Model save
Browse files- README.md +68 -0
- all_results.json +8 -0
- config.json +1 -1
- generation_config.json +14 -0
- train_results.json +8 -0
- trainer_state.json +2967 -0
README.md
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: Qwen/Qwen2.5-3B-Instruct
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: qwen-2.5-3b-r1-countdown
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- grpo
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for qwen-2.5-3b-r1-countdown
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct).
|
| 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="pullpull/qwen-2.5-3b-r1-countdown", 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": 9.454646318052558e-05,
|
| 4 |
+
"train_runtime": 16471.8629,
|
| 5 |
+
"train_samples": 45000,
|
| 6 |
+
"train_samples_per_second": 0.219,
|
| 7 |
+
"train_steps_per_second": 0.027
|
| 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 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"do_sample": true,
|
| 4 |
+
"eos_token_id": [
|
| 5 |
+
151645,
|
| 6 |
+
151643
|
| 7 |
+
],
|
| 8 |
+
"pad_token_id": 151643,
|
| 9 |
+
"repetition_penalty": 1.05,
|
| 10 |
+
"temperature": 0.7,
|
| 11 |
+
"top_k": 20,
|
| 12 |
+
"top_p": 0.8,
|
| 13 |
+
"transformers_version": "4.48.1"
|
| 14 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 0.0,
|
| 3 |
+
"train_loss": 9.454646318052558e-05,
|
| 4 |
+
"train_runtime": 16471.8629,
|
| 5 |
+
"train_samples": 45000,
|
| 6 |
+
"train_samples_per_second": 0.219,
|
| 7 |
+
"train_steps_per_second": 0.027
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2967 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.08,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 450,
|
| 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": 558.21875,
|
| 13 |
+
"epoch": 0.00035555555555555557,
|
| 14 |
+
"grad_norm": 0.16319611465157086,
|
| 15 |
+
"kl": 0.0,
|
| 16 |
+
"learning_rate": 7.142857142857142e-08,
|
| 17 |
+
"loss": -0.0,
|
| 18 |
+
"reward": 0.28125,
|
| 19 |
+
"reward_std": 0.4586672969162464,
|
| 20 |
+
"rewards/equation_reward_func": 0.015625,
|
| 21 |
+
"rewards/format_reward_func": 0.265625,
|
| 22 |
+
"step": 2
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"completion_length": 516.21875,
|
| 26 |
+
"epoch": 0.0007111111111111111,
|
| 27 |
+
"grad_norm": 0.17418126905911543,
|
| 28 |
+
"kl": 0.00036662817001342773,
|
| 29 |
+
"learning_rate": 1.4285714285714285e-07,
|
| 30 |
+
"loss": 0.0,
|
| 31 |
+
"reward": 0.3125,
|
| 32 |
+
"reward_std": 0.4291788823902607,
|
| 33 |
+
"rewards/equation_reward_func": 0.0703125,
|
| 34 |
+
"rewards/format_reward_func": 0.2421875,
|
| 35 |
+
"step": 4
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"completion_length": 511.15625,
|
| 39 |
+
"epoch": 0.0010666666666666667,
|
| 40 |
+
"grad_norm": 0.18686057402139436,
|
| 41 |
+
"kl": 0.0004163980484008789,
|
| 42 |
+
"learning_rate": 2.1428571428571426e-07,
|
| 43 |
+
"loss": 0.0,
|
| 44 |
+
"reward": 0.2578125,
|
| 45 |
+
"reward_std": 0.4385656677186489,
|
| 46 |
+
"rewards/equation_reward_func": 0.0625,
|
| 47 |
+
"rewards/format_reward_func": 0.1953125,
|
| 48 |
+
"step": 6
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"completion_length": 477.4609375,
|
| 52 |
+
"epoch": 0.0014222222222222223,
|
| 53 |
+
"grad_norm": 0.1777894589379156,
|
| 54 |
+
"kl": 0.00041472911834716797,
|
| 55 |
+
"learning_rate": 2.857142857142857e-07,
|
| 56 |
+
"loss": 0.0,
|
| 57 |
+
"reward": 0.3203125,
|
| 58 |
+
"reward_std": 0.446503184735775,
|
| 59 |
+
"rewards/equation_reward_func": 0.046875,
|
| 60 |
+
"rewards/format_reward_func": 0.2734375,
|
| 61 |
+
"step": 8
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"completion_length": 506.890625,
|
| 65 |
+
"epoch": 0.0017777777777777779,
|
| 66 |
+
"grad_norm": 0.2354010939066695,
|
| 67 |
+
"kl": 0.0004538297653198242,
|
| 68 |
+
"learning_rate": 3.5714285714285716e-07,
|
| 69 |
+
"loss": 0.0,
|
| 70 |
+
"reward": 0.3828125,
|
| 71 |
+
"reward_std": 0.4801974333822727,
|
| 72 |
+
"rewards/equation_reward_func": 0.0703125,
|
| 73 |
+
"rewards/format_reward_func": 0.3125,
|
| 74 |
+
"step": 10
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"completion_length": 475.9765625,
|
| 78 |
+
"epoch": 0.0021333333333333334,
|
| 79 |
+
"grad_norm": 0.17183303573120517,
|
| 80 |
+
"kl": 0.0004565715789794922,
|
| 81 |
+
"learning_rate": 4.285714285714285e-07,
|
| 82 |
+
"loss": 0.0,
|
| 83 |
+
"reward": 0.2578125,
|
| 84 |
+
"reward_std": 0.4146980494260788,
|
| 85 |
+
"rewards/equation_reward_func": 0.015625,
|
| 86 |
+
"rewards/format_reward_func": 0.2421875,
|
| 87 |
+
"step": 12
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"completion_length": 522.140625,
|
| 91 |
+
"epoch": 0.002488888888888889,
|
| 92 |
+
"grad_norm": 0.19584841533297276,
|
| 93 |
+
"kl": 0.0005878210067749023,
|
| 94 |
+
"learning_rate": 5e-07,
|
| 95 |
+
"loss": 0.0,
|
| 96 |
+
"reward": 0.3984375,
|
| 97 |
+
"reward_std": 0.5455609019845724,
|
| 98 |
+
"rewards/equation_reward_func": 0.0625,
|
| 99 |
+
"rewards/format_reward_func": 0.3359375,
|
| 100 |
+
"step": 14
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"completion_length": 504.8359375,
|
| 104 |
+
"epoch": 0.0028444444444444446,
|
| 105 |
+
"grad_norm": 0.1963897788415132,
|
| 106 |
+
"kl": 0.0007581710815429688,
|
| 107 |
+
"learning_rate": 4.999740409224932e-07,
|
| 108 |
+
"loss": 0.0,
|
| 109 |
+
"reward": 0.4453125,
|
| 110 |
+
"reward_std": 0.507007647305727,
|
| 111 |
+
"rewards/equation_reward_func": 0.0390625,
|
| 112 |
+
"rewards/format_reward_func": 0.40625,
|
| 113 |
+
"step": 16
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"completion_length": 439.265625,
|
| 117 |
+
"epoch": 0.0032,
|
| 118 |
+
"grad_norm": 0.17301487165752755,
|
| 119 |
+
"kl": 0.0015461444854736328,
|
| 120 |
+
"learning_rate": 4.998961690809627e-07,
|
| 121 |
+
"loss": 0.0,
|
| 122 |
+
"reward": 0.6015625,
|
| 123 |
+
"reward_std": 0.5544095803052187,
|
| 124 |
+
"rewards/equation_reward_func": 0.078125,
|
| 125 |
+
"rewards/format_reward_func": 0.5234375,
|
| 126 |
+
"step": 18
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"completion_length": 440.75,
|
| 130 |
+
"epoch": 0.0035555555555555557,
|
| 131 |
+
"grad_norm": 0.1656037102176249,
|
| 132 |
+
"kl": 0.004067420959472656,
|
| 133 |
+
"learning_rate": 4.997664006472578e-07,
|
| 134 |
+
"loss": 0.0,
|
| 135 |
+
"reward": 0.734375,
|
| 136 |
+
"reward_std": 0.5903380792587996,
|
| 137 |
+
"rewards/equation_reward_func": 0.125,
|
| 138 |
+
"rewards/format_reward_func": 0.609375,
|
| 139 |
+
"step": 20
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"completion_length": 411.96875,
|
| 143 |
+
"epoch": 0.003911111111111111,
|
| 144 |
+
"grad_norm": 0.1908015876481413,
|
| 145 |
+
"kl": 0.00445556640625,
|
| 146 |
+
"learning_rate": 4.995847625707292e-07,
|
| 147 |
+
"loss": 0.0,
|
| 148 |
+
"reward": 0.765625,
|
| 149 |
+
"reward_std": 0.5255697090178728,
|
| 150 |
+
"rewards/equation_reward_func": 0.0859375,
|
| 151 |
+
"rewards/format_reward_func": 0.6796875,
|
| 152 |
+
"step": 22
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"completion_length": 445.640625,
|
| 156 |
+
"epoch": 0.004266666666666667,
|
| 157 |
+
"grad_norm": 0.19251232902264467,
|
| 158 |
+
"kl": 0.005351066589355469,
|
| 159 |
+
"learning_rate": 4.993512925726318e-07,
|
| 160 |
+
"loss": 0.0,
|
| 161 |
+
"reward": 0.8203125,
|
| 162 |
+
"reward_std": 0.4650081805884838,
|
| 163 |
+
"rewards/equation_reward_func": 0.0625,
|
| 164 |
+
"rewards/format_reward_func": 0.7578125,
|
| 165 |
+
"step": 24
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"completion_length": 450.7890625,
|
| 169 |
+
"epoch": 0.004622222222222222,
|
| 170 |
+
"grad_norm": 0.1362690788735505,
|
| 171 |
+
"kl": 0.0061969757080078125,
|
| 172 |
+
"learning_rate": 4.990660391382923e-07,
|
| 173 |
+
"loss": 0.0,
|
| 174 |
+
"reward": 0.8828125,
|
| 175 |
+
"reward_std": 0.49951413460075855,
|
| 176 |
+
"rewards/equation_reward_func": 0.1015625,
|
| 177 |
+
"rewards/format_reward_func": 0.78125,
|
| 178 |
+
"step": 26
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"completion_length": 427.34375,
|
| 182 |
+
"epoch": 0.004977777777777778,
|
| 183 |
+
"grad_norm": 0.1397441605336418,
|
| 184 |
+
"kl": 0.0075531005859375,
|
| 185 |
+
"learning_rate": 4.987290615070384e-07,
|
| 186 |
+
"loss": 0.0,
|
| 187 |
+
"reward": 1.03125,
|
| 188 |
+
"reward_std": 0.3855034988373518,
|
| 189 |
+
"rewards/equation_reward_func": 0.140625,
|
| 190 |
+
"rewards/format_reward_func": 0.890625,
|
| 191 |
+
"step": 28
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"completion_length": 449.6953125,
|
| 195 |
+
"epoch": 0.005333333333333333,
|
| 196 |
+
"grad_norm": 0.13266541127233653,
|
| 197 |
+
"kl": 0.008485794067382812,
|
| 198 |
+
"learning_rate": 4.983404296598978e-07,
|
| 199 |
+
"loss": 0.0,
|
| 200 |
+
"reward": 0.90625,
|
| 201 |
+
"reward_std": 0.42431665025651455,
|
| 202 |
+
"rewards/equation_reward_func": 0.109375,
|
| 203 |
+
"rewards/format_reward_func": 0.796875,
|
| 204 |
+
"step": 30
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"completion_length": 449.546875,
|
| 208 |
+
"epoch": 0.005688888888888889,
|
| 209 |
+
"grad_norm": 0.17828848506675551,
|
| 210 |
+
"kl": 0.009744644165039062,
|
| 211 |
+
"learning_rate": 4.979002243050646e-07,
|
| 212 |
+
"loss": 0.0,
|
| 213 |
+
"reward": 0.9765625,
|
| 214 |
+
"reward_std": 0.4631799254566431,
|
| 215 |
+
"rewards/equation_reward_func": 0.1171875,
|
| 216 |
+
"rewards/format_reward_func": 0.859375,
|
| 217 |
+
"step": 32
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"completion_length": 513.921875,
|
| 221 |
+
"epoch": 0.006044444444444444,
|
| 222 |
+
"grad_norm": 0.14516749872112322,
|
| 223 |
+
"kl": 0.009546279907226562,
|
| 224 |
+
"learning_rate": 4.974085368611381e-07,
|
| 225 |
+
"loss": 0.0,
|
| 226 |
+
"reward": 0.9453125,
|
| 227 |
+
"reward_std": 0.37294892966747284,
|
| 228 |
+
"rewards/equation_reward_func": 0.0625,
|
| 229 |
+
"rewards/format_reward_func": 0.8828125,
|
| 230 |
+
"step": 34
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"completion_length": 471.6796875,
|
| 234 |
+
"epoch": 0.0064,
|
| 235 |
+
"grad_norm": 0.1472615847955791,
|
| 236 |
+
"kl": 0.010135650634765625,
|
| 237 |
+
"learning_rate": 4.968654694381379e-07,
|
| 238 |
+
"loss": 0.0,
|
| 239 |
+
"reward": 1.0078125,
|
| 240 |
+
"reward_std": 0.36952908150851727,
|
| 241 |
+
"rewards/equation_reward_func": 0.109375,
|
| 242 |
+
"rewards/format_reward_func": 0.8984375,
|
| 243 |
+
"step": 36
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"completion_length": 502.9921875,
|
| 247 |
+
"epoch": 0.0067555555555555554,
|
| 248 |
+
"grad_norm": 0.11856505582596304,
|
| 249 |
+
"kl": 0.016201019287109375,
|
| 250 |
+
"learning_rate": 4.962711348162987e-07,
|
| 251 |
+
"loss": 0.0,
|
| 252 |
+
"reward": 0.9453125,
|
| 253 |
+
"reward_std": 0.33866361156105995,
|
| 254 |
+
"rewards/equation_reward_func": 0.078125,
|
| 255 |
+
"rewards/format_reward_func": 0.8671875,
|
| 256 |
+
"step": 38
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"completion_length": 438.1875,
|
| 260 |
+
"epoch": 0.0071111111111111115,
|
| 261 |
+
"grad_norm": 0.1491753865500343,
|
| 262 |
+
"kl": 0.01209259033203125,
|
| 263 |
+
"learning_rate": 4.956256564226487e-07,
|
| 264 |
+
"loss": 0.0,
|
| 265 |
+
"reward": 1.0234375,
|
| 266 |
+
"reward_std": 0.29591915011405945,
|
| 267 |
+
"rewards/equation_reward_func": 0.09375,
|
| 268 |
+
"rewards/format_reward_func": 0.9296875,
|
| 269 |
+
"step": 40
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"completion_length": 428.75,
|
| 273 |
+
"epoch": 0.007466666666666667,
|
| 274 |
+
"grad_norm": 0.10965524897254234,
|
| 275 |
+
"kl": 0.01348876953125,
|
| 276 |
+
"learning_rate": 4.949291683053768e-07,
|
| 277 |
+
"loss": 0.0,
|
| 278 |
+
"reward": 1.0703125,
|
| 279 |
+
"reward_std": 0.3557952530682087,
|
| 280 |
+
"rewards/equation_reward_func": 0.1328125,
|
| 281 |
+
"rewards/format_reward_func": 0.9375,
|
| 282 |
+
"step": 42
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"completion_length": 483.65625,
|
| 286 |
+
"epoch": 0.007822222222222222,
|
| 287 |
+
"grad_norm": 0.07034266397094524,
|
| 288 |
+
"kl": 0.019378662109375,
|
| 289 |
+
"learning_rate": 4.941818151059955e-07,
|
| 290 |
+
"loss": 0.0,
|
| 291 |
+
"reward": 0.9921875,
|
| 292 |
+
"reward_std": 0.182886628434062,
|
| 293 |
+
"rewards/equation_reward_func": 0.0390625,
|
| 294 |
+
"rewards/format_reward_func": 0.953125,
|
| 295 |
+
"step": 44
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"completion_length": 425.375,
|
| 299 |
+
"epoch": 0.008177777777777779,
|
| 300 |
+
"grad_norm": 0.14631602871481755,
|
| 301 |
+
"kl": 0.01434326171875,
|
| 302 |
+
"learning_rate": 4.933837520293017e-07,
|
| 303 |
+
"loss": 0.0,
|
| 304 |
+
"reward": 1.0625,
|
| 305 |
+
"reward_std": 0.35122798942029476,
|
| 306 |
+
"rewards/equation_reward_func": 0.1328125,
|
| 307 |
+
"rewards/format_reward_func": 0.9296875,
|
| 308 |
+
"step": 46
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"completion_length": 435.6328125,
|
| 312 |
+
"epoch": 0.008533333333333334,
|
| 313 |
+
"grad_norm": 0.16370041909573974,
|
| 314 |
+
"kl": 0.01432037353515625,
|
| 315 |
+
"learning_rate": 4.925351448111454e-07,
|
| 316 |
+
"loss": 0.0,
|
| 317 |
+
"reward": 1.109375,
|
| 318 |
+
"reward_std": 0.3174794539809227,
|
| 319 |
+
"rewards/equation_reward_func": 0.1640625,
|
| 320 |
+
"rewards/format_reward_func": 0.9453125,
|
| 321 |
+
"step": 48
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"completion_length": 442.1171875,
|
| 325 |
+
"epoch": 0.008888888888888889,
|
| 326 |
+
"grad_norm": 0.1311062256072383,
|
| 327 |
+
"kl": 0.014678955078125,
|
| 328 |
+
"learning_rate": 4.91636169684011e-07,
|
| 329 |
+
"loss": 0.0,
|
| 330 |
+
"reward": 1.0703125,
|
| 331 |
+
"reward_std": 0.30886387825012207,
|
| 332 |
+
"rewards/equation_reward_func": 0.125,
|
| 333 |
+
"rewards/format_reward_func": 0.9453125,
|
| 334 |
+
"step": 50
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"completion_length": 430.7578125,
|
| 338 |
+
"epoch": 0.009244444444444444,
|
| 339 |
+
"grad_norm": 0.13267168261468218,
|
| 340 |
+
"kl": 0.01804351806640625,
|
| 341 |
+
"learning_rate": 4.906870133404186e-07,
|
| 342 |
+
"loss": 0.0,
|
| 343 |
+
"reward": 1.0703125,
|
| 344 |
+
"reward_std": 0.2635674234479666,
|
| 345 |
+
"rewards/equation_reward_func": 0.109375,
|
| 346 |
+
"rewards/format_reward_func": 0.9609375,
|
| 347 |
+
"step": 52
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"completion_length": 411.578125,
|
| 351 |
+
"epoch": 0.0096,
|
| 352 |
+
"grad_norm": 0.17179126795155464,
|
| 353 |
+
"kl": 0.01618194580078125,
|
| 354 |
+
"learning_rate": 4.896878728941531e-07,
|
| 355 |
+
"loss": 0.0,
|
| 356 |
+
"reward": 1.140625,
|
| 357 |
+
"reward_std": 0.29143064096570015,
|
| 358 |
+
"rewards/equation_reward_func": 0.15625,
|
| 359 |
+
"rewards/format_reward_func": 0.984375,
|
| 360 |
+
"step": 54
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"completion_length": 398.4453125,
|
| 364 |
+
"epoch": 0.009955555555555556,
|
| 365 |
+
"grad_norm": 0.14196611468000556,
|
| 366 |
+
"kl": 0.016681671142578125,
|
| 367 |
+
"learning_rate": 4.886389558393284e-07,
|
| 368 |
+
"loss": 0.0,
|
| 369 |
+
"reward": 1.109375,
|
| 370 |
+
"reward_std": 0.26196980848908424,
|
| 371 |
+
"rewards/equation_reward_func": 0.125,
|
| 372 |
+
"rewards/format_reward_func": 0.984375,
|
| 373 |
+
"step": 56
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"completion_length": 389.7890625,
|
| 377 |
+
"epoch": 0.010311111111111111,
|
| 378 |
+
"grad_norm": 0.1476241924443653,
|
| 379 |
+
"kl": 0.019023895263671875,
|
| 380 |
+
"learning_rate": 4.875404800072976e-07,
|
| 381 |
+
"loss": 0.0,
|
| 382 |
+
"reward": 1.1484375,
|
| 383 |
+
"reward_std": 0.3381521496921778,
|
| 384 |
+
"rewards/equation_reward_func": 0.171875,
|
| 385 |
+
"rewards/format_reward_func": 0.9765625,
|
| 386 |
+
"step": 58
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"completion_length": 388.6015625,
|
| 390 |
+
"epoch": 0.010666666666666666,
|
| 391 |
+
"grad_norm": 0.11394197670203084,
|
| 392 |
+
"kl": 0.0216827392578125,
|
| 393 |
+
"learning_rate": 4.86392673521415e-07,
|
| 394 |
+
"loss": 0.0,
|
| 395 |
+
"reward": 1.0703125,
|
| 396 |
+
"reward_std": 0.22548809088766575,
|
| 397 |
+
"rewards/equation_reward_func": 0.09375,
|
| 398 |
+
"rewards/format_reward_func": 0.9765625,
|
| 399 |
+
"step": 60
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"completion_length": 368.5625,
|
| 403 |
+
"epoch": 0.011022222222222221,
|
| 404 |
+
"grad_norm": 0.13423259129208498,
|
| 405 |
+
"kl": 0.018878936767578125,
|
| 406 |
+
"learning_rate": 4.851957747496606e-07,
|
| 407 |
+
"loss": 0.0,
|
| 408 |
+
"reward": 1.140625,
|
| 409 |
+
"reward_std": 0.30397036112844944,
|
| 410 |
+
"rewards/equation_reward_func": 0.1640625,
|
| 411 |
+
"rewards/format_reward_func": 0.9765625,
|
| 412 |
+
"step": 62
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"completion_length": 400.7734375,
|
| 416 |
+
"epoch": 0.011377777777777778,
|
| 417 |
+
"grad_norm": 0.08843954217928876,
|
| 418 |
+
"kl": 0.0186004638671875,
|
| 419 |
+
"learning_rate": 4.839500322551386e-07,
|
| 420 |
+
"loss": 0.0,
|
| 421 |
+
"reward": 1.0625,
|
| 422 |
+
"reward_std": 0.25172007270157337,
|
| 423 |
+
"rewards/equation_reward_func": 0.1015625,
|
| 424 |
+
"rewards/format_reward_func": 0.9609375,
|
| 425 |
+
"step": 64
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"completion_length": 393.4296875,
|
| 429 |
+
"epoch": 0.011733333333333333,
|
| 430 |
+
"grad_norm": 0.11714237597674881,
|
| 431 |
+
"kl": 0.02496337890625,
|
| 432 |
+
"learning_rate": 4.826557047444563e-07,
|
| 433 |
+
"loss": 0.0,
|
| 434 |
+
"reward": 1.1640625,
|
| 435 |
+
"reward_std": 0.27939521707594395,
|
| 436 |
+
"rewards/equation_reward_func": 0.1953125,
|
| 437 |
+
"rewards/format_reward_func": 0.96875,
|
| 438 |
+
"step": 66
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"completion_length": 417.25,
|
| 442 |
+
"epoch": 0.012088888888888889,
|
| 443 |
+
"grad_norm": 0.1544015474537589,
|
| 444 |
+
"kl": 0.020420074462890625,
|
| 445 |
+
"learning_rate": 4.813130610139993e-07,
|
| 446 |
+
"loss": 0.0,
|
| 447 |
+
"reward": 1.109375,
|
| 448 |
+
"reward_std": 0.33851570822298527,
|
| 449 |
+
"rewards/equation_reward_func": 0.1484375,
|
| 450 |
+
"rewards/format_reward_func": 0.9609375,
|
| 451 |
+
"step": 68
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"completion_length": 432.0703125,
|
| 455 |
+
"epoch": 0.012444444444444444,
|
| 456 |
+
"grad_norm": 0.12342065446113715,
|
| 457 |
+
"kl": 0.023410797119140625,
|
| 458 |
+
"learning_rate": 4.799223798941089e-07,
|
| 459 |
+
"loss": 0.0,
|
| 460 |
+
"reward": 1.09375,
|
| 461 |
+
"reward_std": 0.36288236267864704,
|
| 462 |
+
"rewards/equation_reward_func": 0.1484375,
|
| 463 |
+
"rewards/format_reward_func": 0.9453125,
|
| 464 |
+
"step": 70
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"completion_length": 389.6640625,
|
| 468 |
+
"epoch": 0.0128,
|
| 469 |
+
"grad_norm": 0.14656284486731108,
|
| 470 |
+
"kl": 0.0208740234375,
|
| 471 |
+
"learning_rate": 4.78483950191177e-07,
|
| 472 |
+
"loss": 0.0,
|
| 473 |
+
"reward": 1.125,
|
| 474 |
+
"reward_std": 0.2977413833141327,
|
| 475 |
+
"rewards/equation_reward_func": 0.140625,
|
| 476 |
+
"rewards/format_reward_func": 0.984375,
|
| 477 |
+
"step": 72
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"completion_length": 386.0234375,
|
| 481 |
+
"epoch": 0.013155555555555556,
|
| 482 |
+
"grad_norm": 0.1322785515012084,
|
| 483 |
+
"kl": 0.0208892822265625,
|
| 484 |
+
"learning_rate": 4.769980706276687e-07,
|
| 485 |
+
"loss": 0.0,
|
| 486 |
+
"reward": 1.1875,
|
| 487 |
+
"reward_std": 0.23033758997917175,
|
| 488 |
+
"rewards/equation_reward_func": 0.1953125,
|
| 489 |
+
"rewards/format_reward_func": 0.9921875,
|
| 490 |
+
"step": 74
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"completion_length": 391.1171875,
|
| 494 |
+
"epoch": 0.013511111111111111,
|
| 495 |
+
"grad_norm": 0.137897578511538,
|
| 496 |
+
"kl": 0.02099609375,
|
| 497 |
+
"learning_rate": 4.7546504978008595e-07,
|
| 498 |
+
"loss": 0.0,
|
| 499 |
+
"reward": 1.1640625,
|
| 500 |
+
"reward_std": 0.3311441335827112,
|
| 501 |
+
"rewards/equation_reward_func": 0.1953125,
|
| 502 |
+
"rewards/format_reward_func": 0.96875,
|
| 503 |
+
"step": 76
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"completion_length": 418.8125,
|
| 507 |
+
"epoch": 0.013866666666666666,
|
| 508 |
+
"grad_norm": 0.10892568772438567,
|
| 509 |
+
"kl": 0.021595001220703125,
|
| 510 |
+
"learning_rate": 4.738852060148848e-07,
|
| 511 |
+
"loss": 0.0,
|
| 512 |
+
"reward": 1.09375,
|
| 513 |
+
"reward_std": 0.2359209954738617,
|
| 514 |
+
"rewards/equation_reward_func": 0.1328125,
|
| 515 |
+
"rewards/format_reward_func": 0.9609375,
|
| 516 |
+
"step": 78
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"completion_length": 369.0,
|
| 520 |
+
"epoch": 0.014222222222222223,
|
| 521 |
+
"grad_norm": 0.12919106909292583,
|
| 522 |
+
"kl": 0.02898406982421875,
|
| 523 |
+
"learning_rate": 4.722588674223593e-07,
|
| 524 |
+
"loss": 0.0,
|
| 525 |
+
"reward": 1.1484375,
|
| 526 |
+
"reward_std": 0.21382390893995762,
|
| 527 |
+
"rewards/equation_reward_func": 0.1484375,
|
| 528 |
+
"rewards/format_reward_func": 1.0,
|
| 529 |
+
"step": 80
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"completion_length": 370.3515625,
|
| 533 |
+
"epoch": 0.014577777777777778,
|
| 534 |
+
"grad_norm": 0.15310962735407482,
|
| 535 |
+
"kl": 0.02521514892578125,
|
| 536 |
+
"learning_rate": 4.70586371748506e-07,
|
| 537 |
+
"loss": 0.0,
|
| 538 |
+
"reward": 1.1484375,
|
| 539 |
+
"reward_std": 0.36273445934057236,
|
| 540 |
+
"rewards/equation_reward_func": 0.1875,
|
| 541 |
+
"rewards/format_reward_func": 0.9609375,
|
| 542 |
+
"step": 82
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"completion_length": 396.390625,
|
| 546 |
+
"epoch": 0.014933333333333333,
|
| 547 |
+
"grad_norm": 0.13989921599508573,
|
| 548 |
+
"kl": 0.0299530029296875,
|
| 549 |
+
"learning_rate": 4.6886806632488363e-07,
|
| 550 |
+
"loss": 0.0,
|
| 551 |
+
"reward": 1.1171875,
|
| 552 |
+
"reward_std": 0.3056321032345295,
|
| 553 |
+
"rewards/equation_reward_func": 0.140625,
|
| 554 |
+
"rewards/format_reward_func": 0.9765625,
|
| 555 |
+
"step": 84
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"completion_length": 400.7734375,
|
| 559 |
+
"epoch": 0.015288888888888888,
|
| 560 |
+
"grad_norm": 0.16427882618928974,
|
| 561 |
+
"kl": 0.0267181396484375,
|
| 562 |
+
"learning_rate": 4.6710430799648143e-07,
|
| 563 |
+
"loss": 0.0,
|
| 564 |
+
"reward": 1.21875,
|
| 565 |
+
"reward_std": 0.3713865801692009,
|
| 566 |
+
"rewards/equation_reward_func": 0.2421875,
|
| 567 |
+
"rewards/format_reward_func": 0.9765625,
|
| 568 |
+
"step": 86
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"completion_length": 395.515625,
|
| 572 |
+
"epoch": 0.015644444444444443,
|
| 573 |
+
"grad_norm": 0.15324522374897853,
|
| 574 |
+
"kl": 0.030181884765625,
|
| 575 |
+
"learning_rate": 4.652954630476127e-07,
|
| 576 |
+
"loss": 0.0,
|
| 577 |
+
"reward": 1.1484375,
|
| 578 |
+
"reward_std": 0.25547589734196663,
|
| 579 |
+
"rewards/equation_reward_func": 0.171875,
|
| 580 |
+
"rewards/format_reward_func": 0.9765625,
|
| 581 |
+
"step": 88
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"completion_length": 425.8984375,
|
| 585 |
+
"epoch": 0.016,
|
| 586 |
+
"grad_norm": 0.1254448594399879,
|
| 587 |
+
"kl": 0.0306243896484375,
|
| 588 |
+
"learning_rate": 4.6344190712584713e-07,
|
| 589 |
+
"loss": 0.0,
|
| 590 |
+
"reward": 1.046875,
|
| 591 |
+
"reward_std": 0.23911622539162636,
|
| 592 |
+
"rewards/equation_reward_func": 0.078125,
|
| 593 |
+
"rewards/format_reward_func": 0.96875,
|
| 594 |
+
"step": 90
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"completion_length": 392.8828125,
|
| 598 |
+
"epoch": 0.016355555555555557,
|
| 599 |
+
"grad_norm": 0.11207079919662942,
|
| 600 |
+
"kl": 0.03009033203125,
|
| 601 |
+
"learning_rate": 4.615440251639995e-07,
|
| 602 |
+
"loss": 0.0,
|
| 603 |
+
"reward": 1.1484375,
|
| 604 |
+
"reward_std": 0.2369818277657032,
|
| 605 |
+
"rewards/equation_reward_func": 0.15625,
|
| 606 |
+
"rewards/format_reward_func": 0.9921875,
|
| 607 |
+
"step": 92
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"completion_length": 384.5546875,
|
| 611 |
+
"epoch": 0.01671111111111111,
|
| 612 |
+
"grad_norm": 0.14199339358701857,
|
| 613 |
+
"kl": 0.04071044921875,
|
| 614 |
+
"learning_rate": 4.596022113001894e-07,
|
| 615 |
+
"loss": 0.0,
|
| 616 |
+
"reward": 1.25,
|
| 617 |
+
"reward_std": 0.3733201716095209,
|
| 618 |
+
"rewards/equation_reward_func": 0.28125,
|
| 619 |
+
"rewards/format_reward_func": 0.96875,
|
| 620 |
+
"step": 94
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"completion_length": 437.5625,
|
| 624 |
+
"epoch": 0.017066666666666667,
|
| 625 |
+
"grad_norm": 0.12806187866277957,
|
| 626 |
+
"kl": 0.041961669921875,
|
| 627 |
+
"learning_rate": 4.576168687959895e-07,
|
| 628 |
+
"loss": 0.0,
|
| 629 |
+
"reward": 1.1015625,
|
| 630 |
+
"reward_std": 0.20016413368284702,
|
| 631 |
+
"rewards/equation_reward_func": 0.109375,
|
| 632 |
+
"rewards/format_reward_func": 0.9921875,
|
| 633 |
+
"step": 96
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"completion_length": 384.7734375,
|
| 637 |
+
"epoch": 0.01742222222222222,
|
| 638 |
+
"grad_norm": 0.16560343482417905,
|
| 639 |
+
"kl": 0.0358428955078125,
|
| 640 |
+
"learning_rate": 4.555884099526793e-07,
|
| 641 |
+
"loss": 0.0,
|
| 642 |
+
"reward": 1.2109375,
|
| 643 |
+
"reward_std": 0.3128028716892004,
|
| 644 |
+
"rewards/equation_reward_func": 0.21875,
|
| 645 |
+
"rewards/format_reward_func": 0.9921875,
|
| 646 |
+
"step": 98
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"completion_length": 377.8046875,
|
| 650 |
+
"epoch": 0.017777777777777778,
|
| 651 |
+
"grad_norm": 0.06544152830497535,
|
| 652 |
+
"kl": 0.0405426025390625,
|
| 653 |
+
"learning_rate": 4.5351725602562174e-07,
|
| 654 |
+
"loss": 0.0,
|
| 655 |
+
"reward": 1.109375,
|
| 656 |
+
"reward_std": 0.14123954437673092,
|
| 657 |
+
"rewards/equation_reward_func": 0.1171875,
|
| 658 |
+
"rewards/format_reward_func": 0.9921875,
|
| 659 |
+
"step": 100
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"completion_length": 360.8828125,
|
| 663 |
+
"epoch": 0.018133333333333335,
|
| 664 |
+
"grad_norm": 0.11192315546197115,
|
| 665 |
+
"kl": 0.037384033203125,
|
| 666 |
+
"learning_rate": 4.514038371367791e-07,
|
| 667 |
+
"loss": 0.0,
|
| 668 |
+
"reward": 1.15625,
|
| 669 |
+
"reward_std": 0.2957183439284563,
|
| 670 |
+
"rewards/equation_reward_func": 0.1796875,
|
| 671 |
+
"rewards/format_reward_func": 0.9765625,
|
| 672 |
+
"step": 102
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"completion_length": 355.203125,
|
| 676 |
+
"epoch": 0.018488888888888888,
|
| 677 |
+
"grad_norm": 0.12420501860194036,
|
| 678 |
+
"kl": 0.06029510498046875,
|
| 679 |
+
"learning_rate": 4.4924859218538936e-07,
|
| 680 |
+
"loss": 0.0001,
|
| 681 |
+
"reward": 1.15625,
|
| 682 |
+
"reward_std": 0.25118329003453255,
|
| 683 |
+
"rewards/equation_reward_func": 0.171875,
|
| 684 |
+
"rewards/format_reward_func": 0.984375,
|
| 685 |
+
"step": 104
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"completion_length": 378.75,
|
| 689 |
+
"epoch": 0.018844444444444445,
|
| 690 |
+
"grad_norm": 0.12758866020524365,
|
| 691 |
+
"kl": 0.0380859375,
|
| 692 |
+
"learning_rate": 4.470519687568185e-07,
|
| 693 |
+
"loss": 0.0,
|
| 694 |
+
"reward": 1.1015625,
|
| 695 |
+
"reward_std": 0.2606765292584896,
|
| 696 |
+
"rewards/equation_reward_func": 0.1171875,
|
| 697 |
+
"rewards/format_reward_func": 0.984375,
|
| 698 |
+
"step": 106
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"completion_length": 380.8203125,
|
| 702 |
+
"epoch": 0.0192,
|
| 703 |
+
"grad_norm": 0.16123721776657848,
|
| 704 |
+
"kl": 0.04666900634765625,
|
| 705 |
+
"learning_rate": 4.4481442302960923e-07,
|
| 706 |
+
"loss": 0.0,
|
| 707 |
+
"reward": 1.1171875,
|
| 708 |
+
"reward_std": 0.2664504889398813,
|
| 709 |
+
"rewards/equation_reward_func": 0.1328125,
|
| 710 |
+
"rewards/format_reward_func": 0.984375,
|
| 711 |
+
"step": 108
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"completion_length": 384.2578125,
|
| 715 |
+
"epoch": 0.019555555555555555,
|
| 716 |
+
"grad_norm": 0.1514840733121254,
|
| 717 |
+
"kl": 0.0447998046875,
|
| 718 |
+
"learning_rate": 4.4253641968074505e-07,
|
| 719 |
+
"loss": 0.0,
|
| 720 |
+
"reward": 1.2265625,
|
| 721 |
+
"reward_std": 0.30851516872644424,
|
| 722 |
+
"rewards/equation_reward_func": 0.2265625,
|
| 723 |
+
"rewards/format_reward_func": 1.0,
|
| 724 |
+
"step": 110
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"completion_length": 374.7578125,
|
| 728 |
+
"epoch": 0.019911111111111112,
|
| 729 |
+
"grad_norm": 0.13918628422029689,
|
| 730 |
+
"kl": 0.0448150634765625,
|
| 731 |
+
"learning_rate": 4.402184317891501e-07,
|
| 732 |
+
"loss": 0.0,
|
| 733 |
+
"reward": 1.2109375,
|
| 734 |
+
"reward_std": 0.32930916734039783,
|
| 735 |
+
"rewards/equation_reward_func": 0.234375,
|
| 736 |
+
"rewards/format_reward_func": 0.9765625,
|
| 737 |
+
"step": 112
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"completion_length": 365.015625,
|
| 741 |
+
"epoch": 0.020266666666666665,
|
| 742 |
+
"grad_norm": 0.1357853470773616,
|
| 743 |
+
"kl": 0.0449066162109375,
|
| 744 |
+
"learning_rate": 4.37860940737443e-07,
|
| 745 |
+
"loss": 0.0,
|
| 746 |
+
"reward": 1.3125,
|
| 747 |
+
"reward_std": 0.3374421428889036,
|
| 748 |
+
"rewards/equation_reward_func": 0.3203125,
|
| 749 |
+
"rewards/format_reward_func": 0.9921875,
|
| 750 |
+
"step": 114
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"completion_length": 373.3671875,
|
| 754 |
+
"epoch": 0.020622222222222222,
|
| 755 |
+
"grad_norm": 0.14960512254555802,
|
| 756 |
+
"kl": 0.0490875244140625,
|
| 757 |
+
"learning_rate": 4.354644361119671e-07,
|
| 758 |
+
"loss": 0.0,
|
| 759 |
+
"reward": 1.1796875,
|
| 760 |
+
"reward_std": 0.21382390893995762,
|
| 761 |
+
"rewards/equation_reward_func": 0.1875,
|
| 762 |
+
"rewards/format_reward_func": 0.9921875,
|
| 763 |
+
"step": 116
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"completion_length": 378.625,
|
| 767 |
+
"epoch": 0.02097777777777778,
|
| 768 |
+
"grad_norm": 0.1510812591190034,
|
| 769 |
+
"kl": 0.0518798828125,
|
| 770 |
+
"learning_rate": 4.3302941560111716e-07,
|
| 771 |
+
"loss": 0.0001,
|
| 772 |
+
"reward": 1.2734375,
|
| 773 |
+
"reward_std": 0.31605015881359577,
|
| 774 |
+
"rewards/equation_reward_func": 0.28125,
|
| 775 |
+
"rewards/format_reward_func": 0.9921875,
|
| 776 |
+
"step": 118
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"completion_length": 380.46875,
|
| 780 |
+
"epoch": 0.021333333333333333,
|
| 781 |
+
"grad_norm": 0.14734857757565625,
|
| 782 |
+
"kl": 0.0499420166015625,
|
| 783 |
+
"learning_rate": 4.3055638489198236e-07,
|
| 784 |
+
"loss": 0.0,
|
| 785 |
+
"reward": 1.1484375,
|
| 786 |
+
"reward_std": 0.28012007288634777,
|
| 787 |
+
"rewards/equation_reward_func": 0.171875,
|
| 788 |
+
"rewards/format_reward_func": 0.9765625,
|
| 789 |
+
"step": 120
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"completion_length": 378.453125,
|
| 793 |
+
"epoch": 0.02168888888888889,
|
| 794 |
+
"grad_norm": 0.1541652488299738,
|
| 795 |
+
"kl": 0.0516815185546875,
|
| 796 |
+
"learning_rate": 4.280458575653296e-07,
|
| 797 |
+
"loss": 0.0001,
|
| 798 |
+
"reward": 1.21875,
|
| 799 |
+
"reward_std": 0.30403448827564716,
|
| 800 |
+
"rewards/equation_reward_func": 0.234375,
|
| 801 |
+
"rewards/format_reward_func": 0.984375,
|
| 802 |
+
"step": 122
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"completion_length": 331.6953125,
|
| 806 |
+
"epoch": 0.022044444444444443,
|
| 807 |
+
"grad_norm": 0.1800491418335961,
|
| 808 |
+
"kl": 0.0565338134765625,
|
| 809 |
+
"learning_rate": 4.2549835498894665e-07,
|
| 810 |
+
"loss": 0.0001,
|
| 811 |
+
"reward": 1.34375,
|
| 812 |
+
"reward_std": 0.34010059013962746,
|
| 813 |
+
"rewards/equation_reward_func": 0.3515625,
|
| 814 |
+
"rewards/format_reward_func": 0.9921875,
|
| 815 |
+
"step": 124
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"completion_length": 348.4140625,
|
| 819 |
+
"epoch": 0.0224,
|
| 820 |
+
"grad_norm": 0.1794948955818512,
|
| 821 |
+
"kl": 0.057586669921875,
|
| 822 |
+
"learning_rate": 4.229144062093679e-07,
|
| 823 |
+
"loss": 0.0001,
|
| 824 |
+
"reward": 1.296875,
|
| 825 |
+
"reward_std": 0.3322049658745527,
|
| 826 |
+
"rewards/equation_reward_func": 0.3046875,
|
| 827 |
+
"rewards/format_reward_func": 0.9921875,
|
| 828 |
+
"step": 126
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"completion_length": 388.171875,
|
| 832 |
+
"epoch": 0.022755555555555557,
|
| 833 |
+
"grad_norm": 0.143172951066265,
|
| 834 |
+
"kl": 0.0540313720703125,
|
| 835 |
+
"learning_rate": 4.2029454784200675e-07,
|
| 836 |
+
"loss": 0.0001,
|
| 837 |
+
"reward": 1.1875,
|
| 838 |
+
"reward_std": 0.31116871163249016,
|
| 839 |
+
"rewards/equation_reward_func": 0.203125,
|
| 840 |
+
"rewards/format_reward_func": 0.984375,
|
| 841 |
+
"step": 128
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"completion_length": 328.7578125,
|
| 845 |
+
"epoch": 0.02311111111111111,
|
| 846 |
+
"grad_norm": 0.15975290384389043,
|
| 847 |
+
"kl": 0.073944091796875,
|
| 848 |
+
"learning_rate": 4.1763932395971433e-07,
|
| 849 |
+
"loss": 0.0001,
|
| 850 |
+
"reward": 1.3359375,
|
| 851 |
+
"reward_std": 0.35031135752797127,
|
| 852 |
+
"rewards/equation_reward_func": 0.3515625,
|
| 853 |
+
"rewards/format_reward_func": 0.984375,
|
| 854 |
+
"step": 130
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"completion_length": 392.6953125,
|
| 858 |
+
"epoch": 0.023466666666666667,
|
| 859 |
+
"grad_norm": 0.15405003814264823,
|
| 860 |
+
"kl": 0.0763092041015625,
|
| 861 |
+
"learning_rate": 4.1494928597979117e-07,
|
| 862 |
+
"loss": 0.0001,
|
| 863 |
+
"reward": 1.203125,
|
| 864 |
+
"reward_std": 0.3124668952077627,
|
| 865 |
+
"rewards/equation_reward_func": 0.2421875,
|
| 866 |
+
"rewards/format_reward_func": 0.9609375,
|
| 867 |
+
"step": 132
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"completion_length": 353.1484375,
|
| 871 |
+
"epoch": 0.023822222222222224,
|
| 872 |
+
"grad_norm": 0.19135038627848672,
|
| 873 |
+
"kl": 0.0674285888671875,
|
| 874 |
+
"learning_rate": 4.122249925494726e-07,
|
| 875 |
+
"loss": 0.0001,
|
| 876 |
+
"reward": 1.203125,
|
| 877 |
+
"reward_std": 0.2993340939283371,
|
| 878 |
+
"rewards/equation_reward_func": 0.203125,
|
| 879 |
+
"rewards/format_reward_func": 1.0,
|
| 880 |
+
"step": 134
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"completion_length": 381.6796875,
|
| 884 |
+
"epoch": 0.024177777777777777,
|
| 885 |
+
"grad_norm": 0.15650361206530433,
|
| 886 |
+
"kl": 0.065765380859375,
|
| 887 |
+
"learning_rate": 4.094670094299131e-07,
|
| 888 |
+
"loss": 0.0001,
|
| 889 |
+
"reward": 1.2734375,
|
| 890 |
+
"reward_std": 0.34999459609389305,
|
| 891 |
+
"rewards/equation_reward_func": 0.3125,
|
| 892 |
+
"rewards/format_reward_func": 0.9609375,
|
| 893 |
+
"step": 136
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"completion_length": 408.96875,
|
| 897 |
+
"epoch": 0.024533333333333334,
|
| 898 |
+
"grad_norm": 0.1551337440537955,
|
| 899 |
+
"kl": 0.0649566650390625,
|
| 900 |
+
"learning_rate": 4.066759093786931e-07,
|
| 901 |
+
"loss": 0.0001,
|
| 902 |
+
"reward": 1.2734375,
|
| 903 |
+
"reward_std": 0.3595039639621973,
|
| 904 |
+
"rewards/equation_reward_func": 0.3125,
|
| 905 |
+
"rewards/format_reward_func": 0.9609375,
|
| 906 |
+
"step": 138
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"completion_length": 365.265625,
|
| 910 |
+
"epoch": 0.024888888888888887,
|
| 911 |
+
"grad_norm": 0.14349181297141309,
|
| 912 |
+
"kl": 0.0655364990234375,
|
| 913 |
+
"learning_rate": 4.038522720308732e-07,
|
| 914 |
+
"loss": 0.0001,
|
| 915 |
+
"reward": 1.34375,
|
| 916 |
+
"reward_std": 0.28242738731205463,
|
| 917 |
+
"rewards/equation_reward_func": 0.359375,
|
| 918 |
+
"rewards/format_reward_func": 0.984375,
|
| 919 |
+
"step": 140
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"completion_length": 417.859375,
|
| 923 |
+
"epoch": 0.025244444444444444,
|
| 924 |
+
"grad_norm": 0.12417767012343875,
|
| 925 |
+
"kl": 0.0717010498046875,
|
| 926 |
+
"learning_rate": 4.009966837786194e-07,
|
| 927 |
+
"loss": 0.0001,
|
| 928 |
+
"reward": 1.1484375,
|
| 929 |
+
"reward_std": 0.2211486864835024,
|
| 930 |
+
"rewards/equation_reward_func": 0.15625,
|
| 931 |
+
"rewards/format_reward_func": 0.9921875,
|
| 932 |
+
"step": 142
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"completion_length": 413.171875,
|
| 936 |
+
"epoch": 0.0256,
|
| 937 |
+
"grad_norm": 0.15012499276377514,
|
| 938 |
+
"kl": 0.0660858154296875,
|
| 939 |
+
"learning_rate": 3.981097376494259e-07,
|
| 940 |
+
"loss": 0.0001,
|
| 941 |
+
"reward": 1.2734375,
|
| 942 |
+
"reward_std": 0.33003650419414043,
|
| 943 |
+
"rewards/equation_reward_func": 0.2890625,
|
| 944 |
+
"rewards/format_reward_func": 0.984375,
|
| 945 |
+
"step": 144
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"completion_length": 359.765625,
|
| 949 |
+
"epoch": 0.025955555555555555,
|
| 950 |
+
"grad_norm": 0.15995574283554995,
|
| 951 |
+
"kl": 0.0708770751953125,
|
| 952 |
+
"learning_rate": 3.951920331829592e-07,
|
| 953 |
+
"loss": 0.0001,
|
| 954 |
+
"reward": 1.2421875,
|
| 955 |
+
"reward_std": 0.277237007394433,
|
| 956 |
+
"rewards/equation_reward_func": 0.2734375,
|
| 957 |
+
"rewards/format_reward_func": 0.96875,
|
| 958 |
+
"step": 146
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"completion_length": 417.5078125,
|
| 962 |
+
"epoch": 0.02631111111111111,
|
| 963 |
+
"grad_norm": 0.17208499354721293,
|
| 964 |
+
"kl": 0.0723419189453125,
|
| 965 |
+
"learning_rate": 3.922441763065506e-07,
|
| 966 |
+
"loss": 0.0001,
|
| 967 |
+
"reward": 1.234375,
|
| 968 |
+
"reward_std": 0.3679391499608755,
|
| 969 |
+
"rewards/equation_reward_func": 0.2734375,
|
| 970 |
+
"rewards/format_reward_func": 0.9609375,
|
| 971 |
+
"step": 148
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"completion_length": 417.3671875,
|
| 975 |
+
"epoch": 0.02666666666666667,
|
| 976 |
+
"grad_norm": 0.17080855226672484,
|
| 977 |
+
"kl": 0.0657806396484375,
|
| 978 |
+
"learning_rate": 3.8926677920936093e-07,
|
| 979 |
+
"loss": 0.0001,
|
| 980 |
+
"reward": 1.2109375,
|
| 981 |
+
"reward_std": 0.37137173116207123,
|
| 982 |
+
"rewards/equation_reward_func": 0.2578125,
|
| 983 |
+
"rewards/format_reward_func": 0.953125,
|
| 984 |
+
"step": 150
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"completion_length": 452.265625,
|
| 988 |
+
"epoch": 0.027022222222222222,
|
| 989 |
+
"grad_norm": 0.13098468802735794,
|
| 990 |
+
"kl": 0.062225341796875,
|
| 991 |
+
"learning_rate": 3.862604602152464e-07,
|
| 992 |
+
"loss": 0.0001,
|
| 993 |
+
"reward": 1.3046875,
|
| 994 |
+
"reward_std": 0.3546949904412031,
|
| 995 |
+
"rewards/equation_reward_func": 0.3515625,
|
| 996 |
+
"rewards/format_reward_func": 0.953125,
|
| 997 |
+
"step": 152
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"completion_length": 456.4609375,
|
| 1001 |
+
"epoch": 0.02737777777777778,
|
| 1002 |
+
"grad_norm": 0.13309465770298312,
|
| 1003 |
+
"kl": 0.0681304931640625,
|
| 1004 |
+
"learning_rate": 3.8322584365434934e-07,
|
| 1005 |
+
"loss": 0.0001,
|
| 1006 |
+
"reward": 1.25,
|
| 1007 |
+
"reward_std": 0.3172913808375597,
|
| 1008 |
+
"rewards/equation_reward_func": 0.3046875,
|
| 1009 |
+
"rewards/format_reward_func": 0.9453125,
|
| 1010 |
+
"step": 154
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"completion_length": 414.6875,
|
| 1014 |
+
"epoch": 0.027733333333333332,
|
| 1015 |
+
"grad_norm": 0.1765554041111672,
|
| 1016 |
+
"kl": 0.0706787109375,
|
| 1017 |
+
"learning_rate": 3.8016355973344173e-07,
|
| 1018 |
+
"loss": 0.0001,
|
| 1019 |
+
"reward": 1.2578125,
|
| 1020 |
+
"reward_std": 0.3864140044897795,
|
| 1021 |
+
"rewards/equation_reward_func": 0.296875,
|
| 1022 |
+
"rewards/format_reward_func": 0.9609375,
|
| 1023 |
+
"step": 156
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"completion_length": 419.3203125,
|
| 1027 |
+
"epoch": 0.02808888888888889,
|
| 1028 |
+
"grad_norm": 0.1525317919499253,
|
| 1029 |
+
"kl": 0.0763397216796875,
|
| 1030 |
+
"learning_rate": 3.7707424440504863e-07,
|
| 1031 |
+
"loss": 0.0001,
|
| 1032 |
+
"reward": 1.21875,
|
| 1033 |
+
"reward_std": 0.32123037427663803,
|
| 1034 |
+
"rewards/equation_reward_func": 0.2578125,
|
| 1035 |
+
"rewards/format_reward_func": 0.9609375,
|
| 1036 |
+
"step": 158
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"completion_length": 458.296875,
|
| 1040 |
+
"epoch": 0.028444444444444446,
|
| 1041 |
+
"grad_norm": 0.13661366975115602,
|
| 1042 |
+
"kl": 0.0710906982421875,
|
| 1043 |
+
"learning_rate": 3.739585392353787e-07,
|
| 1044 |
+
"loss": 0.0001,
|
| 1045 |
+
"reward": 1.171875,
|
| 1046 |
+
"reward_std": 0.3388565890491009,
|
| 1047 |
+
"rewards/equation_reward_func": 0.25,
|
| 1048 |
+
"rewards/format_reward_func": 0.921875,
|
| 1049 |
+
"step": 160
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"completion_length": 407.5234375,
|
| 1053 |
+
"epoch": 0.0288,
|
| 1054 |
+
"grad_norm": 0.14265288581900928,
|
| 1055 |
+
"kl": 0.090179443359375,
|
| 1056 |
+
"learning_rate": 3.7081709127108767e-07,
|
| 1057 |
+
"loss": 0.0001,
|
| 1058 |
+
"reward": 1.34375,
|
| 1059 |
+
"reward_std": 0.4125017859041691,
|
| 1060 |
+
"rewards/equation_reward_func": 0.3671875,
|
| 1061 |
+
"rewards/format_reward_func": 0.9765625,
|
| 1062 |
+
"step": 162
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"completion_length": 457.34375,
|
| 1066 |
+
"epoch": 0.029155555555555556,
|
| 1067 |
+
"grad_norm": 0.14147277763916563,
|
| 1068 |
+
"kl": 0.079864501953125,
|
| 1069 |
+
"learning_rate": 3.6765055290490513e-07,
|
| 1070 |
+
"loss": 0.0001,
|
| 1071 |
+
"reward": 1.1015625,
|
| 1072 |
+
"reward_std": 0.26880460046231747,
|
| 1073 |
+
"rewards/equation_reward_func": 0.1328125,
|
| 1074 |
+
"rewards/format_reward_func": 0.96875,
|
| 1075 |
+
"step": 164
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"completion_length": 357.3515625,
|
| 1079 |
+
"epoch": 0.02951111111111111,
|
| 1080 |
+
"grad_norm": 0.1715828805345642,
|
| 1081 |
+
"kl": 0.0870819091796875,
|
| 1082 |
+
"learning_rate": 3.644595817401501e-07,
|
| 1083 |
+
"loss": 0.0001,
|
| 1084 |
+
"reward": 1.3515625,
|
| 1085 |
+
"reward_std": 0.35194794088602066,
|
| 1086 |
+
"rewards/equation_reward_func": 0.3828125,
|
| 1087 |
+
"rewards/format_reward_func": 0.96875,
|
| 1088 |
+
"step": 166
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"completion_length": 425.90625,
|
| 1092 |
+
"epoch": 0.029866666666666666,
|
| 1093 |
+
"grad_norm": 0.14100531969548477,
|
| 1094 |
+
"kl": 0.07891845703125,
|
| 1095 |
+
"learning_rate": 3.6124484045416483e-07,
|
| 1096 |
+
"loss": 0.0001,
|
| 1097 |
+
"reward": 1.2890625,
|
| 1098 |
+
"reward_std": 0.28513263165950775,
|
| 1099 |
+
"rewards/equation_reward_func": 0.3203125,
|
| 1100 |
+
"rewards/format_reward_func": 0.96875,
|
| 1101 |
+
"step": 168
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"completion_length": 366.703125,
|
| 1105 |
+
"epoch": 0.030222222222222223,
|
| 1106 |
+
"grad_norm": 0.16622145587450843,
|
| 1107 |
+
"kl": 0.0851593017578125,
|
| 1108 |
+
"learning_rate": 3.580069966606949e-07,
|
| 1109 |
+
"loss": 0.0001,
|
| 1110 |
+
"reward": 1.4296875,
|
| 1111 |
+
"reward_std": 0.31588183902204037,
|
| 1112 |
+
"rewards/equation_reward_func": 0.4296875,
|
| 1113 |
+
"rewards/format_reward_func": 1.0,
|
| 1114 |
+
"step": 170
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"completion_length": 371.0234375,
|
| 1118 |
+
"epoch": 0.030577777777777777,
|
| 1119 |
+
"grad_norm": 0.137987763899706,
|
| 1120 |
+
"kl": 0.1009979248046875,
|
| 1121 |
+
"learning_rate": 3.547467227712444e-07,
|
| 1122 |
+
"loss": 0.0001,
|
| 1123 |
+
"reward": 1.4375,
|
| 1124 |
+
"reward_std": 0.2630355451256037,
|
| 1125 |
+
"rewards/equation_reward_func": 0.453125,
|
| 1126 |
+
"rewards/format_reward_func": 0.984375,
|
| 1127 |
+
"step": 172
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"completion_length": 389.546875,
|
| 1131 |
+
"epoch": 0.030933333333333334,
|
| 1132 |
+
"grad_norm": 0.1711519127577815,
|
| 1133 |
+
"kl": 0.0822601318359375,
|
| 1134 |
+
"learning_rate": 3.5146469585543386e-07,
|
| 1135 |
+
"loss": 0.0001,
|
| 1136 |
+
"reward": 1.3125,
|
| 1137 |
+
"reward_std": 0.3842947632074356,
|
| 1138 |
+
"rewards/equation_reward_func": 0.359375,
|
| 1139 |
+
"rewards/format_reward_func": 0.953125,
|
| 1140 |
+
"step": 174
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"completion_length": 438.453125,
|
| 1144 |
+
"epoch": 0.03128888888888889,
|
| 1145 |
+
"grad_norm": 0.12051965947246142,
|
| 1146 |
+
"kl": 0.0817413330078125,
|
| 1147 |
+
"learning_rate": 3.481615975003922e-07,
|
| 1148 |
+
"loss": 0.0001,
|
| 1149 |
+
"reward": 1.2421875,
|
| 1150 |
+
"reward_std": 0.31843096762895584,
|
| 1151 |
+
"rewards/equation_reward_func": 0.2890625,
|
| 1152 |
+
"rewards/format_reward_func": 0.953125,
|
| 1153 |
+
"step": 176
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"completion_length": 381.703125,
|
| 1157 |
+
"epoch": 0.03164444444444445,
|
| 1158 |
+
"grad_norm": 0.15318076967674452,
|
| 1159 |
+
"kl": 0.0802001953125,
|
| 1160 |
+
"learning_rate": 3.448381136692089e-07,
|
| 1161 |
+
"loss": 0.0001,
|
| 1162 |
+
"reward": 1.4375,
|
| 1163 |
+
"reward_std": 0.23410366661846638,
|
| 1164 |
+
"rewards/equation_reward_func": 0.4453125,
|
| 1165 |
+
"rewards/format_reward_func": 0.9921875,
|
| 1166 |
+
"step": 178
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"completion_length": 404.2109375,
|
| 1170 |
+
"epoch": 0.032,
|
| 1171 |
+
"grad_norm": 0.18128262823423888,
|
| 1172 |
+
"kl": 0.078521728515625,
|
| 1173 |
+
"learning_rate": 3.4149493455847897e-07,
|
| 1174 |
+
"loss": 0.0001,
|
| 1175 |
+
"reward": 1.3046875,
|
| 1176 |
+
"reward_std": 0.2657256331294775,
|
| 1177 |
+
"rewards/equation_reward_func": 0.3515625,
|
| 1178 |
+
"rewards/format_reward_func": 0.953125,
|
| 1179 |
+
"step": 180
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"completion_length": 424.4609375,
|
| 1183 |
+
"epoch": 0.032355555555555554,
|
| 1184 |
+
"grad_norm": 0.14981419450826716,
|
| 1185 |
+
"kl": 0.083251953125,
|
| 1186 |
+
"learning_rate": 3.3813275445496766e-07,
|
| 1187 |
+
"loss": 0.0001,
|
| 1188 |
+
"reward": 1.15625,
|
| 1189 |
+
"reward_std": 0.33486341312527657,
|
| 1190 |
+
"rewards/equation_reward_func": 0.203125,
|
| 1191 |
+
"rewards/format_reward_func": 0.953125,
|
| 1192 |
+
"step": 182
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"completion_length": 345.890625,
|
| 1196 |
+
"epoch": 0.032711111111111114,
|
| 1197 |
+
"grad_norm": 0.15455090626242537,
|
| 1198 |
+
"kl": 0.0913238525390625,
|
| 1199 |
+
"learning_rate": 3.347522715914262e-07,
|
| 1200 |
+
"loss": 0.0001,
|
| 1201 |
+
"reward": 1.4765625,
|
| 1202 |
+
"reward_std": 0.3798140827566385,
|
| 1203 |
+
"rewards/equation_reward_func": 0.5,
|
| 1204 |
+
"rewards/format_reward_func": 0.9765625,
|
| 1205 |
+
"step": 184
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"completion_length": 330.0859375,
|
| 1209 |
+
"epoch": 0.03306666666666667,
|
| 1210 |
+
"grad_norm": 0.17923333714255735,
|
| 1211 |
+
"kl": 0.0919342041015625,
|
| 1212 |
+
"learning_rate": 3.313541880015877e-07,
|
| 1213 |
+
"loss": 0.0001,
|
| 1214 |
+
"reward": 1.4609375,
|
| 1215 |
+
"reward_std": 0.28012007288634777,
|
| 1216 |
+
"rewards/equation_reward_func": 0.4765625,
|
| 1217 |
+
"rewards/format_reward_func": 0.984375,
|
| 1218 |
+
"step": 186
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"completion_length": 419.8828125,
|
| 1222 |
+
"epoch": 0.03342222222222222,
|
| 1223 |
+
"grad_norm": 0.2229598300520282,
|
| 1224 |
+
"kl": 0.089447021484375,
|
| 1225 |
+
"learning_rate": 3.279392093743747e-07,
|
| 1226 |
+
"loss": 0.0001,
|
| 1227 |
+
"reward": 1.265625,
|
| 1228 |
+
"reward_std": 0.3122788220643997,
|
| 1229 |
+
"rewards/equation_reward_func": 0.328125,
|
| 1230 |
+
"rewards/format_reward_func": 0.9375,
|
| 1231 |
+
"step": 188
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"completion_length": 400.140625,
|
| 1235 |
+
"epoch": 0.033777777777777775,
|
| 1236 |
+
"grad_norm": 0.12889272306222627,
|
| 1237 |
+
"kl": 0.085418701171875,
|
| 1238 |
+
"learning_rate": 3.245080449073459e-07,
|
| 1239 |
+
"loss": 0.0001,
|
| 1240 |
+
"reward": 1.25,
|
| 1241 |
+
"reward_std": 0.24054957926273346,
|
| 1242 |
+
"rewards/equation_reward_func": 0.2734375,
|
| 1243 |
+
"rewards/format_reward_func": 0.9765625,
|
| 1244 |
+
"step": 190
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"completion_length": 363.1171875,
|
| 1248 |
+
"epoch": 0.034133333333333335,
|
| 1249 |
+
"grad_norm": 0.16266797929879878,
|
| 1250 |
+
"kl": 0.09893798828125,
|
| 1251 |
+
"learning_rate": 3.210614071594162e-07,
|
| 1252 |
+
"loss": 0.0001,
|
| 1253 |
+
"reward": 1.484375,
|
| 1254 |
+
"reward_std": 0.24435340240597725,
|
| 1255 |
+
"rewards/equation_reward_func": 0.4921875,
|
| 1256 |
+
"rewards/format_reward_func": 0.9921875,
|
| 1257 |
+
"step": 192
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"completion_length": 385.84375,
|
| 1261 |
+
"epoch": 0.03448888888888889,
|
| 1262 |
+
"grad_norm": 0.1286538346420234,
|
| 1263 |
+
"kl": 0.0966796875,
|
| 1264 |
+
"learning_rate": 3.1760001190287695e-07,
|
| 1265 |
+
"loss": 0.0001,
|
| 1266 |
+
"reward": 1.296875,
|
| 1267 |
+
"reward_std": 0.29644319973886013,
|
| 1268 |
+
"rewards/equation_reward_func": 0.3359375,
|
| 1269 |
+
"rewards/format_reward_func": 0.9609375,
|
| 1270 |
+
"step": 194
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"completion_length": 491.7578125,
|
| 1274 |
+
"epoch": 0.03484444444444444,
|
| 1275 |
+
"grad_norm": 0.13692671785492067,
|
| 1276 |
+
"kl": 0.082366943359375,
|
| 1277 |
+
"learning_rate": 3.141245779747502e-07,
|
| 1278 |
+
"loss": 0.0001,
|
| 1279 |
+
"reward": 1.1875,
|
| 1280 |
+
"reward_std": 0.39936898462474346,
|
| 1281 |
+
"rewards/equation_reward_func": 0.28125,
|
| 1282 |
+
"rewards/format_reward_func": 0.90625,
|
| 1283 |
+
"step": 196
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"completion_length": 385.8515625,
|
| 1287 |
+
"epoch": 0.0352,
|
| 1288 |
+
"grad_norm": 0.15906297032986566,
|
| 1289 |
+
"kl": 0.0865325927734375,
|
| 1290 |
+
"learning_rate": 3.106358271275056e-07,
|
| 1291 |
+
"loss": 0.0001,
|
| 1292 |
+
"reward": 1.3203125,
|
| 1293 |
+
"reward_std": 0.23891541920602322,
|
| 1294 |
+
"rewards/equation_reward_func": 0.3359375,
|
| 1295 |
+
"rewards/format_reward_func": 0.984375,
|
| 1296 |
+
"step": 198
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"completion_length": 345.7734375,
|
| 1300 |
+
"epoch": 0.035555555555555556,
|
| 1301 |
+
"grad_norm": 0.10916438327428707,
|
| 1302 |
+
"kl": 0.10546875,
|
| 1303 |
+
"learning_rate": 3.0713448387917227e-07,
|
| 1304 |
+
"loss": 0.0001,
|
| 1305 |
+
"reward": 1.484375,
|
| 1306 |
+
"reward_std": 0.21754318848252296,
|
| 1307 |
+
"rewards/equation_reward_func": 0.5078125,
|
| 1308 |
+
"rewards/format_reward_func": 0.9765625,
|
| 1309 |
+
"step": 200
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"completion_length": 469.6328125,
|
| 1313 |
+
"epoch": 0.03591111111111111,
|
| 1314 |
+
"grad_norm": 0.13427396937511493,
|
| 1315 |
+
"kl": 0.1422882080078125,
|
| 1316 |
+
"learning_rate": 3.0362127536287636e-07,
|
| 1317 |
+
"loss": 0.0001,
|
| 1318 |
+
"reward": 1.296875,
|
| 1319 |
+
"reward_std": 0.4032180178910494,
|
| 1320 |
+
"rewards/equation_reward_func": 0.359375,
|
| 1321 |
+
"rewards/format_reward_func": 0.9375,
|
| 1322 |
+
"step": 202
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"completion_length": 462.484375,
|
| 1326 |
+
"epoch": 0.03626666666666667,
|
| 1327 |
+
"grad_norm": 0.14175335721131654,
|
| 1328 |
+
"kl": 0.084930419921875,
|
| 1329 |
+
"learning_rate": 3.0009693117583523e-07,
|
| 1330 |
+
"loss": 0.0001,
|
| 1331 |
+
"reward": 1.2109375,
|
| 1332 |
+
"reward_std": 0.3685295097529888,
|
| 1333 |
+
"rewards/equation_reward_func": 0.2734375,
|
| 1334 |
+
"rewards/format_reward_func": 0.9375,
|
| 1335 |
+
"step": 204
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"completion_length": 442.421875,
|
| 1339 |
+
"epoch": 0.03662222222222222,
|
| 1340 |
+
"grad_norm": 0.1605545110922793,
|
| 1341 |
+
"kl": 0.092681884765625,
|
| 1342 |
+
"learning_rate": 2.965621832278401e-07,
|
| 1343 |
+
"loss": 0.0001,
|
| 1344 |
+
"reward": 1.2265625,
|
| 1345 |
+
"reward_std": 0.36273445934057236,
|
| 1346 |
+
"rewards/equation_reward_func": 0.3125,
|
| 1347 |
+
"rewards/format_reward_func": 0.9140625,
|
| 1348 |
+
"step": 206
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"completion_length": 487.59375,
|
| 1352 |
+
"epoch": 0.036977777777777776,
|
| 1353 |
+
"grad_norm": 0.17241738987093455,
|
| 1354 |
+
"kl": 0.083953857421875,
|
| 1355 |
+
"learning_rate": 2.9301776558925875e-07,
|
| 1356 |
+
"loss": 0.0001,
|
| 1357 |
+
"reward": 1.2421875,
|
| 1358 |
+
"reward_std": 0.3413938693702221,
|
| 1359 |
+
"rewards/equation_reward_func": 0.2890625,
|
| 1360 |
+
"rewards/format_reward_func": 0.953125,
|
| 1361 |
+
"step": 208
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"completion_length": 385.3359375,
|
| 1365 |
+
"epoch": 0.037333333333333336,
|
| 1366 |
+
"grad_norm": 0.18550539158073817,
|
| 1367 |
+
"kl": 0.099517822265625,
|
| 1368 |
+
"learning_rate": 2.894644143385885e-07,
|
| 1369 |
+
"loss": 0.0001,
|
| 1370 |
+
"reward": 1.46875,
|
| 1371 |
+
"reward_std": 0.40134032256901264,
|
| 1372 |
+
"rewards/equation_reward_func": 0.5,
|
| 1373 |
+
"rewards/format_reward_func": 0.96875,
|
| 1374 |
+
"step": 210
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"completion_length": 407.0703125,
|
| 1378 |
+
"epoch": 0.03768888888888889,
|
| 1379 |
+
"grad_norm": 0.10848019065658879,
|
| 1380 |
+
"kl": 0.1204833984375,
|
| 1381 |
+
"learning_rate": 2.859028674095937e-07,
|
| 1382 |
+
"loss": 0.0001,
|
| 1383 |
+
"reward": 1.375,
|
| 1384 |
+
"reward_std": 0.311705494299531,
|
| 1385 |
+
"rewards/equation_reward_func": 0.421875,
|
| 1386 |
+
"rewards/format_reward_func": 0.953125,
|
| 1387 |
+
"step": 212
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"completion_length": 430.4921875,
|
| 1391 |
+
"epoch": 0.03804444444444444,
|
| 1392 |
+
"grad_norm": 0.09779461620640097,
|
| 1393 |
+
"kl": 0.1276397705078125,
|
| 1394 |
+
"learning_rate": 2.823338644380566e-07,
|
| 1395 |
+
"loss": 0.0001,
|
| 1396 |
+
"reward": 1.28125,
|
| 1397 |
+
"reward_std": 0.24435340240597725,
|
| 1398 |
+
"rewards/equation_reward_func": 0.359375,
|
| 1399 |
+
"rewards/format_reward_func": 0.921875,
|
| 1400 |
+
"step": 214
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"completion_length": 337.546875,
|
| 1404 |
+
"epoch": 0.0384,
|
| 1405 |
+
"grad_norm": 0.11632584351646018,
|
| 1406 |
+
"kl": 0.169647216796875,
|
| 1407 |
+
"learning_rate": 2.7875814660817504e-07,
|
| 1408 |
+
"loss": 0.0002,
|
| 1409 |
+
"reward": 1.5625,
|
| 1410 |
+
"reward_std": 0.20069601200520992,
|
| 1411 |
+
"rewards/equation_reward_func": 0.578125,
|
| 1412 |
+
"rewards/format_reward_func": 0.984375,
|
| 1413 |
+
"step": 216
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"completion_length": 452.1171875,
|
| 1417 |
+
"epoch": 0.03875555555555556,
|
| 1418 |
+
"grad_norm": 0.12440617923159014,
|
| 1419 |
+
"kl": 0.1014556884765625,
|
| 1420 |
+
"learning_rate": 2.751764564986396e-07,
|
| 1421 |
+
"loss": 0.0001,
|
| 1422 |
+
"reward": 1.234375,
|
| 1423 |
+
"reward_std": 0.28065195120871067,
|
| 1424 |
+
"rewards/equation_reward_func": 0.28125,
|
| 1425 |
+
"rewards/format_reward_func": 0.953125,
|
| 1426 |
+
"step": 218
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"completion_length": 456.3359375,
|
| 1430 |
+
"epoch": 0.03911111111111111,
|
| 1431 |
+
"grad_norm": 0.19414546978036584,
|
| 1432 |
+
"kl": 0.1228485107421875,
|
| 1433 |
+
"learning_rate": 2.715895379284194e-07,
|
| 1434 |
+
"loss": 0.0001,
|
| 1435 |
+
"reward": 1.265625,
|
| 1436 |
+
"reward_std": 0.27309720776975155,
|
| 1437 |
+
"rewards/equation_reward_func": 0.3046875,
|
| 1438 |
+
"rewards/format_reward_func": 0.9609375,
|
| 1439 |
+
"step": 220
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"completion_length": 422.109375,
|
| 1443 |
+
"epoch": 0.039466666666666664,
|
| 1444 |
+
"grad_norm": 0.18691301250047884,
|
| 1445 |
+
"kl": 0.1074981689453125,
|
| 1446 |
+
"learning_rate": 2.6799813580229174e-07,
|
| 1447 |
+
"loss": 0.0001,
|
| 1448 |
+
"reward": 1.3125,
|
| 1449 |
+
"reward_std": 0.31189604848623276,
|
| 1450 |
+
"rewards/equation_reward_func": 0.3828125,
|
| 1451 |
+
"rewards/format_reward_func": 0.9296875,
|
| 1452 |
+
"step": 222
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"completion_length": 375.328125,
|
| 1456 |
+
"epoch": 0.039822222222222224,
|
| 1457 |
+
"grad_norm": 0.1870312456814116,
|
| 1458 |
+
"kl": 0.12664794921875,
|
| 1459 |
+
"learning_rate": 2.6440299595614606e-07,
|
| 1460 |
+
"loss": 0.0001,
|
| 1461 |
+
"reward": 1.484375,
|
| 1462 |
+
"reward_std": 0.271799024194479,
|
| 1463 |
+
"rewards/equation_reward_func": 0.5078125,
|
| 1464 |
+
"rewards/format_reward_func": 0.9765625,
|
| 1465 |
+
"step": 224
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"completion_length": 441.59375,
|
| 1469 |
+
"epoch": 0.04017777777777778,
|
| 1470 |
+
"grad_norm": 0.1395588541862425,
|
| 1471 |
+
"kl": 0.099700927734375,
|
| 1472 |
+
"learning_rate": 2.6080486500209347e-07,
|
| 1473 |
+
"loss": 0.0001,
|
| 1474 |
+
"reward": 1.34375,
|
| 1475 |
+
"reward_std": 0.3171109911054373,
|
| 1476 |
+
"rewards/equation_reward_func": 0.40625,
|
| 1477 |
+
"rewards/format_reward_func": 0.9375,
|
| 1478 |
+
"step": 226
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"completion_length": 412.484375,
|
| 1482 |
+
"epoch": 0.04053333333333333,
|
| 1483 |
+
"grad_norm": 0.17905219731684363,
|
| 1484 |
+
"kl": 0.1191253662109375,
|
| 1485 |
+
"learning_rate": 2.572044901734166e-07,
|
| 1486 |
+
"loss": 0.0001,
|
| 1487 |
+
"reward": 1.2890625,
|
| 1488 |
+
"reward_std": 0.28012007288634777,
|
| 1489 |
+
"rewards/equation_reward_func": 0.3515625,
|
| 1490 |
+
"rewards/format_reward_func": 0.9375,
|
| 1491 |
+
"step": 228
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"completion_length": 476.296875,
|
| 1495 |
+
"epoch": 0.04088888888888889,
|
| 1496 |
+
"grad_norm": 0.1304738846155797,
|
| 1497 |
+
"kl": 0.1032562255859375,
|
| 1498 |
+
"learning_rate": 2.536026191693893e-07,
|
| 1499 |
+
"loss": 0.0001,
|
| 1500 |
+
"reward": 1.25,
|
| 1501 |
+
"reward_std": 0.23391559347510338,
|
| 1502 |
+
"rewards/equation_reward_func": 0.296875,
|
| 1503 |
+
"rewards/format_reward_func": 0.953125,
|
| 1504 |
+
"step": 230
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"completion_length": 382.609375,
|
| 1508 |
+
"epoch": 0.041244444444444445,
|
| 1509 |
+
"grad_norm": 0.14335342765956446,
|
| 1510 |
+
"kl": 0.172271728515625,
|
| 1511 |
+
"learning_rate": 2.5e-07,
|
| 1512 |
+
"loss": 0.0002,
|
| 1513 |
+
"reward": 1.4140625,
|
| 1514 |
+
"reward_std": 0.23068872280418873,
|
| 1515 |
+
"rewards/equation_reward_func": 0.46875,
|
| 1516 |
+
"rewards/format_reward_func": 0.9453125,
|
| 1517 |
+
"step": 232
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"completion_length": 413.921875,
|
| 1521 |
+
"epoch": 0.0416,
|
| 1522 |
+
"grad_norm": 0.11409022946228915,
|
| 1523 |
+
"kl": 0.150146484375,
|
| 1524 |
+
"learning_rate": 2.4639738083061073e-07,
|
| 1525 |
+
"loss": 0.0002,
|
| 1526 |
+
"reward": 1.3828125,
|
| 1527 |
+
"reward_std": 0.14966704696416855,
|
| 1528 |
+
"rewards/equation_reward_func": 0.4140625,
|
| 1529 |
+
"rewards/format_reward_func": 0.96875,
|
| 1530 |
+
"step": 234
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"completion_length": 435.7265625,
|
| 1534 |
+
"epoch": 0.04195555555555556,
|
| 1535 |
+
"grad_norm": 0.19382477904173945,
|
| 1536 |
+
"kl": 0.108245849609375,
|
| 1537 |
+
"learning_rate": 2.4279550982658345e-07,
|
| 1538 |
+
"loss": 0.0001,
|
| 1539 |
+
"reward": 1.296875,
|
| 1540 |
+
"reward_std": 0.4299037382006645,
|
| 1541 |
+
"rewards/equation_reward_func": 0.3671875,
|
| 1542 |
+
"rewards/format_reward_func": 0.9296875,
|
| 1543 |
+
"step": 236
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"completion_length": 422.359375,
|
| 1547 |
+
"epoch": 0.04231111111111111,
|
| 1548 |
+
"grad_norm": 0.07742598896983852,
|
| 1549 |
+
"kl": 0.134552001953125,
|
| 1550 |
+
"learning_rate": 2.3919513499790646e-07,
|
| 1551 |
+
"loss": 0.0001,
|
| 1552 |
+
"reward": 1.2734375,
|
| 1553 |
+
"reward_std": 0.28402500227093697,
|
| 1554 |
+
"rewards/equation_reward_func": 0.3359375,
|
| 1555 |
+
"rewards/format_reward_func": 0.9375,
|
| 1556 |
+
"step": 238
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"completion_length": 446.703125,
|
| 1560 |
+
"epoch": 0.042666666666666665,
|
| 1561 |
+
"grad_norm": 0.13048074165206625,
|
| 1562 |
+
"kl": 0.10418701171875,
|
| 1563 |
+
"learning_rate": 2.3559700404385394e-07,
|
| 1564 |
+
"loss": 0.0001,
|
| 1565 |
+
"reward": 1.2734375,
|
| 1566 |
+
"reward_std": 0.20680594816803932,
|
| 1567 |
+
"rewards/equation_reward_func": 0.34375,
|
| 1568 |
+
"rewards/format_reward_func": 0.9296875,
|
| 1569 |
+
"step": 240
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"completion_length": 380.8671875,
|
| 1573 |
+
"epoch": 0.043022222222222226,
|
| 1574 |
+
"grad_norm": 0.15195815320180947,
|
| 1575 |
+
"kl": 0.134918212890625,
|
| 1576 |
+
"learning_rate": 2.3200186419770823e-07,
|
| 1577 |
+
"loss": 0.0001,
|
| 1578 |
+
"reward": 1.4375,
|
| 1579 |
+
"reward_std": 0.17176413349807262,
|
| 1580 |
+
"rewards/equation_reward_func": 0.453125,
|
| 1581 |
+
"rewards/format_reward_func": 0.984375,
|
| 1582 |
+
"step": 242
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"completion_length": 317.765625,
|
| 1586 |
+
"epoch": 0.04337777777777778,
|
| 1587 |
+
"grad_norm": 0.20510319394941706,
|
| 1588 |
+
"kl": 0.139068603515625,
|
| 1589 |
+
"learning_rate": 2.284104620715807e-07,
|
| 1590 |
+
"loss": 0.0001,
|
| 1591 |
+
"reward": 1.6015625,
|
| 1592 |
+
"reward_std": 0.2688144091516733,
|
| 1593 |
+
"rewards/equation_reward_func": 0.609375,
|
| 1594 |
+
"rewards/format_reward_func": 0.9921875,
|
| 1595 |
+
"step": 244
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"completion_length": 401.015625,
|
| 1599 |
+
"epoch": 0.04373333333333333,
|
| 1600 |
+
"grad_norm": 0.17580532584141528,
|
| 1601 |
+
"kl": 0.115447998046875,
|
| 1602 |
+
"learning_rate": 2.2482354350136043e-07,
|
| 1603 |
+
"loss": 0.0001,
|
| 1604 |
+
"reward": 1.3359375,
|
| 1605 |
+
"reward_std": 0.2896084077656269,
|
| 1606 |
+
"rewards/equation_reward_func": 0.3828125,
|
| 1607 |
+
"rewards/format_reward_func": 0.953125,
|
| 1608 |
+
"step": 246
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"completion_length": 374.3828125,
|
| 1612 |
+
"epoch": 0.044088888888888886,
|
| 1613 |
+
"grad_norm": 0.13703363908634722,
|
| 1614 |
+
"kl": 0.13311767578125,
|
| 1615 |
+
"learning_rate": 2.2124185339182496e-07,
|
| 1616 |
+
"loss": 0.0001,
|
| 1617 |
+
"reward": 1.453125,
|
| 1618 |
+
"reward_std": 0.19958838261663914,
|
| 1619 |
+
"rewards/equation_reward_func": 0.484375,
|
| 1620 |
+
"rewards/format_reward_func": 0.96875,
|
| 1621 |
+
"step": 248
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"completion_length": 410.9921875,
|
| 1625 |
+
"epoch": 0.044444444444444446,
|
| 1626 |
+
"grad_norm": 0.12840967985250826,
|
| 1627 |
+
"kl": 0.124114990234375,
|
| 1628 |
+
"learning_rate": 2.1766613556194344e-07,
|
| 1629 |
+
"loss": 0.0001,
|
| 1630 |
+
"reward": 1.265625,
|
| 1631 |
+
"reward_std": 0.29555559158325195,
|
| 1632 |
+
"rewards/equation_reward_func": 0.328125,
|
| 1633 |
+
"rewards/format_reward_func": 0.9375,
|
| 1634 |
+
"step": 250
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"completion_length": 370.46875,
|
| 1638 |
+
"epoch": 0.0448,
|
| 1639 |
+
"grad_norm": 0.14200419109331727,
|
| 1640 |
+
"kl": 0.12237548828125,
|
| 1641 |
+
"learning_rate": 2.1409713259040628e-07,
|
| 1642 |
+
"loss": 0.0001,
|
| 1643 |
+
"reward": 1.40625,
|
| 1644 |
+
"reward_std": 0.17965975776314735,
|
| 1645 |
+
"rewards/equation_reward_func": 0.421875,
|
| 1646 |
+
"rewards/format_reward_func": 0.984375,
|
| 1647 |
+
"step": 252
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"completion_length": 372.2265625,
|
| 1651 |
+
"epoch": 0.04515555555555555,
|
| 1652 |
+
"grad_norm": 0.23225534396386804,
|
| 1653 |
+
"kl": 0.143218994140625,
|
| 1654 |
+
"learning_rate": 2.105355856614115e-07,
|
| 1655 |
+
"loss": 0.0001,
|
| 1656 |
+
"reward": 1.3046875,
|
| 1657 |
+
"reward_std": 0.20411095581948757,
|
| 1658 |
+
"rewards/equation_reward_func": 0.34375,
|
| 1659 |
+
"rewards/format_reward_func": 0.9609375,
|
| 1660 |
+
"step": 254
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"completion_length": 285.4921875,
|
| 1664 |
+
"epoch": 0.04551111111111111,
|
| 1665 |
+
"grad_norm": 0.24766446564024155,
|
| 1666 |
+
"kl": 0.151763916015625,
|
| 1667 |
+
"learning_rate": 2.069822344107413e-07,
|
| 1668 |
+
"loss": 0.0002,
|
| 1669 |
+
"reward": 1.59375,
|
| 1670 |
+
"reward_std": 0.25689312256872654,
|
| 1671 |
+
"rewards/equation_reward_func": 0.609375,
|
| 1672 |
+
"rewards/format_reward_func": 0.984375,
|
| 1673 |
+
"step": 256
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"completion_length": 313.8203125,
|
| 1677 |
+
"epoch": 0.04586666666666667,
|
| 1678 |
+
"grad_norm": 0.16165568014293,
|
| 1679 |
+
"kl": 0.146575927734375,
|
| 1680 |
+
"learning_rate": 2.034378167721599e-07,
|
| 1681 |
+
"loss": 0.0001,
|
| 1682 |
+
"reward": 1.59375,
|
| 1683 |
+
"reward_std": 0.13781969621777534,
|
| 1684 |
+
"rewards/equation_reward_func": 0.6015625,
|
| 1685 |
+
"rewards/format_reward_func": 0.9921875,
|
| 1686 |
+
"step": 258
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"completion_length": 415.359375,
|
| 1690 |
+
"epoch": 0.04622222222222222,
|
| 1691 |
+
"grad_norm": 0.09688841854391178,
|
| 1692 |
+
"kl": 0.12677001953125,
|
| 1693 |
+
"learning_rate": 1.9990306882416485e-07,
|
| 1694 |
+
"loss": 0.0001,
|
| 1695 |
+
"reward": 1.2890625,
|
| 1696 |
+
"reward_std": 0.15756267122924328,
|
| 1697 |
+
"rewards/equation_reward_func": 0.328125,
|
| 1698 |
+
"rewards/format_reward_func": 0.9609375,
|
| 1699 |
+
"step": 260
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"completion_length": 373.765625,
|
| 1703 |
+
"epoch": 0.04657777777777778,
|
| 1704 |
+
"grad_norm": 0.11053378420987946,
|
| 1705 |
+
"kl": 0.12457275390625,
|
| 1706 |
+
"learning_rate": 1.9637872463712362e-07,
|
| 1707 |
+
"loss": 0.0001,
|
| 1708 |
+
"reward": 1.4140625,
|
| 1709 |
+
"reward_std": 0.17859892547130585,
|
| 1710 |
+
"rewards/equation_reward_func": 0.4453125,
|
| 1711 |
+
"rewards/format_reward_func": 0.96875,
|
| 1712 |
+
"step": 262
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"completion_length": 399.046875,
|
| 1716 |
+
"epoch": 0.046933333333333334,
|
| 1717 |
+
"grad_norm": 0.09951068355965577,
|
| 1718 |
+
"kl": 0.12164306640625,
|
| 1719 |
+
"learning_rate": 1.9286551612082773e-07,
|
| 1720 |
+
"loss": 0.0001,
|
| 1721 |
+
"reward": 1.3828125,
|
| 1722 |
+
"reward_std": 0.23231797851622105,
|
| 1723 |
+
"rewards/equation_reward_func": 0.4296875,
|
| 1724 |
+
"rewards/format_reward_func": 0.953125,
|
| 1725 |
+
"step": 264
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"completion_length": 410.6484375,
|
| 1729 |
+
"epoch": 0.04728888888888889,
|
| 1730 |
+
"grad_norm": 0.12843842474279485,
|
| 1731 |
+
"kl": 0.109649658203125,
|
| 1732 |
+
"learning_rate": 1.8936417287249446e-07,
|
| 1733 |
+
"loss": 0.0001,
|
| 1734 |
+
"reward": 1.3828125,
|
| 1735 |
+
"reward_std": 0.21937325038015842,
|
| 1736 |
+
"rewards/equation_reward_func": 0.3984375,
|
| 1737 |
+
"rewards/format_reward_func": 0.984375,
|
| 1738 |
+
"step": 266
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"completion_length": 338.921875,
|
| 1742 |
+
"epoch": 0.04764444444444445,
|
| 1743 |
+
"grad_norm": 0.1702909908327491,
|
| 1744 |
+
"kl": 0.172119140625,
|
| 1745 |
+
"learning_rate": 1.8587542202524985e-07,
|
| 1746 |
+
"loss": 0.0002,
|
| 1747 |
+
"reward": 1.390625,
|
| 1748 |
+
"reward_std": 0.18543371744453907,
|
| 1749 |
+
"rewards/equation_reward_func": 0.421875,
|
| 1750 |
+
"rewards/format_reward_func": 0.96875,
|
| 1751 |
+
"step": 268
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"completion_length": 415.5546875,
|
| 1755 |
+
"epoch": 0.048,
|
| 1756 |
+
"grad_norm": 0.15257037646762198,
|
| 1757 |
+
"kl": 0.125518798828125,
|
| 1758 |
+
"learning_rate": 1.82399988097123e-07,
|
| 1759 |
+
"loss": 0.0001,
|
| 1760 |
+
"reward": 1.3515625,
|
| 1761 |
+
"reward_std": 0.2567375358194113,
|
| 1762 |
+
"rewards/equation_reward_func": 0.3671875,
|
| 1763 |
+
"rewards/format_reward_func": 0.984375,
|
| 1764 |
+
"step": 270
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"completion_length": 491.0703125,
|
| 1768 |
+
"epoch": 0.048355555555555554,
|
| 1769 |
+
"grad_norm": 0.1015784600459762,
|
| 1770 |
+
"kl": 0.1043853759765625,
|
| 1771 |
+
"learning_rate": 1.7893859284058378e-07,
|
| 1772 |
+
"loss": 0.0001,
|
| 1773 |
+
"reward": 1.203125,
|
| 1774 |
+
"reward_std": 0.28407179936766624,
|
| 1775 |
+
"rewards/equation_reward_func": 0.2734375,
|
| 1776 |
+
"rewards/format_reward_func": 0.9296875,
|
| 1777 |
+
"step": 272
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"completion_length": 320.1484375,
|
| 1781 |
+
"epoch": 0.04871111111111111,
|
| 1782 |
+
"grad_norm": 0.07568286413210676,
|
| 1783 |
+
"kl": 0.19097900390625,
|
| 1784 |
+
"learning_rate": 1.7549195509265407e-07,
|
| 1785 |
+
"loss": 0.0002,
|
| 1786 |
+
"reward": 1.5625,
|
| 1787 |
+
"reward_std": 0.16097761504352093,
|
| 1788 |
+
"rewards/equation_reward_func": 0.5625,
|
| 1789 |
+
"rewards/format_reward_func": 1.0,
|
| 1790 |
+
"step": 274
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"completion_length": 397.09375,
|
| 1794 |
+
"epoch": 0.04906666666666667,
|
| 1795 |
+
"grad_norm": 0.15664401529813785,
|
| 1796 |
+
"kl": 0.12188720703125,
|
| 1797 |
+
"learning_rate": 1.7206079062562536e-07,
|
| 1798 |
+
"loss": 0.0001,
|
| 1799 |
+
"reward": 1.375,
|
| 1800 |
+
"reward_std": 0.2890843581408262,
|
| 1801 |
+
"rewards/equation_reward_func": 0.4296875,
|
| 1802 |
+
"rewards/format_reward_func": 0.9453125,
|
| 1803 |
+
"step": 276
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"completion_length": 412.875,
|
| 1807 |
+
"epoch": 0.04942222222222222,
|
| 1808 |
+
"grad_norm": 0.152716469033041,
|
| 1809 |
+
"kl": 0.215545654296875,
|
| 1810 |
+
"learning_rate": 1.6864581199841226e-07,
|
| 1811 |
+
"loss": 0.0002,
|
| 1812 |
+
"reward": 1.4140625,
|
| 1813 |
+
"reward_std": 0.24452662654221058,
|
| 1814 |
+
"rewards/equation_reward_func": 0.4453125,
|
| 1815 |
+
"rewards/format_reward_func": 0.96875,
|
| 1816 |
+
"step": 278
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"completion_length": 322.53125,
|
| 1820 |
+
"epoch": 0.049777777777777775,
|
| 1821 |
+
"grad_norm": 0.06465172658536934,
|
| 1822 |
+
"kl": 0.12664794921875,
|
| 1823 |
+
"learning_rate": 1.6524772840857388e-07,
|
| 1824 |
+
"loss": 0.0001,
|
| 1825 |
+
"reward": 1.59375,
|
| 1826 |
+
"reward_std": 0.14123954437673092,
|
| 1827 |
+
"rewards/equation_reward_func": 0.59375,
|
| 1828 |
+
"rewards/format_reward_func": 1.0,
|
| 1829 |
+
"step": 280
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"completion_length": 338.296875,
|
| 1833 |
+
"epoch": 0.050133333333333335,
|
| 1834 |
+
"grad_norm": 0.13308881483334925,
|
| 1835 |
+
"kl": 0.119415283203125,
|
| 1836 |
+
"learning_rate": 1.6186724554503237e-07,
|
| 1837 |
+
"loss": 0.0001,
|
| 1838 |
+
"reward": 1.625,
|
| 1839 |
+
"reward_std": 0.2380426600575447,
|
| 1840 |
+
"rewards/equation_reward_func": 0.6328125,
|
| 1841 |
+
"rewards/format_reward_func": 0.9921875,
|
| 1842 |
+
"step": 282
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"completion_length": 408.3515625,
|
| 1846 |
+
"epoch": 0.05048888888888889,
|
| 1847 |
+
"grad_norm": 0.12489806113236866,
|
| 1848 |
+
"kl": 0.099822998046875,
|
| 1849 |
+
"learning_rate": 1.5850506544152103e-07,
|
| 1850 |
+
"loss": 0.0001,
|
| 1851 |
+
"reward": 1.3359375,
|
| 1852 |
+
"reward_std": 0.2407503854483366,
|
| 1853 |
+
"rewards/equation_reward_func": 0.3671875,
|
| 1854 |
+
"rewards/format_reward_func": 0.96875,
|
| 1855 |
+
"step": 284
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"completion_length": 436.9921875,
|
| 1859 |
+
"epoch": 0.05084444444444444,
|
| 1860 |
+
"grad_norm": 0.1660027330528205,
|
| 1861 |
+
"kl": 0.111358642578125,
|
| 1862 |
+
"learning_rate": 1.5516188633079107e-07,
|
| 1863 |
+
"loss": 0.0001,
|
| 1864 |
+
"reward": 1.2890625,
|
| 1865 |
+
"reward_std": 0.2788218893110752,
|
| 1866 |
+
"rewards/equation_reward_func": 0.3203125,
|
| 1867 |
+
"rewards/format_reward_func": 0.96875,
|
| 1868 |
+
"step": 286
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"completion_length": 381.5390625,
|
| 1872 |
+
"epoch": 0.0512,
|
| 1873 |
+
"grad_norm": 0.1773949520743432,
|
| 1874 |
+
"kl": 0.1202392578125,
|
| 1875 |
+
"learning_rate": 1.5183840249960784e-07,
|
| 1876 |
+
"loss": 0.0001,
|
| 1877 |
+
"reward": 1.4375,
|
| 1878 |
+
"reward_std": 0.23410366661846638,
|
| 1879 |
+
"rewards/equation_reward_func": 0.46875,
|
| 1880 |
+
"rewards/format_reward_func": 0.96875,
|
| 1881 |
+
"step": 288
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"completion_length": 363.2890625,
|
| 1885 |
+
"epoch": 0.051555555555555556,
|
| 1886 |
+
"grad_norm": 0.12908690460870859,
|
| 1887 |
+
"kl": 0.12237548828125,
|
| 1888 |
+
"learning_rate": 1.4853530414456612e-07,
|
| 1889 |
+
"loss": 0.0001,
|
| 1890 |
+
"reward": 1.4140625,
|
| 1891 |
+
"reward_std": 0.2874350417405367,
|
| 1892 |
+
"rewards/equation_reward_func": 0.4375,
|
| 1893 |
+
"rewards/format_reward_func": 0.9765625,
|
| 1894 |
+
"step": 290
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"completion_length": 398.25,
|
| 1898 |
+
"epoch": 0.05191111111111111,
|
| 1899 |
+
"grad_norm": 0.1223189167367173,
|
| 1900 |
+
"kl": 0.10498046875,
|
| 1901 |
+
"learning_rate": 1.4525327722875568e-07,
|
| 1902 |
+
"loss": 0.0001,
|
| 1903 |
+
"reward": 1.421875,
|
| 1904 |
+
"reward_std": 0.20517178811132908,
|
| 1905 |
+
"rewards/equation_reward_func": 0.4609375,
|
| 1906 |
+
"rewards/format_reward_func": 0.9609375,
|
| 1907 |
+
"step": 292
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"completion_length": 299.5390625,
|
| 1911 |
+
"epoch": 0.05226666666666667,
|
| 1912 |
+
"grad_norm": 0.1475307814276985,
|
| 1913 |
+
"kl": 0.138275146484375,
|
| 1914 |
+
"learning_rate": 1.4199300333930515e-07,
|
| 1915 |
+
"loss": 0.0001,
|
| 1916 |
+
"reward": 1.6171875,
|
| 1917 |
+
"reward_std": 0.15773099102079868,
|
| 1918 |
+
"rewards/equation_reward_func": 0.6328125,
|
| 1919 |
+
"rewards/format_reward_func": 0.984375,
|
| 1920 |
+
"step": 294
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"completion_length": 408.1484375,
|
| 1924 |
+
"epoch": 0.05262222222222222,
|
| 1925 |
+
"grad_norm": 0.15893771408986387,
|
| 1926 |
+
"kl": 0.12841796875,
|
| 1927 |
+
"learning_rate": 1.3875515954583523e-07,
|
| 1928 |
+
"loss": 0.0001,
|
| 1929 |
+
"reward": 1.375,
|
| 1930 |
+
"reward_std": 0.22043408267199993,
|
| 1931 |
+
"rewards/equation_reward_func": 0.421875,
|
| 1932 |
+
"rewards/format_reward_func": 0.953125,
|
| 1933 |
+
"step": 296
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"completion_length": 386.34375,
|
| 1937 |
+
"epoch": 0.052977777777777776,
|
| 1938 |
+
"grad_norm": 0.09802748245282768,
|
| 1939 |
+
"kl": 0.12982177734375,
|
| 1940 |
+
"learning_rate": 1.3554041825985e-07,
|
| 1941 |
+
"loss": 0.0001,
|
| 1942 |
+
"reward": 1.3828125,
|
| 1943 |
+
"reward_std": 0.19691260531544685,
|
| 1944 |
+
"rewards/equation_reward_func": 0.4140625,
|
| 1945 |
+
"rewards/format_reward_func": 0.96875,
|
| 1946 |
+
"step": 298
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"completion_length": 376.203125,
|
| 1950 |
+
"epoch": 0.05333333333333334,
|
| 1951 |
+
"grad_norm": 0.0792523065441085,
|
| 1952 |
+
"kl": 0.11859130859375,
|
| 1953 |
+
"learning_rate": 1.323494470950949e-07,
|
| 1954 |
+
"loss": 0.0001,
|
| 1955 |
+
"reward": 1.46875,
|
| 1956 |
+
"reward_std": 0.13098980858922005,
|
| 1957 |
+
"rewards/equation_reward_func": 0.4765625,
|
| 1958 |
+
"rewards/format_reward_func": 0.9921875,
|
| 1959 |
+
"step": 300
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"completion_length": 353.7421875,
|
| 1963 |
+
"epoch": 0.05368888888888889,
|
| 1964 |
+
"grad_norm": 0.0887805404195599,
|
| 1965 |
+
"kl": 0.125640869140625,
|
| 1966 |
+
"learning_rate": 1.2918290872891236e-07,
|
| 1967 |
+
"loss": 0.0001,
|
| 1968 |
+
"reward": 1.4921875,
|
| 1969 |
+
"reward_std": 0.16762433387339115,
|
| 1970 |
+
"rewards/equation_reward_func": 0.5,
|
| 1971 |
+
"rewards/format_reward_func": 0.9921875,
|
| 1972 |
+
"step": 302
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"completion_length": 440.4375,
|
| 1976 |
+
"epoch": 0.054044444444444444,
|
| 1977 |
+
"grad_norm": 0.10652059405770911,
|
| 1978 |
+
"kl": 0.104339599609375,
|
| 1979 |
+
"learning_rate": 1.260414607646213e-07,
|
| 1980 |
+
"loss": 0.0001,
|
| 1981 |
+
"reward": 1.21875,
|
| 1982 |
+
"reward_std": 0.2987973112612963,
|
| 1983 |
+
"rewards/equation_reward_func": 0.2734375,
|
| 1984 |
+
"rewards/format_reward_func": 0.9453125,
|
| 1985 |
+
"step": 304
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"completion_length": 369.453125,
|
| 1989 |
+
"epoch": 0.0544,
|
| 1990 |
+
"grad_norm": 0.10132439230172154,
|
| 1991 |
+
"kl": 0.143829345703125,
|
| 1992 |
+
"learning_rate": 1.2292575559495143e-07,
|
| 1993 |
+
"loss": 0.0001,
|
| 1994 |
+
"reward": 1.4609375,
|
| 1995 |
+
"reward_std": 0.22584448382258415,
|
| 1996 |
+
"rewards/equation_reward_func": 0.5,
|
| 1997 |
+
"rewards/format_reward_func": 0.9609375,
|
| 1998 |
+
"step": 306
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"completion_length": 424.7109375,
|
| 2002 |
+
"epoch": 0.05475555555555556,
|
| 2003 |
+
"grad_norm": 0.14438365212654558,
|
| 2004 |
+
"kl": 0.131591796875,
|
| 2005 |
+
"learning_rate": 1.1983644026655835e-07,
|
| 2006 |
+
"loss": 0.0001,
|
| 2007 |
+
"reward": 1.3203125,
|
| 2008 |
+
"reward_std": 0.2513565141707659,
|
| 2009 |
+
"rewards/equation_reward_func": 0.359375,
|
| 2010 |
+
"rewards/format_reward_func": 0.9609375,
|
| 2011 |
+
"step": 308
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"completion_length": 346.9375,
|
| 2015 |
+
"epoch": 0.05511111111111111,
|
| 2016 |
+
"grad_norm": 0.15713203326183722,
|
| 2017 |
+
"kl": 0.12139892578125,
|
| 2018 |
+
"learning_rate": 1.1677415634565066e-07,
|
| 2019 |
+
"loss": 0.0001,
|
| 2020 |
+
"reward": 1.546875,
|
| 2021 |
+
"reward_std": 0.1645657829940319,
|
| 2022 |
+
"rewards/equation_reward_func": 0.578125,
|
| 2023 |
+
"rewards/format_reward_func": 0.96875,
|
| 2024 |
+
"step": 310
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"completion_length": 373.9453125,
|
| 2028 |
+
"epoch": 0.055466666666666664,
|
| 2029 |
+
"grad_norm": 0.08389325726469715,
|
| 2030 |
+
"kl": 0.11663818359375,
|
| 2031 |
+
"learning_rate": 1.1373953978475353e-07,
|
| 2032 |
+
"loss": 0.0001,
|
| 2033 |
+
"reward": 1.40625,
|
| 2034 |
+
"reward_std": 0.21595830656588078,
|
| 2035 |
+
"rewards/equation_reward_func": 0.421875,
|
| 2036 |
+
"rewards/format_reward_func": 0.984375,
|
| 2037 |
+
"step": 312
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"completion_length": 363.4140625,
|
| 2041 |
+
"epoch": 0.055822222222222224,
|
| 2042 |
+
"grad_norm": 0.14102614011841097,
|
| 2043 |
+
"kl": 0.332305908203125,
|
| 2044 |
+
"learning_rate": 1.1073322079063913e-07,
|
| 2045 |
+
"loss": 0.0003,
|
| 2046 |
+
"reward": 1.484375,
|
| 2047 |
+
"reward_std": 0.19039457477629185,
|
| 2048 |
+
"rewards/equation_reward_func": 0.5,
|
| 2049 |
+
"rewards/format_reward_func": 0.984375,
|
| 2050 |
+
"step": 314
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"completion_length": 443.2109375,
|
| 2054 |
+
"epoch": 0.05617777777777778,
|
| 2055 |
+
"grad_norm": 0.19363787374064304,
|
| 2056 |
+
"kl": 0.104736328125,
|
| 2057 |
+
"learning_rate": 1.0775582369344946e-07,
|
| 2058 |
+
"loss": 0.0001,
|
| 2059 |
+
"reward": 1.2109375,
|
| 2060 |
+
"reward_std": 0.15991678275167942,
|
| 2061 |
+
"rewards/equation_reward_func": 0.2265625,
|
| 2062 |
+
"rewards/format_reward_func": 0.984375,
|
| 2063 |
+
"step": 316
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"completion_length": 382.5703125,
|
| 2067 |
+
"epoch": 0.05653333333333333,
|
| 2068 |
+
"grad_norm": 0.09292657488720626,
|
| 2069 |
+
"kl": 0.16693115234375,
|
| 2070 |
+
"learning_rate": 1.0480796681704077e-07,
|
| 2071 |
+
"loss": 0.0002,
|
| 2072 |
+
"reward": 1.4375,
|
| 2073 |
+
"reward_std": 0.2001592293381691,
|
| 2074 |
+
"rewards/equation_reward_func": 0.46875,
|
| 2075 |
+
"rewards/format_reward_func": 0.96875,
|
| 2076 |
+
"step": 318
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"completion_length": 327.9765625,
|
| 2080 |
+
"epoch": 0.05688888888888889,
|
| 2081 |
+
"grad_norm": 0.10755681115568093,
|
| 2082 |
+
"kl": 0.128448486328125,
|
| 2083 |
+
"learning_rate": 1.018902623505741e-07,
|
| 2084 |
+
"loss": 0.0001,
|
| 2085 |
+
"reward": 1.4765625,
|
| 2086 |
+
"reward_std": 0.20858673192560673,
|
| 2087 |
+
"rewards/equation_reward_func": 0.5,
|
| 2088 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2089 |
+
"step": 320
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"completion_length": 391.078125,
|
| 2093 |
+
"epoch": 0.057244444444444445,
|
| 2094 |
+
"grad_norm": 0.1090245479572572,
|
| 2095 |
+
"kl": 0.11785888671875,
|
| 2096 |
+
"learning_rate": 9.900331622138063e-08,
|
| 2097 |
+
"loss": 0.0001,
|
| 2098 |
+
"reward": 1.4765625,
|
| 2099 |
+
"reward_std": 0.1649293415248394,
|
| 2100 |
+
"rewards/equation_reward_func": 0.4921875,
|
| 2101 |
+
"rewards/format_reward_func": 0.984375,
|
| 2102 |
+
"step": 322
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"completion_length": 355.6015625,
|
| 2106 |
+
"epoch": 0.0576,
|
| 2107 |
+
"grad_norm": 0.10346963676894178,
|
| 2108 |
+
"kl": 0.13653564453125,
|
| 2109 |
+
"learning_rate": 9.614772796912681e-08,
|
| 2110 |
+
"loss": 0.0001,
|
| 2111 |
+
"reward": 1.5,
|
| 2112 |
+
"reward_std": 0.15650183893740177,
|
| 2113 |
+
"rewards/equation_reward_func": 0.515625,
|
| 2114 |
+
"rewards/format_reward_func": 0.984375,
|
| 2115 |
+
"step": 324
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"completion_length": 356.0,
|
| 2119 |
+
"epoch": 0.05795555555555556,
|
| 2120 |
+
"grad_norm": 0.1293537003023565,
|
| 2121 |
+
"kl": 0.134002685546875,
|
| 2122 |
+
"learning_rate": 9.332409062130686e-08,
|
| 2123 |
+
"loss": 0.0001,
|
| 2124 |
+
"reward": 1.3984375,
|
| 2125 |
+
"reward_std": 0.19833699613809586,
|
| 2126 |
+
"rewards/equation_reward_func": 0.421875,
|
| 2127 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2128 |
+
"step": 326
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"completion_length": 385.75,
|
| 2132 |
+
"epoch": 0.05831111111111111,
|
| 2133 |
+
"grad_norm": 0.09266992787765642,
|
| 2134 |
+
"kl": 0.327056884765625,
|
| 2135 |
+
"learning_rate": 9.053299057008699e-08,
|
| 2136 |
+
"loss": 0.0003,
|
| 2137 |
+
"reward": 1.328125,
|
| 2138 |
+
"reward_std": 0.09863808192312717,
|
| 2139 |
+
"rewards/equation_reward_func": 0.34375,
|
| 2140 |
+
"rewards/format_reward_func": 0.984375,
|
| 2141 |
+
"step": 328
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"completion_length": 448.9609375,
|
| 2145 |
+
"epoch": 0.058666666666666666,
|
| 2146 |
+
"grad_norm": 0.10795996998035878,
|
| 2147 |
+
"kl": 0.119293212890625,
|
| 2148 |
+
"learning_rate": 8.777500745052743e-08,
|
| 2149 |
+
"loss": 0.0001,
|
| 2150 |
+
"reward": 1.2421875,
|
| 2151 |
+
"reward_std": 0.15308689512312412,
|
| 2152 |
+
"rewards/equation_reward_func": 0.25,
|
| 2153 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2154 |
+
"step": 330
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"completion_length": 382.96875,
|
| 2158 |
+
"epoch": 0.05902222222222222,
|
| 2159 |
+
"grad_norm": 0.16527711592141753,
|
| 2160 |
+
"kl": 0.118896484375,
|
| 2161 |
+
"learning_rate": 8.505071402020892e-08,
|
| 2162 |
+
"loss": 0.0001,
|
| 2163 |
+
"reward": 1.46875,
|
| 2164 |
+
"reward_std": 0.2238539308309555,
|
| 2165 |
+
"rewards/equation_reward_func": 0.4921875,
|
| 2166 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2167 |
+
"step": 332
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"completion_length": 330.5625,
|
| 2171 |
+
"epoch": 0.05937777777777778,
|
| 2172 |
+
"grad_norm": 0.09535910660673434,
|
| 2173 |
+
"kl": 0.156768798828125,
|
| 2174 |
+
"learning_rate": 8.236067604028562e-08,
|
| 2175 |
+
"loss": 0.0002,
|
| 2176 |
+
"reward": 1.4921875,
|
| 2177 |
+
"reward_std": 0.17859892547130585,
|
| 2178 |
+
"rewards/equation_reward_func": 0.5078125,
|
| 2179 |
+
"rewards/format_reward_func": 0.984375,
|
| 2180 |
+
"step": 334
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"completion_length": 331.328125,
|
| 2184 |
+
"epoch": 0.05973333333333333,
|
| 2185 |
+
"grad_norm": 0.10463153558356299,
|
| 2186 |
+
"kl": 0.165557861328125,
|
| 2187 |
+
"learning_rate": 7.970545215799327e-08,
|
| 2188 |
+
"loss": 0.0002,
|
| 2189 |
+
"reward": 1.5703125,
|
| 2190 |
+
"reward_std": 0.15360604040324688,
|
| 2191 |
+
"rewards/equation_reward_func": 0.5703125,
|
| 2192 |
+
"rewards/format_reward_func": 1.0,
|
| 2193 |
+
"step": 336
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"completion_length": 355.625,
|
| 2197 |
+
"epoch": 0.060088888888888886,
|
| 2198 |
+
"grad_norm": 0.1913964658824772,
|
| 2199 |
+
"kl": 0.151611328125,
|
| 2200 |
+
"learning_rate": 7.708559379063204e-08,
|
| 2201 |
+
"loss": 0.0002,
|
| 2202 |
+
"reward": 1.4609375,
|
| 2203 |
+
"reward_std": 0.25224412232637405,
|
| 2204 |
+
"rewards/equation_reward_func": 0.484375,
|
| 2205 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2206 |
+
"step": 338
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"completion_length": 404.421875,
|
| 2210 |
+
"epoch": 0.060444444444444446,
|
| 2211 |
+
"grad_norm": 0.14713103619518464,
|
| 2212 |
+
"kl": 0.11785888671875,
|
| 2213 |
+
"learning_rate": 7.45016450110534e-08,
|
| 2214 |
+
"loss": 0.0001,
|
| 2215 |
+
"reward": 1.40625,
|
| 2216 |
+
"reward_std": 0.32176715694367886,
|
| 2217 |
+
"rewards/equation_reward_func": 0.4375,
|
| 2218 |
+
"rewards/format_reward_func": 0.96875,
|
| 2219 |
+
"step": 340
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"completion_length": 422.7421875,
|
| 2223 |
+
"epoch": 0.0608,
|
| 2224 |
+
"grad_norm": 0.1327901388939736,
|
| 2225 |
+
"kl": 0.125274658203125,
|
| 2226 |
+
"learning_rate": 7.195414243467029e-08,
|
| 2227 |
+
"loss": 0.0001,
|
| 2228 |
+
"reward": 1.359375,
|
| 2229 |
+
"reward_std": 0.17464719898998737,
|
| 2230 |
+
"rewards/equation_reward_func": 0.359375,
|
| 2231 |
+
"rewards/format_reward_func": 1.0,
|
| 2232 |
+
"step": 342
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"completion_length": 405.5078125,
|
| 2236 |
+
"epoch": 0.06115555555555555,
|
| 2237 |
+
"grad_norm": 0.15507152679168051,
|
| 2238 |
+
"kl": 0.1173095703125,
|
| 2239 |
+
"learning_rate": 6.944361510801763e-08,
|
| 2240 |
+
"loss": 0.0001,
|
| 2241 |
+
"reward": 1.4609375,
|
| 2242 |
+
"reward_std": 0.2535423059016466,
|
| 2243 |
+
"rewards/equation_reward_func": 0.46875,
|
| 2244 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2245 |
+
"step": 344
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"completion_length": 390.125,
|
| 2249 |
+
"epoch": 0.061511111111111114,
|
| 2250 |
+
"grad_norm": 0.09613630702325471,
|
| 2251 |
+
"kl": 0.133941650390625,
|
| 2252 |
+
"learning_rate": 6.697058439888283e-08,
|
| 2253 |
+
"loss": 0.0001,
|
| 2254 |
+
"reward": 1.3828125,
|
| 2255 |
+
"reward_std": 0.16781240701675415,
|
| 2256 |
+
"rewards/equation_reward_func": 0.40625,
|
| 2257 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2258 |
+
"step": 346
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"completion_length": 322.09375,
|
| 2262 |
+
"epoch": 0.06186666666666667,
|
| 2263 |
+
"grad_norm": 0.13920858828391514,
|
| 2264 |
+
"kl": 0.144439697265625,
|
| 2265 |
+
"learning_rate": 6.453556388803288e-08,
|
| 2266 |
+
"loss": 0.0001,
|
| 2267 |
+
"reward": 1.5703125,
|
| 2268 |
+
"reward_std": 0.17854722402989864,
|
| 2269 |
+
"rewards/equation_reward_func": 0.5859375,
|
| 2270 |
+
"rewards/format_reward_func": 0.984375,
|
| 2271 |
+
"step": 348
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"completion_length": 442.453125,
|
| 2275 |
+
"epoch": 0.06222222222222222,
|
| 2276 |
+
"grad_norm": 0.10472362379988483,
|
| 2277 |
+
"kl": 0.11431884765625,
|
| 2278 |
+
"learning_rate": 6.213905926255697e-08,
|
| 2279 |
+
"loss": 0.0001,
|
| 2280 |
+
"reward": 1.25,
|
| 2281 |
+
"reward_std": 0.11279274709522724,
|
| 2282 |
+
"rewards/equation_reward_func": 0.2734375,
|
| 2283 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2284 |
+
"step": 350
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"completion_length": 318.703125,
|
| 2288 |
+
"epoch": 0.06257777777777777,
|
| 2289 |
+
"grad_norm": 0.13606380997377993,
|
| 2290 |
+
"kl": 0.167205810546875,
|
| 2291 |
+
"learning_rate": 5.978156821084987e-08,
|
| 2292 |
+
"loss": 0.0002,
|
| 2293 |
+
"reward": 1.6171875,
|
| 2294 |
+
"reward_std": 0.21070839650928974,
|
| 2295 |
+
"rewards/equation_reward_func": 0.625,
|
| 2296 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2297 |
+
"step": 352
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"completion_length": 400.1328125,
|
| 2301 |
+
"epoch": 0.06293333333333333,
|
| 2302 |
+
"grad_norm": 0.06894639690652522,
|
| 2303 |
+
"kl": 0.1187744140625,
|
| 2304 |
+
"learning_rate": 5.7463580319254853e-08,
|
| 2305 |
+
"loss": 0.0001,
|
| 2306 |
+
"reward": 1.375,
|
| 2307 |
+
"reward_std": 0.0883883461356163,
|
| 2308 |
+
"rewards/equation_reward_func": 0.3828125,
|
| 2309 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2310 |
+
"step": 354
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"completion_length": 396.875,
|
| 2314 |
+
"epoch": 0.0632888888888889,
|
| 2315 |
+
"grad_norm": 0.17070490178918976,
|
| 2316 |
+
"kl": 0.140167236328125,
|
| 2317 |
+
"learning_rate": 5.518557697039081e-08,
|
| 2318 |
+
"loss": 0.0001,
|
| 2319 |
+
"reward": 1.375,
|
| 2320 |
+
"reward_std": 0.30904704704880714,
|
| 2321 |
+
"rewards/equation_reward_func": 0.3828125,
|
| 2322 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2323 |
+
"step": 356
|
| 2324 |
+
},
|
| 2325 |
+
{
|
| 2326 |
+
"completion_length": 414.6640625,
|
| 2327 |
+
"epoch": 0.06364444444444445,
|
| 2328 |
+
"grad_norm": 0.15235608953158414,
|
| 2329 |
+
"kl": 0.129058837890625,
|
| 2330 |
+
"learning_rate": 5.294803124318145e-08,
|
| 2331 |
+
"loss": 0.0001,
|
| 2332 |
+
"reward": 1.3671875,
|
| 2333 |
+
"reward_std": 0.34888696670532227,
|
| 2334 |
+
"rewards/equation_reward_func": 0.40625,
|
| 2335 |
+
"rewards/format_reward_func": 0.9609375,
|
| 2336 |
+
"step": 358
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"completion_length": 363.8984375,
|
| 2340 |
+
"epoch": 0.064,
|
| 2341 |
+
"grad_norm": 0.14985273857090523,
|
| 2342 |
+
"kl": 0.124176025390625,
|
| 2343 |
+
"learning_rate": 5.07514078146106e-08,
|
| 2344 |
+
"loss": 0.0001,
|
| 2345 |
+
"reward": 1.4609375,
|
| 2346 |
+
"reward_std": 0.2845958489924669,
|
| 2347 |
+
"rewards/equation_reward_func": 0.4609375,
|
| 2348 |
+
"rewards/format_reward_func": 1.0,
|
| 2349 |
+
"step": 360
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"completion_length": 424.5625,
|
| 2353 |
+
"epoch": 0.06435555555555555,
|
| 2354 |
+
"grad_norm": 0.11123767098843869,
|
| 2355 |
+
"kl": 0.125823974609375,
|
| 2356 |
+
"learning_rate": 4.859616286322094e-08,
|
| 2357 |
+
"loss": 0.0001,
|
| 2358 |
+
"reward": 1.296875,
|
| 2359 |
+
"reward_std": 0.2488291785120964,
|
| 2360 |
+
"rewards/equation_reward_func": 0.3359375,
|
| 2361 |
+
"rewards/format_reward_func": 0.9609375,
|
| 2362 |
+
"step": 362
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"completion_length": 350.8828125,
|
| 2366 |
+
"epoch": 0.06471111111111111,
|
| 2367 |
+
"grad_norm": 0.09977313666369174,
|
| 2368 |
+
"kl": 0.148895263671875,
|
| 2369 |
+
"learning_rate": 4.648274397437829e-08,
|
| 2370 |
+
"loss": 0.0001,
|
| 2371 |
+
"reward": 1.5,
|
| 2372 |
+
"reward_std": 0.1462521031498909,
|
| 2373 |
+
"rewards/equation_reward_func": 0.5078125,
|
| 2374 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2375 |
+
"step": 364
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"completion_length": 298.3125,
|
| 2379 |
+
"epoch": 0.06506666666666666,
|
| 2380 |
+
"grad_norm": 0.14198944071871591,
|
| 2381 |
+
"kl": 0.1343994140625,
|
| 2382 |
+
"learning_rate": 4.4411590047320617e-08,
|
| 2383 |
+
"loss": 0.0001,
|
| 2384 |
+
"reward": 1.6328125,
|
| 2385 |
+
"reward_std": 0.18649454973638058,
|
| 2386 |
+
"rewards/equation_reward_func": 0.6484375,
|
| 2387 |
+
"rewards/format_reward_func": 0.984375,
|
| 2388 |
+
"step": 366
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"completion_length": 405.453125,
|
| 2392 |
+
"epoch": 0.06542222222222223,
|
| 2393 |
+
"grad_norm": 0.12595363853356176,
|
| 2394 |
+
"kl": 0.1236572265625,
|
| 2395 |
+
"learning_rate": 4.2383131204010494e-08,
|
| 2396 |
+
"loss": 0.0001,
|
| 2397 |
+
"reward": 1.3828125,
|
| 2398 |
+
"reward_std": 0.2517249770462513,
|
| 2399 |
+
"rewards/equation_reward_func": 0.4140625,
|
| 2400 |
+
"rewards/format_reward_func": 0.96875,
|
| 2401 |
+
"step": 368
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"completion_length": 337.7890625,
|
| 2405 |
+
"epoch": 0.06577777777777778,
|
| 2406 |
+
"grad_norm": 0.15385943707708435,
|
| 2407 |
+
"kl": 0.12908935546875,
|
| 2408 |
+
"learning_rate": 4.039778869981064e-08,
|
| 2409 |
+
"loss": 0.0001,
|
| 2410 |
+
"reward": 1.5546875,
|
| 2411 |
+
"reward_std": 0.21778053976595402,
|
| 2412 |
+
"rewards/equation_reward_func": 0.5546875,
|
| 2413 |
+
"rewards/format_reward_func": 1.0,
|
| 2414 |
+
"step": 370
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"completion_length": 321.765625,
|
| 2418 |
+
"epoch": 0.06613333333333334,
|
| 2419 |
+
"grad_norm": 0.17092942527418264,
|
| 2420 |
+
"kl": 0.1295166015625,
|
| 2421 |
+
"learning_rate": 3.845597483600049e-08,
|
| 2422 |
+
"loss": 0.0001,
|
| 2423 |
+
"reward": 1.6328125,
|
| 2424 |
+
"reward_std": 0.1701665185391903,
|
| 2425 |
+
"rewards/equation_reward_func": 0.6328125,
|
| 2426 |
+
"rewards/format_reward_func": 1.0,
|
| 2427 |
+
"step": 372
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"completion_length": 403.3671875,
|
| 2431 |
+
"epoch": 0.06648888888888889,
|
| 2432 |
+
"grad_norm": 0.1280050801798591,
|
| 2433 |
+
"kl": 0.208740234375,
|
| 2434 |
+
"learning_rate": 3.655809287415284e-08,
|
| 2435 |
+
"loss": 0.0002,
|
| 2436 |
+
"reward": 1.375,
|
| 2437 |
+
"reward_std": 0.21542152389883995,
|
| 2438 |
+
"rewards/equation_reward_func": 0.3984375,
|
| 2439 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2440 |
+
"step": 374
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"completion_length": 404.90625,
|
| 2444 |
+
"epoch": 0.06684444444444444,
|
| 2445 |
+
"grad_norm": 0.18022228135215232,
|
| 2446 |
+
"kl": 0.125,
|
| 2447 |
+
"learning_rate": 3.4704536952387285e-08,
|
| 2448 |
+
"loss": 0.0001,
|
| 2449 |
+
"reward": 1.375,
|
| 2450 |
+
"reward_std": 0.29555559158325195,
|
| 2451 |
+
"rewards/equation_reward_func": 0.4140625,
|
| 2452 |
+
"rewards/format_reward_func": 0.9609375,
|
| 2453 |
+
"step": 376
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"completion_length": 314.6953125,
|
| 2457 |
+
"epoch": 0.0672,
|
| 2458 |
+
"grad_norm": 0.14022337811828436,
|
| 2459 |
+
"kl": 0.13714599609375,
|
| 2460 |
+
"learning_rate": 3.2895692003518575e-08,
|
| 2461 |
+
"loss": 0.0001,
|
| 2462 |
+
"reward": 1.6171875,
|
| 2463 |
+
"reward_std": 0.24056999571621418,
|
| 2464 |
+
"rewards/equation_reward_func": 0.6484375,
|
| 2465 |
+
"rewards/format_reward_func": 0.96875,
|
| 2466 |
+
"step": 378
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"completion_length": 370.1875,
|
| 2470 |
+
"epoch": 0.06755555555555555,
|
| 2471 |
+
"grad_norm": 0.0019306138405977334,
|
| 2472 |
+
"kl": 0.1578369140625,
|
| 2473 |
+
"learning_rate": 3.113193367511635e-08,
|
| 2474 |
+
"loss": 0.0002,
|
| 2475 |
+
"reward": 1.3984375,
|
| 2476 |
+
"reward_std": 0.15756267122924328,
|
| 2477 |
+
"rewards/equation_reward_func": 0.40625,
|
| 2478 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2479 |
+
"step": 380
|
| 2480 |
+
},
|
| 2481 |
+
{
|
| 2482 |
+
"completion_length": 382.078125,
|
| 2483 |
+
"epoch": 0.06791111111111112,
|
| 2484 |
+
"grad_norm": 0.11573688091221201,
|
| 2485 |
+
"kl": 0.148834228515625,
|
| 2486 |
+
"learning_rate": 2.9413628251493934e-08,
|
| 2487 |
+
"loss": 0.0001,
|
| 2488 |
+
"reward": 1.515625,
|
| 2489 |
+
"reward_std": 0.25513992086052895,
|
| 2490 |
+
"rewards/equation_reward_func": 0.5390625,
|
| 2491 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2492 |
+
"step": 382
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"completion_length": 324.90625,
|
| 2496 |
+
"epoch": 0.06826666666666667,
|
| 2497 |
+
"grad_norm": 0.17809327173629308,
|
| 2498 |
+
"kl": 0.16400146484375,
|
| 2499 |
+
"learning_rate": 2.774113257764066e-08,
|
| 2500 |
+
"loss": 0.0002,
|
| 2501 |
+
"reward": 1.5546875,
|
| 2502 |
+
"reward_std": 0.17859892547130585,
|
| 2503 |
+
"rewards/equation_reward_func": 0.5625,
|
| 2504 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2505 |
+
"step": 384
|
| 2506 |
+
},
|
| 2507 |
+
{
|
| 2508 |
+
"completion_length": 332.84375,
|
| 2509 |
+
"epoch": 0.06862222222222222,
|
| 2510 |
+
"grad_norm": 0.06792283881086807,
|
| 2511 |
+
"kl": 0.132476806640625,
|
| 2512 |
+
"learning_rate": 2.611479398511518e-08,
|
| 2513 |
+
"loss": 0.0001,
|
| 2514 |
+
"reward": 1.5390625,
|
| 2515 |
+
"reward_std": 0.1237865537405014,
|
| 2516 |
+
"rewards/equation_reward_func": 0.5546875,
|
| 2517 |
+
"rewards/format_reward_func": 0.984375,
|
| 2518 |
+
"step": 386
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"completion_length": 423.6015625,
|
| 2522 |
+
"epoch": 0.06897777777777778,
|
| 2523 |
+
"grad_norm": 0.13085284140188322,
|
| 2524 |
+
"kl": 0.140411376953125,
|
| 2525 |
+
"learning_rate": 2.4534950219914057e-08,
|
| 2526 |
+
"loss": 0.0001,
|
| 2527 |
+
"reward": 1.296875,
|
| 2528 |
+
"reward_std": 0.21900969184935093,
|
| 2529 |
+
"rewards/equation_reward_func": 0.3359375,
|
| 2530 |
+
"rewards/format_reward_func": 0.9609375,
|
| 2531 |
+
"step": 388
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"completion_length": 375.953125,
|
| 2535 |
+
"epoch": 0.06933333333333333,
|
| 2536 |
+
"grad_norm": 0.21027045668269104,
|
| 2537 |
+
"kl": 0.14263916015625,
|
| 2538 |
+
"learning_rate": 2.300192937233128e-08,
|
| 2539 |
+
"loss": 0.0001,
|
| 2540 |
+
"reward": 1.375,
|
| 2541 |
+
"reward_std": 0.24777325056493282,
|
| 2542 |
+
"rewards/equation_reward_func": 0.390625,
|
| 2543 |
+
"rewards/format_reward_func": 0.984375,
|
| 2544 |
+
"step": 390
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"completion_length": 259.9609375,
|
| 2548 |
+
"epoch": 0.06968888888888888,
|
| 2549 |
+
"grad_norm": 0.06740235405182507,
|
| 2550 |
+
"kl": 0.177703857421875,
|
| 2551 |
+
"learning_rate": 2.1516049808822935e-08,
|
| 2552 |
+
"loss": 0.0002,
|
| 2553 |
+
"reward": 1.7109375,
|
| 2554 |
+
"reward_std": 0.09969891421496868,
|
| 2555 |
+
"rewards/equation_reward_func": 0.71875,
|
| 2556 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2557 |
+
"step": 392
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"completion_length": 408.7734375,
|
| 2561 |
+
"epoch": 0.07004444444444445,
|
| 2562 |
+
"grad_norm": 0.12862412677248317,
|
| 2563 |
+
"kl": 0.12994384765625,
|
| 2564 |
+
"learning_rate": 2.007762010589098e-08,
|
| 2565 |
+
"loss": 0.0001,
|
| 2566 |
+
"reward": 1.3125,
|
| 2567 |
+
"reward_std": 0.25707351230084896,
|
| 2568 |
+
"rewards/equation_reward_func": 0.3828125,
|
| 2569 |
+
"rewards/format_reward_func": 0.9296875,
|
| 2570 |
+
"step": 394
|
| 2571 |
+
},
|
| 2572 |
+
{
|
| 2573 |
+
"completion_length": 376.734375,
|
| 2574 |
+
"epoch": 0.0704,
|
| 2575 |
+
"grad_norm": 0.12418206496249037,
|
| 2576 |
+
"kl": 0.124755859375,
|
| 2577 |
+
"learning_rate": 1.8686938986000627e-08,
|
| 2578 |
+
"loss": 0.0001,
|
| 2579 |
+
"reward": 1.375,
|
| 2580 |
+
"reward_std": 0.20593318901956081,
|
| 2581 |
+
"rewards/equation_reward_func": 0.375,
|
| 2582 |
+
"rewards/format_reward_func": 1.0,
|
| 2583 |
+
"step": 396
|
| 2584 |
+
},
|
| 2585 |
+
{
|
| 2586 |
+
"completion_length": 449.6328125,
|
| 2587 |
+
"epoch": 0.07075555555555556,
|
| 2588 |
+
"grad_norm": 0.19881539319315003,
|
| 2589 |
+
"kl": 0.1287078857421875,
|
| 2590 |
+
"learning_rate": 1.734429525554365e-08,
|
| 2591 |
+
"loss": 0.0001,
|
| 2592 |
+
"reward": 1.2109375,
|
| 2593 |
+
"reward_std": 0.31295688077807426,
|
| 2594 |
+
"rewards/equation_reward_func": 0.28125,
|
| 2595 |
+
"rewards/format_reward_func": 0.9296875,
|
| 2596 |
+
"step": 398
|
| 2597 |
+
},
|
| 2598 |
+
{
|
| 2599 |
+
"completion_length": 375.9140625,
|
| 2600 |
+
"epoch": 0.07111111111111111,
|
| 2601 |
+
"grad_norm": 0.14489917455194734,
|
| 2602 |
+
"kl": 0.122894287109375,
|
| 2603 |
+
"learning_rate": 1.604996774486145e-08,
|
| 2604 |
+
"loss": 0.0001,
|
| 2605 |
+
"reward": 1.4296875,
|
| 2606 |
+
"reward_std": 0.30127294547855854,
|
| 2607 |
+
"rewards/equation_reward_func": 0.4609375,
|
| 2608 |
+
"rewards/format_reward_func": 0.96875,
|
| 2609 |
+
"step": 400
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"completion_length": 274.84375,
|
| 2613 |
+
"epoch": 0.07146666666666666,
|
| 2614 |
+
"grad_norm": 0.2041697074289669,
|
| 2615 |
+
"kl": 0.155670166015625,
|
| 2616 |
+
"learning_rate": 1.4804225250339281e-08,
|
| 2617 |
+
"loss": 0.0002,
|
| 2618 |
+
"reward": 1.640625,
|
| 2619 |
+
"reward_std": 0.15992168709635735,
|
| 2620 |
+
"rewards/equation_reward_func": 0.65625,
|
| 2621 |
+
"rewards/format_reward_func": 0.984375,
|
| 2622 |
+
"step": 402
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"completion_length": 309.875,
|
| 2626 |
+
"epoch": 0.07182222222222222,
|
| 2627 |
+
"grad_norm": 0.11843339954382205,
|
| 2628 |
+
"kl": 0.169708251953125,
|
| 2629 |
+
"learning_rate": 1.360732647858498e-08,
|
| 2630 |
+
"loss": 0.0002,
|
| 2631 |
+
"reward": 1.625,
|
| 2632 |
+
"reward_std": 0.17176413349807262,
|
| 2633 |
+
"rewards/equation_reward_func": 0.640625,
|
| 2634 |
+
"rewards/format_reward_func": 0.984375,
|
| 2635 |
+
"step": 404
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"completion_length": 421.8984375,
|
| 2639 |
+
"epoch": 0.07217777777777777,
|
| 2640 |
+
"grad_norm": 0.1323247597342515,
|
| 2641 |
+
"kl": 0.11968994140625,
|
| 2642 |
+
"learning_rate": 1.2459519992702311e-08,
|
| 2643 |
+
"loss": 0.0001,
|
| 2644 |
+
"reward": 1.3671875,
|
| 2645 |
+
"reward_std": 0.28508093021810055,
|
| 2646 |
+
"rewards/equation_reward_func": 0.3984375,
|
| 2647 |
+
"rewards/format_reward_func": 0.96875,
|
| 2648 |
+
"step": 406
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"completion_length": 353.96875,
|
| 2652 |
+
"epoch": 0.07253333333333334,
|
| 2653 |
+
"grad_norm": 0.13268080952152958,
|
| 2654 |
+
"kl": 0.137847900390625,
|
| 2655 |
+
"learning_rate": 1.1361044160671629e-08,
|
| 2656 |
+
"loss": 0.0001,
|
| 2657 |
+
"reward": 1.4765625,
|
| 2658 |
+
"reward_std": 0.15756267122924328,
|
| 2659 |
+
"rewards/equation_reward_func": 0.515625,
|
| 2660 |
+
"rewards/format_reward_func": 0.9609375,
|
| 2661 |
+
"step": 408
|
| 2662 |
+
},
|
| 2663 |
+
{
|
| 2664 |
+
"completion_length": 382.5078125,
|
| 2665 |
+
"epoch": 0.07288888888888889,
|
| 2666 |
+
"grad_norm": 0.08449676532401729,
|
| 2667 |
+
"kl": 0.125823974609375,
|
| 2668 |
+
"learning_rate": 1.0312127105846947e-08,
|
| 2669 |
+
"loss": 0.0001,
|
| 2670 |
+
"reward": 1.484375,
|
| 2671 |
+
"reward_std": 0.13204573653638363,
|
| 2672 |
+
"rewards/equation_reward_func": 0.484375,
|
| 2673 |
+
"rewards/format_reward_func": 1.0,
|
| 2674 |
+
"step": 410
|
| 2675 |
+
},
|
| 2676 |
+
{
|
| 2677 |
+
"completion_length": 325.953125,
|
| 2678 |
+
"epoch": 0.07324444444444445,
|
| 2679 |
+
"grad_norm": 0.1977511977193377,
|
| 2680 |
+
"kl": 0.147247314453125,
|
| 2681 |
+
"learning_rate": 9.312986659581301e-09,
|
| 2682 |
+
"loss": 0.0001,
|
| 2683 |
+
"reward": 1.46875,
|
| 2684 |
+
"reward_std": 0.2290911078453064,
|
| 2685 |
+
"rewards/equation_reward_func": 0.46875,
|
| 2686 |
+
"rewards/format_reward_func": 1.0,
|
| 2687 |
+
"step": 412
|
| 2688 |
+
},
|
| 2689 |
+
{
|
| 2690 |
+
"completion_length": 381.5390625,
|
| 2691 |
+
"epoch": 0.0736,
|
| 2692 |
+
"grad_norm": 0.13775437493448475,
|
| 2693 |
+
"kl": 0.1502685546875,
|
| 2694 |
+
"learning_rate": 8.363830315988945e-09,
|
| 2695 |
+
"loss": 0.0002,
|
| 2696 |
+
"reward": 1.453125,
|
| 2697 |
+
"reward_std": 0.2527858093380928,
|
| 2698 |
+
"rewards/equation_reward_func": 0.4921875,
|
| 2699 |
+
"rewards/format_reward_func": 0.9609375,
|
| 2700 |
+
"step": 414
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"completion_length": 391.96875,
|
| 2704 |
+
"epoch": 0.07395555555555555,
|
| 2705 |
+
"grad_norm": 0.16627374730160457,
|
| 2706 |
+
"kl": 0.130828857421875,
|
| 2707 |
+
"learning_rate": 7.46485518885462e-09,
|
| 2708 |
+
"loss": 0.0001,
|
| 2709 |
+
"reward": 1.4375,
|
| 2710 |
+
"reward_std": 0.2887158952653408,
|
| 2711 |
+
"rewards/equation_reward_func": 0.46875,
|
| 2712 |
+
"rewards/format_reward_func": 0.96875,
|
| 2713 |
+
"step": 416
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"completion_length": 379.6328125,
|
| 2717 |
+
"epoch": 0.0743111111111111,
|
| 2718 |
+
"grad_norm": 0.20602437534961654,
|
| 2719 |
+
"kl": 0.15301513671875,
|
| 2720 |
+
"learning_rate": 6.616247970698319e-09,
|
| 2721 |
+
"loss": 0.0002,
|
| 2722 |
+
"reward": 1.4453125,
|
| 2723 |
+
"reward_std": 0.21095065213739872,
|
| 2724 |
+
"rewards/equation_reward_func": 0.4609375,
|
| 2725 |
+
"rewards/format_reward_func": 0.984375,
|
| 2726 |
+
"step": 418
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"completion_length": 443.8515625,
|
| 2730 |
+
"epoch": 0.07466666666666667,
|
| 2731 |
+
"grad_norm": 0.0967545808782271,
|
| 2732 |
+
"kl": 0.118255615234375,
|
| 2733 |
+
"learning_rate": 5.8181848940044855e-09,
|
| 2734 |
+
"loss": 0.0001,
|
| 2735 |
+
"reward": 1.2578125,
|
| 2736 |
+
"reward_std": 0.2395910546183586,
|
| 2737 |
+
"rewards/equation_reward_func": 0.3125,
|
| 2738 |
+
"rewards/format_reward_func": 0.9453125,
|
| 2739 |
+
"step": 420
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"completion_length": 428.65625,
|
| 2743 |
+
"epoch": 0.07502222222222223,
|
| 2744 |
+
"grad_norm": 0.15101166210394215,
|
| 2745 |
+
"kl": 0.174652099609375,
|
| 2746 |
+
"learning_rate": 5.070831694623135e-09,
|
| 2747 |
+
"loss": 0.0002,
|
| 2748 |
+
"reward": 1.3125,
|
| 2749 |
+
"reward_std": 0.23410366661846638,
|
| 2750 |
+
"rewards/equation_reward_func": 0.3515625,
|
| 2751 |
+
"rewards/format_reward_func": 0.9609375,
|
| 2752 |
+
"step": 422
|
| 2753 |
+
},
|
| 2754 |
+
{
|
| 2755 |
+
"completion_length": 395.5078125,
|
| 2756 |
+
"epoch": 0.07537777777777778,
|
| 2757 |
+
"grad_norm": 0.15966375032382485,
|
| 2758 |
+
"kl": 0.12115478515625,
|
| 2759 |
+
"learning_rate": 4.374343577351336e-09,
|
| 2760 |
+
"loss": 0.0001,
|
| 2761 |
+
"reward": 1.3515625,
|
| 2762 |
+
"reward_std": 0.3044727724045515,
|
| 2763 |
+
"rewards/equation_reward_func": 0.390625,
|
| 2764 |
+
"rewards/format_reward_func": 0.9609375,
|
| 2765 |
+
"step": 424
|
| 2766 |
+
},
|
| 2767 |
+
{
|
| 2768 |
+
"completion_length": 358.5625,
|
| 2769 |
+
"epoch": 0.07573333333333333,
|
| 2770 |
+
"grad_norm": 0.11716508707747059,
|
| 2771 |
+
"kl": 0.128387451171875,
|
| 2772 |
+
"learning_rate": 3.7288651837012745e-09,
|
| 2773 |
+
"loss": 0.0001,
|
| 2774 |
+
"reward": 1.5390625,
|
| 2775 |
+
"reward_std": 0.19728106819093227,
|
| 2776 |
+
"rewards/equation_reward_func": 0.5625,
|
| 2777 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2778 |
+
"step": 426
|
| 2779 |
+
},
|
| 2780 |
+
{
|
| 2781 |
+
"completion_length": 399.2109375,
|
| 2782 |
+
"epoch": 0.07608888888888889,
|
| 2783 |
+
"grad_norm": 0.06036882983312084,
|
| 2784 |
+
"kl": 0.11102294921875,
|
| 2785 |
+
"learning_rate": 3.134530561862081e-09,
|
| 2786 |
+
"loss": 0.0001,
|
| 2787 |
+
"reward": 1.4765625,
|
| 2788 |
+
"reward_std": 0.14465448819100857,
|
| 2789 |
+
"rewards/equation_reward_func": 0.4765625,
|
| 2790 |
+
"rewards/format_reward_func": 1.0,
|
| 2791 |
+
"step": 428
|
| 2792 |
+
},
|
| 2793 |
+
{
|
| 2794 |
+
"completion_length": 382.3359375,
|
| 2795 |
+
"epoch": 0.07644444444444444,
|
| 2796 |
+
"grad_norm": 0.612962185519629,
|
| 2797 |
+
"kl": 0.1231689453125,
|
| 2798 |
+
"learning_rate": 2.5914631388619103e-09,
|
| 2799 |
+
"loss": 0.0001,
|
| 2800 |
+
"reward": 1.4765625,
|
| 2801 |
+
"reward_std": 0.15756267122924328,
|
| 2802 |
+
"rewards/equation_reward_func": 0.4921875,
|
| 2803 |
+
"rewards/format_reward_func": 0.984375,
|
| 2804 |
+
"step": 430
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"completion_length": 241.8125,
|
| 2808 |
+
"epoch": 0.0768,
|
| 2809 |
+
"grad_norm": 0.12381992051598001,
|
| 2810 |
+
"kl": 0.15179443359375,
|
| 2811 |
+
"learning_rate": 2.0997756949353297e-09,
|
| 2812 |
+
"loss": 0.0002,
|
| 2813 |
+
"reward": 1.8046875,
|
| 2814 |
+
"reward_std": 0.19637582264840603,
|
| 2815 |
+
"rewards/equation_reward_func": 0.8125,
|
| 2816 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2817 |
+
"step": 432
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"completion_length": 392.546875,
|
| 2821 |
+
"epoch": 0.07715555555555556,
|
| 2822 |
+
"grad_norm": 0.10938935972335283,
|
| 2823 |
+
"kl": 0.1282958984375,
|
| 2824 |
+
"learning_rate": 1.6595703401020844e-09,
|
| 2825 |
+
"loss": 0.0001,
|
| 2826 |
+
"reward": 1.4296875,
|
| 2827 |
+
"reward_std": 0.16834918968379498,
|
| 2828 |
+
"rewards/equation_reward_func": 0.4453125,
|
| 2829 |
+
"rewards/format_reward_func": 0.984375,
|
| 2830 |
+
"step": 434
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"completion_length": 301.4609375,
|
| 2834 |
+
"epoch": 0.07751111111111111,
|
| 2835 |
+
"grad_norm": 0.11017165117960269,
|
| 2836 |
+
"kl": 0.1378173828125,
|
| 2837 |
+
"learning_rate": 1.2709384929615596e-09,
|
| 2838 |
+
"loss": 0.0001,
|
| 2839 |
+
"reward": 1.6640625,
|
| 2840 |
+
"reward_std": 0.18630647659301758,
|
| 2841 |
+
"rewards/equation_reward_func": 0.6875,
|
| 2842 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2843 |
+
"step": 436
|
| 2844 |
+
},
|
| 2845 |
+
{
|
| 2846 |
+
"completion_length": 336.8359375,
|
| 2847 |
+
"epoch": 0.07786666666666667,
|
| 2848 |
+
"grad_norm": 0.13100046865623674,
|
| 2849 |
+
"kl": 0.13287353515625,
|
| 2850 |
+
"learning_rate": 9.339608617077165e-10,
|
| 2851 |
+
"loss": 0.0001,
|
| 2852 |
+
"reward": 1.5546875,
|
| 2853 |
+
"reward_std": 0.277237007394433,
|
| 2854 |
+
"rewards/equation_reward_func": 0.5703125,
|
| 2855 |
+
"rewards/format_reward_func": 0.984375,
|
| 2856 |
+
"step": 438
|
| 2857 |
+
},
|
| 2858 |
+
{
|
| 2859 |
+
"completion_length": 412.5625,
|
| 2860 |
+
"epoch": 0.07822222222222222,
|
| 2861 |
+
"grad_norm": 0.13909584980828368,
|
| 2862 |
+
"kl": 0.115570068359375,
|
| 2863 |
+
"learning_rate": 6.487074273681114e-10,
|
| 2864 |
+
"loss": 0.0001,
|
| 2865 |
+
"reward": 1.3125,
|
| 2866 |
+
"reward_std": 0.21884137205779552,
|
| 2867 |
+
"rewards/equation_reward_func": 0.328125,
|
| 2868 |
+
"rewards/format_reward_func": 0.984375,
|
| 2869 |
+
"step": 440
|
| 2870 |
+
},
|
| 2871 |
+
{
|
| 2872 |
+
"completion_length": 358.8984375,
|
| 2873 |
+
"epoch": 0.07857777777777777,
|
| 2874 |
+
"grad_norm": 0.1263095691692818,
|
| 2875 |
+
"kl": 0.136993408203125,
|
| 2876 |
+
"learning_rate": 4.152374292708538e-10,
|
| 2877 |
+
"loss": 0.0001,
|
| 2878 |
+
"reward": 1.4609375,
|
| 2879 |
+
"reward_std": 0.12073516845703125,
|
| 2880 |
+
"rewards/equation_reward_func": 0.4609375,
|
| 2881 |
+
"rewards/format_reward_func": 1.0,
|
| 2882 |
+
"step": 442
|
| 2883 |
+
},
|
| 2884 |
+
{
|
| 2885 |
+
"completion_length": 414.4140625,
|
| 2886 |
+
"epoch": 0.07893333333333333,
|
| 2887 |
+
"grad_norm": 0.1266338639368109,
|
| 2888 |
+
"kl": 0.13043212890625,
|
| 2889 |
+
"learning_rate": 2.3359935274214204e-10,
|
| 2890 |
+
"loss": 0.0001,
|
| 2891 |
+
"reward": 1.3828125,
|
| 2892 |
+
"reward_std": 0.141234640032053,
|
| 2893 |
+
"rewards/equation_reward_func": 0.390625,
|
| 2894 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2895 |
+
"step": 444
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"completion_length": 314.1640625,
|
| 2899 |
+
"epoch": 0.0792888888888889,
|
| 2900 |
+
"grad_norm": 0.09004410832047449,
|
| 2901 |
+
"kl": 0.15936279296875,
|
| 2902 |
+
"learning_rate": 1.0383091903720665e-10,
|
| 2903 |
+
"loss": 0.0002,
|
| 2904 |
+
"reward": 1.6171875,
|
| 2905 |
+
"reward_std": 0.10994865000247955,
|
| 2906 |
+
"rewards/equation_reward_func": 0.625,
|
| 2907 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2908 |
+
"step": 446
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"completion_length": 471.28125,
|
| 2912 |
+
"epoch": 0.07964444444444445,
|
| 2913 |
+
"grad_norm": 0.06872113403307542,
|
| 2914 |
+
"kl": 0.1243896484375,
|
| 2915 |
+
"learning_rate": 2.595907750671533e-11,
|
| 2916 |
+
"loss": 0.0001,
|
| 2917 |
+
"reward": 1.234375,
|
| 2918 |
+
"reward_std": 0.08337578736245632,
|
| 2919 |
+
"rewards/equation_reward_func": 0.2421875,
|
| 2920 |
+
"rewards/format_reward_func": 0.9921875,
|
| 2921 |
+
"step": 448
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"completion_length": 365.1015625,
|
| 2925 |
+
"epoch": 0.08,
|
| 2926 |
+
"grad_norm": 0.2033994999207491,
|
| 2927 |
+
"kl": 0.13482666015625,
|
| 2928 |
+
"learning_rate": 0.0,
|
| 2929 |
+
"loss": 0.0001,
|
| 2930 |
+
"reward": 1.4609375,
|
| 2931 |
+
"reward_std": 0.22567616403102875,
|
| 2932 |
+
"rewards/equation_reward_func": 0.484375,
|
| 2933 |
+
"rewards/format_reward_func": 0.9765625,
|
| 2934 |
+
"step": 450
|
| 2935 |
+
},
|
| 2936 |
+
{
|
| 2937 |
+
"epoch": 0.08,
|
| 2938 |
+
"step": 450,
|
| 2939 |
+
"total_flos": 0.0,
|
| 2940 |
+
"train_loss": 9.454646318052558e-05,
|
| 2941 |
+
"train_runtime": 16471.8629,
|
| 2942 |
+
"train_samples_per_second": 0.219,
|
| 2943 |
+
"train_steps_per_second": 0.027
|
| 2944 |
+
}
|
| 2945 |
+
],
|
| 2946 |
+
"logging_steps": 2,
|
| 2947 |
+
"max_steps": 450,
|
| 2948 |
+
"num_input_tokens_seen": 0,
|
| 2949 |
+
"num_train_epochs": 1,
|
| 2950 |
+
"save_steps": 25,
|
| 2951 |
+
"stateful_callbacks": {
|
| 2952 |
+
"TrainerControl": {
|
| 2953 |
+
"args": {
|
| 2954 |
+
"should_epoch_stop": false,
|
| 2955 |
+
"should_evaluate": false,
|
| 2956 |
+
"should_log": false,
|
| 2957 |
+
"should_save": true,
|
| 2958 |
+
"should_training_stop": true
|
| 2959 |
+
},
|
| 2960 |
+
"attributes": {}
|
| 2961 |
+
}
|
| 2962 |
+
},
|
| 2963 |
+
"total_flos": 0.0,
|
| 2964 |
+
"train_batch_size": 1,
|
| 2965 |
+
"trial_name": null,
|
| 2966 |
+
"trial_params": null
|
| 2967 |
+
}
|