829346,
|
| 1202 |
+
"reward_std": 0.40181903541088104,
|
| 1203 |
+
"rewards/correct_code_reward_func": 0.4375000149011612,
|
| 1204 |
+
"rewards/len_reward_func": 0.39280249178409576,
|
| 1205 |
+
"step": 92
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"completion_length": 58.22916793823242,
|
| 1209 |
+
"epoch": 1.48,
|
| 1210 |
+
"grad_norm": 1.8747344082688029,
|
| 1211 |
+
"kl": 0.0426025390625,
|
| 1212 |
+
"learning_rate": 8.143505501859551e-08,
|
| 1213 |
+
"loss": 0.0,
|
| 1214 |
+
"reward": 0.7909549474716187,
|
| 1215 |
+
"reward_std": 0.4536728262901306,
|
| 1216 |
+
"rewards/correct_code_reward_func": 0.458333358168602,
|
| 1217 |
+
"rewards/len_reward_func": 0.33262157440185547,
|
| 1218 |
+
"step": 93
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"completion_length": 125.10417175292969,
|
| 1222 |
+
"epoch": 1.496,
|
| 1223 |
+
"grad_norm": 1.5754029745287528,
|
| 1224 |
+
"kl": 0.02886962890625,
|
| 1225 |
+
"learning_rate": 7.669891799365282e-08,
|
| 1226 |
+
"loss": 0.0,
|
| 1227 |
+
"reward": 0.6297820806503296,
|
| 1228 |
+
"reward_std": 0.5051470398902893,
|
| 1229 |
+
"rewards/correct_code_reward_func": 0.2708333432674408,
|
| 1230 |
+
"rewards/len_reward_func": 0.3589487075805664,
|
| 1231 |
+
"step": 94
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"completion_length": 89.27083587646484,
|
| 1235 |
+
"epoch": 1.512,
|
| 1236 |
+
"grad_norm": 1.698829198816419,
|
| 1237 |
+
"kl": 0.02362060546875,
|
| 1238 |
+
"learning_rate": 7.207959792385998e-08,
|
| 1239 |
+
"loss": 0.0,
|
| 1240 |
+
"reward": 0.7924558222293854,
|
| 1241 |
+
"reward_std": 0.42506614327430725,
|
| 1242 |
+
"rewards/correct_code_reward_func": 0.3541666865348816,
|
| 1243 |
+
"rewards/len_reward_func": 0.4382891356945038,
|
| 1244 |
+
"step": 95
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"completion_length": 82.18750381469727,
|
| 1248 |
+
"epoch": 1.528,
|
| 1249 |
+
"grad_norm": 1.4031599496951968,
|
| 1250 |
+
"kl": 0.03643798828125,
|
| 1251 |
+
"learning_rate": 6.758020855149249e-08,
|
| 1252 |
+
"loss": 0.0,
|
| 1253 |
+
"reward": 0.6851500123739243,
|
| 1254 |
+
"reward_std": 0.2974398583173752,
|
| 1255 |
+
"rewards/correct_code_reward_func": 0.25000000558793545,
|
| 1256 |
+
"rewards/len_reward_func": 0.43515002727508545,
|
| 1257 |
+
"step": 96
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"completion_length": 54.6875,
|
| 1261 |
+
"epoch": 1.544,
|
| 1262 |
+
"grad_norm": 1.4467008481635895,
|
| 1263 |
+
"kl": 0.039306640625,
|
| 1264 |
+
"learning_rate": 6.320378277721342e-08,
|
| 1265 |
+
"loss": 0.0,
|
| 1266 |
+
"reward": 0.7509966492652893,
|
| 1267 |
+
"reward_std": 0.3042096644639969,
|
| 1268 |
+
"rewards/correct_code_reward_func": 0.3125,
|
| 1269 |
+
"rewards/len_reward_func": 0.4384966343641281,
|
| 1270 |
+
"step": 97
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"completion_length": 68.08333587646484,
|
| 1274 |
+
"epoch": 1.56,
|
| 1275 |
+
"grad_norm": 2.082709482850275,
|
| 1276 |
+
"kl": 0.03460693359375,
|
| 1277 |
+
"learning_rate": 5.895327061568775e-08,
|
| 1278 |
+
"loss": 0.0,
|
| 1279 |
+
"reward": 0.7968247532844543,
|
| 1280 |
+
"reward_std": 0.36605267226696014,
|
| 1281 |
+
"rewards/correct_code_reward_func": 0.3750000149011612,
|
| 1282 |
+
"rewards/len_reward_func": 0.42182472348213196,
|
| 1283 |
+
"step": 98
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"completion_length": 56.020835876464844,
|
| 1287 |
+
"epoch": 1.576,
|
| 1288 |
+
"grad_norm": 2.726579074776626,
|
| 1289 |
+
"kl": 0.0662841796875,
|
| 1290 |
+
"learning_rate": 5.483153720706798e-08,
|
| 1291 |
+
"loss": 0.0001,
|
| 1292 |
+
"reward": 0.8111520707607269,
|
| 1293 |
+
"reward_std": 0.548240602016449,
|
| 1294 |
+
"rewards/correct_code_reward_func": 0.4166666716337204,
|
| 1295 |
+
"rewards/len_reward_func": 0.3944854289293289,
|
| 1296 |
+
"step": 99
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"completion_length": 54.25000190734863,
|
| 1300 |
+
"epoch": 1.592,
|
| 1301 |
+
"grad_norm": 2.079061824739654,
|
| 1302 |
+
"kl": 0.0452880859375,
|
| 1303 |
+
"learning_rate": 5.0841360885690996e-08,
|
| 1304 |
+
"loss": 0.0,
|
| 1305 |
+
"reward": 0.9174363613128662,
|
| 1306 |
+
"reward_std": 0.46667972207069397,
|
| 1307 |
+
"rewards/correct_code_reward_func": 0.5416666865348816,
|
| 1308 |
+
"rewards/len_reward_func": 0.375769704580307,
|
| 1309 |
+
"step": 100
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"completion_length": 65.72916793823242,
|
| 1313 |
+
"epoch": 1.608,
|
| 1314 |
+
"grad_norm": 1.5292386933354263,
|
| 1315 |
+
"kl": 0.04522705078125,
|
| 1316 |
+
"learning_rate": 4.698543130728755e-08,
|
| 1317 |
+
"loss": 0.0,
|
| 1318 |
+
"reward": 0.8213175535202026,
|
| 1319 |
+
"reward_std": 0.38392098248004913,
|
| 1320 |
+
"rewards/correct_code_reward_func": 0.458333358168602,
|
| 1321 |
+
"rewards/len_reward_func": 0.3629842549562454,
|
| 1322 |
+
"step": 101
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"completion_length": 67.77083587646484,
|
| 1326 |
+
"epoch": 1.624,
|
| 1327 |
+
"grad_norm": 1.352325105446135,
|
| 1328 |
+
"kl": 0.0390625,
|
| 1329 |
+
"learning_rate": 4.326634763596784e-08,
|
| 1330 |
+
"loss": 0.0,
|
| 1331 |
+
"reward": 0.7263242900371552,
|
| 1332 |
+
"reward_std": 0.37168650329113007,
|
| 1333 |
+
"rewards/correct_code_reward_func": 0.31250002048909664,
|
| 1334 |
+
"rewards/len_reward_func": 0.41382429003715515,
|
| 1335 |
+
"step": 102
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"completion_length": 64.10416793823242,
|
| 1339 |
+
"epoch": 1.6400000000000001,
|
| 1340 |
+
"grad_norm": 1.9987254276022863,
|
| 1341 |
+
"kl": 0.02880859375,
|
| 1342 |
+
"learning_rate": 3.968661679220467e-08,
|
| 1343 |
+
"loss": 0.0,
|
| 1344 |
+
"reward": 1.174392580986023,
|
| 1345 |
+
"reward_std": 0.4813085198402405,
|
| 1346 |
+
"rewards/correct_code_reward_func": 0.7500000298023224,
|
| 1347 |
+
"rewards/len_reward_func": 0.42439255118370056,
|
| 1348 |
+
"step": 103
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"completion_length": 57.437503814697266,
|
| 1352 |
+
"epoch": 1.6560000000000001,
|
| 1353 |
+
"grad_norm": 1.5506203528349733,
|
| 1354 |
+
"kl": 0.041015625,
|
| 1355 |
+
"learning_rate": 3.624865176299499e-08,
|
| 1356 |
+
"loss": 0.0,
|
| 1357 |
+
"reward": 0.9918626546859741,
|
| 1358 |
+
"reward_std": 0.5309067815542221,
|
| 1359 |
+
"rewards/correct_code_reward_func": 0.6666666865348816,
|
| 1360 |
+
"rewards/len_reward_func": 0.3251959830522537,
|
| 1361 |
+
"step": 104
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"completion_length": 114.50000762939453,
|
| 1365 |
+
"epoch": 1.6720000000000002,
|
| 1366 |
+
"grad_norm": 1.538301941895194,
|
| 1367 |
+
"kl": 0.0245361328125,
|
| 1368 |
+
"learning_rate": 3.295476997533905e-08,
|
| 1369 |
+
"loss": 0.0,
|
| 1370 |
+
"reward": 0.9100688099861145,
|
| 1371 |
+
"reward_std": 0.29824198782444,
|
| 1372 |
+
"rewards/correct_code_reward_func": 0.4583333432674408,
|
| 1373 |
+
"rewards/len_reward_func": 0.4517354816198349,
|
| 1374 |
+
"step": 105
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"completion_length": 129.81250381469727,
|
| 1378 |
+
"epoch": 1.688,
|
| 1379 |
+
"grad_norm": 1.3867754807731443,
|
| 1380 |
+
"kl": 0.0283203125,
|
| 1381 |
+
"learning_rate": 2.980719173413396e-08,
|
| 1382 |
+
"loss": 0.0,
|
| 1383 |
+
"reward": 0.818383663892746,
|
| 1384 |
+
"reward_std": 0.5115247815847397,
|
| 1385 |
+
"rewards/correct_code_reward_func": 0.4166666716337204,
|
| 1386 |
+
"rewards/len_reward_func": 0.4017169624567032,
|
| 1387 |
+
"step": 106
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"completion_length": 73.33333587646484,
|
| 1391 |
+
"epoch": 1.704,
|
| 1392 |
+
"grad_norm": 2.2267187145460765,
|
| 1393 |
+
"kl": 0.04461669921875,
|
| 1394 |
+
"learning_rate": 2.680803872553408e-08,
|
| 1395 |
+
"loss": 0.0,
|
| 1396 |
+
"reward": 0.8567679226398468,
|
| 1397 |
+
"reward_std": 0.51302769780159,
|
| 1398 |
+
"rewards/correct_code_reward_func": 0.4375,
|
| 1399 |
+
"rewards/len_reward_func": 0.4192679077386856,
|
| 1400 |
+
"step": 107
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"completion_length": 53.54166793823242,
|
| 1404 |
+
"epoch": 1.72,
|
| 1405 |
+
"grad_norm": 3.1940102299602953,
|
| 1406 |
+
"kl": 0.0521240234375,
|
| 1407 |
+
"learning_rate": 2.395933258678745e-08,
|
| 1408 |
+
"loss": 0.0001,
|
| 1409 |
+
"reward": 0.9940223693847656,
|
| 1410 |
+
"reward_std": 0.46572498977184296,
|
| 1411 |
+
"rewards/correct_code_reward_func": 0.6041666865348816,
|
| 1412 |
+
"rewards/len_reward_func": 0.3898557126522064,
|
| 1413 |
+
"step": 108
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"completion_length": 41.52083396911621,
|
| 1417 |
+
"epoch": 1.736,
|
| 1418 |
+
"grad_norm": 2.0727978566546295,
|
| 1419 |
+
"kl": 0.0655517578125,
|
| 1420 |
+
"learning_rate": 2.1262993543511715e-08,
|
| 1421 |
+
"loss": 0.0001,
|
| 1422 |
+
"reward": 0.9489125609397888,
|
| 1423 |
+
"reward_std": 0.5604254603385925,
|
| 1424 |
+
"rewards/correct_code_reward_func": 0.6250000298023224,
|
| 1425 |
+
"rewards/len_reward_func": 0.32391248643398285,
|
| 1426 |
+
"step": 109
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"completion_length": 106.08333587646484,
|
| 1430 |
+
"epoch": 1.752,
|
| 1431 |
+
"grad_norm": 2.3414859603625806,
|
| 1432 |
+
"kl": 0.03424072265625,
|
| 1433 |
+
"learning_rate": 1.872083911532907e-08,
|
| 1434 |
+
"loss": 0.0,
|
| 1435 |
+
"reward": 0.5710697174072266,
|
| 1436 |
+
"reward_std": 0.4303289204835892,
|
| 1437 |
+
"rewards/correct_code_reward_func": 0.1666666679084301,
|
| 1438 |
+
"rewards/len_reward_func": 0.4044030159711838,
|
| 1439 |
+
"step": 110
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"completion_length": 60.437503814697266,
|
| 1443 |
+
"epoch": 1.768,
|
| 1444 |
+
"grad_norm": 1.5494191116212308,
|
| 1445 |
+
"kl": 0.046875,
|
| 1446 |
+
"learning_rate": 1.6334582890731697e-08,
|
| 1447 |
+
"loss": 0.0,
|
| 1448 |
+
"reward": 1.0543819665908813,
|
| 1449 |
+
"reward_std": 0.4688963294029236,
|
| 1450 |
+
"rewards/correct_code_reward_func": 0.6666666865348816,
|
| 1451 |
+
"rewards/len_reward_func": 0.38771532475948334,
|
| 1452 |
+
"step": 111
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"completion_length": 139.43750381469727,
|
| 1456 |
+
"epoch": 1.784,
|
| 1457 |
+
"grad_norm": 1.8975149766982131,
|
| 1458 |
+
"kl": 0.0323486328125,
|
| 1459 |
+
"learning_rate": 1.4105833372004523e-08,
|
| 1460 |
+
"loss": 0.0,
|
| 1461 |
+
"reward": 0.7198583781719208,
|
| 1462 |
+
"reward_std": 0.2770904451608658,
|
| 1463 |
+
"rewards/correct_code_reward_func": 0.2708333395421505,
|
| 1464 |
+
"rewards/len_reward_func": 0.4490250498056412,
|
| 1465 |
+
"step": 112
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"completion_length": 71.87500190734863,
|
| 1469 |
+
"epoch": 1.8,
|
| 1470 |
+
"grad_norm": 1.8779975481307012,
|
| 1471 |
+
"kl": 0.0350341796875,
|
| 1472 |
+
"learning_rate": 1.2036092890982619e-08,
|
| 1473 |
+
"loss": 0.0,
|
| 1474 |
+
"reward": 0.6213224828243256,
|
| 1475 |
+
"reward_std": 0.39381173253059387,
|
| 1476 |
+
"rewards/correct_code_reward_func": 0.25,
|
| 1477 |
+
"rewards/len_reward_func": 0.3713224530220032,
|
| 1478 |
+
"step": 113
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"completion_length": 73.16666793823242,
|
| 1482 |
+
"epoch": 1.8159999999999998,
|
| 1483 |
+
"grad_norm": 1.625916920606493,
|
| 1484 |
+
"kl": 0.04345703125,
|
| 1485 |
+
"learning_rate": 1.0126756596375685e-08,
|
| 1486 |
+
"loss": 0.0,
|
| 1487 |
+
"reward": 0.8906111121177673,
|
| 1488 |
+
"reward_std": 0.5251133739948273,
|
| 1489 |
+
"rewards/correct_code_reward_func": 0.4791666865348816,
|
| 1490 |
+
"rewards/len_reward_func": 0.41144441068172455,
|
| 1491 |
+
"step": 114
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"completion_length": 39.85416793823242,
|
| 1495 |
+
"epoch": 1.8319999999999999,
|
| 1496 |
+
"grad_norm": 1.8155165345051183,
|
| 1497 |
+
"kl": 0.0440673828125,
|
| 1498 |
+
"learning_rate": 8.379111513340753e-09,
|
| 1499 |
+
"loss": 0.0,
|
| 1500 |
+
"reward": 0.8687795996665955,
|
| 1501 |
+
"reward_std": 0.4838385283946991,
|
| 1502 |
+
"rewards/correct_code_reward_func": 0.4583333358168602,
|
| 1503 |
+
"rewards/len_reward_func": 0.41044625639915466,
|
| 1504 |
+
"step": 115
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"completion_length": 75.58333396911621,
|
| 1508 |
+
"epoch": 1.8479999999999999,
|
| 1509 |
+
"grad_norm": 1.8222797961879316,
|
| 1510 |
+
"kl": 0.03985595703125,
|
| 1511 |
+
"learning_rate": 6.7943356759381785e-09,
|
| 1512 |
+
"loss": 0.0,
|
| 1513 |
+
"reward": 0.9320607483386993,
|
| 1514 |
+
"reward_std": 0.5384509861469269,
|
| 1515 |
+
"rewards/correct_code_reward_func": 0.5416666865348816,
|
| 1516 |
+
"rewards/len_reward_func": 0.39039406180381775,
|
| 1517 |
+
"step": 116
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"completion_length": 68.54166984558105,
|
| 1521 |
+
"epoch": 1.8639999999999999,
|
| 1522 |
+
"grad_norm": 2.0020075086567775,
|
| 1523 |
+
"kl": 0.031982421875,
|
| 1524 |
+
"learning_rate": 5.373497333054616e-09,
|
| 1525 |
+
"loss": 0.0,
|
| 1526 |
+
"reward": 0.9275134801864624,
|
| 1527 |
+
"reward_std": 0.4482097327709198,
|
| 1528 |
+
"rewards/correct_code_reward_func": 0.5000000298023224,
|
| 1529 |
+
"rewards/len_reward_func": 0.4275134950876236,
|
| 1530 |
+
"step": 117
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"completion_length": 73.91666793823242,
|
| 1534 |
+
"epoch": 1.88,
|
| 1535 |
+
"grad_norm": 1.7788611304062052,
|
| 1536 |
+
"kl": 0.03240966796875,
|
| 1537 |
+
"learning_rate": 4.117554228329406e-09,
|
| 1538 |
+
"loss": 0.0,
|
| 1539 |
+
"reward": 0.9304822385311127,
|
| 1540 |
+
"reward_std": 0.5174555033445358,
|
| 1541 |
+
"rewards/correct_code_reward_func": 0.5416666865348816,
|
| 1542 |
+
"rewards/len_reward_func": 0.38881558179855347,
|
| 1543 |
+
"step": 118
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"completion_length": 56.20833396911621,
|
| 1547 |
+
"epoch": 1.896,
|
| 1548 |
+
"grad_norm": 2.1126141119280257,
|
| 1549 |
+
"kl": 0.0341796875,
|
| 1550 |
+
"learning_rate": 3.0273529545687125e-09,
|
| 1551 |
+
"loss": 0.0,
|
| 1552 |
+
"reward": 0.7594221532344818,
|
| 1553 |
+
"reward_std": 0.480338990688324,
|
| 1554 |
+
"rewards/correct_code_reward_func": 0.3958333432674408,
|
| 1555 |
+
"rewards/len_reward_func": 0.3635888248682022,
|
| 1556 |
+
"step": 119
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"completion_length": 72.47916793823242,
|
| 1560 |
+
"epoch": 1.912,
|
| 1561 |
+
"grad_norm": 1.4598566413193612,
|
| 1562 |
+
"kl": 0.03466796875,
|
| 1563 |
+
"learning_rate": 2.1036283830834224e-09,
|
| 1564 |
+
"loss": 0.0,
|
| 1565 |
+
"reward": 0.7889427244663239,
|
| 1566 |
+
"reward_std": 0.48503294587135315,
|
| 1567 |
+
"rewards/correct_code_reward_func": 0.3958333432674408,
|
| 1568 |
+
"rewards/len_reward_func": 0.39310936629772186,
|
| 1569 |
+
"step": 120
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"completion_length": 40.85416793823242,
|
| 1573 |
+
"epoch": 1.928,
|
| 1574 |
+
"grad_norm": 2.335195303935002,
|
| 1575 |
+
"kl": 0.056640625,
|
| 1576 |
+
"learning_rate": 1.347003168334665e-09,
|
| 1577 |
+
"loss": 0.0001,
|
| 1578 |
+
"reward": 1.0662382543087006,
|
| 1579 |
+
"reward_std": 0.2768351137638092,
|
| 1580 |
+
"rewards/correct_code_reward_func": 0.6250000149011612,
|
| 1581 |
+
"rewards/len_reward_func": 0.44123825430870056,
|
| 1582 |
+
"step": 121
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"completion_length": 50.62500190734863,
|
| 1586 |
+
"epoch": 1.944,
|
| 1587 |
+
"grad_norm": 1.8386331097859265,
|
| 1588 |
+
"kl": 0.03173828125,
|
| 1589 |
+
"learning_rate": 7.579873282216598e-10,
|
| 1590 |
+
"loss": 0.0,
|
| 1591 |
+
"reward": 0.8906074166297913,
|
| 1592 |
+
"reward_std": 0.5252098143100739,
|
| 1593 |
+
"rewards/correct_code_reward_func": 0.5833333730697632,
|
| 1594 |
+
"rewards/len_reward_func": 0.30727406591176987,
|
| 1595 |
+
"step": 122
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"completion_length": 99.4375057220459,
|
| 1599 |
+
"epoch": 1.96,
|
| 1600 |
+
"grad_norm": 1.621045537411182,
|
| 1601 |
+
"kl": 0.0238037109375,
|
| 1602 |
+
"learning_rate": 3.3697790029424413e-10,
|
| 1603 |
+
"loss": 0.0,
|
| 1604 |
+
"reward": 0.9505272507667542,
|
| 1605 |
+
"reward_std": 0.5842320024967194,
|
| 1606 |
+
"rewards/correct_code_reward_func": 0.5833333432674408,
|
| 1607 |
+
"rewards/len_reward_func": 0.36719387769699097,
|
| 1608 |
+
"step": 123
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"completion_length": 63.000003814697266,
|
| 1612 |
+
"epoch": 1.976,
|
| 1613 |
+
"grad_norm": 2.157350197672568,
|
| 1614 |
+
"kl": 0.0465087890625,
|
| 1615 |
+
"learning_rate": 8.425867412190091e-11,
|
| 1616 |
+
"loss": 0.0,
|
| 1617 |
+
"reward": 0.9762873649597168,
|
| 1618 |
+
"reward_std": 0.5066816210746765,
|
| 1619 |
+
"rewards/correct_code_reward_func": 0.5833333432674408,
|
| 1620 |
+
"rewards/len_reward_func": 0.3929540067911148,
|
| 1621 |
+
"step": 124
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"completion_length": 126.97917175292969,
|
| 1625 |
+
"epoch": 1.992,
|
| 1626 |
+
"grad_norm": 1.7642641467833304,
|
| 1627 |
+
"kl": 0.02130126953125,
|
| 1628 |
+
"learning_rate": 0.0,
|
| 1629 |
+
"loss": 0.0,
|
| 1630 |
+
"reward": 0.7899810075759888,
|
| 1631 |
+
"reward_std": 0.38732415437698364,
|
| 1632 |
+
"rewards/correct_code_reward_func": 0.3750000149011612,
|
| 1633 |
+
"rewards/len_reward_func": 0.41498102247714996,
|
| 1634 |
+
"step": 125
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.992,
|
| 1638 |
+
"step": 125,
|
| 1639 |
+
"total_flos": 0.0,
|
| 1640 |
+
"train_loss": 1.9367338650191358e-05,
|
| 1641 |
+
"train_runtime": 3648.0047,
|
| 1642 |
+
"train_samples_per_second": 0.206,
|
| 1643 |
+
"train_steps_per_second": 0.034
|
| 1644 |
+
}
|
| 1645 |
+
],
|
| 1646 |
+
"logging_steps": 1,
|
| 1647 |
+
"max_steps": 125,
|
| 1648 |
+
"num_input_tokens_seen": 0,
|
| 1649 |
+
"num_train_epochs": 3,
|
| 1650 |
+
"save_steps": 25,
|
| 1651 |
+
"stateful_callbacks": {
|
| 1652 |
+
"TrainerControl": {
|
| 1653 |
+
"args": {
|
| 1654 |
+
"should_epoch_stop": false,
|
| 1655 |
+
"should_evaluate": false,
|
| 1656 |
+
"should_log": false,
|
| 1657 |
+
"should_save": true,
|
| 1658 |
+
"should_training_stop": true
|
| 1659 |
+
},
|
| 1660 |
+
"attributes": {}
|
| 1661 |
+
}
|
| 1662 |
+
},
|
| 1663 |
+
"total_flos": 0.0,
|
| 1664 |
+
"train_batch_size": 1,
|
| 1665 |
+
"trial_name": null,
|
| 1666 |
+
"trial_params": null
|
| 1667 |
+
}
|