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/Qwen2.5-0.5B-Instruct
|
| 3 |
+
library_name: transformers
|
| 4 |
+
model_name: qwen-2.5-0.5b-r1-countdown_lr1.0e-6
|
| 5 |
+
tags:
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- trl
|
| 8 |
+
- grpo
|
| 9 |
+
licence: license
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Model Card for qwen-2.5-0.5b-r1-countdown_lr1.0e-6
|
| 13 |
+
|
| 14 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-0.5B-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="hyunw3/qwen-2.5-0.5b-r1-countdown_lr1.0e-6", 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": 0.00035568679777363087,
|
| 4 |
+
"train_runtime": 8958.7654,
|
| 5 |
+
"train_samples": 45000,
|
| 6 |
+
"train_samples_per_second": 1.206,
|
| 7 |
+
"train_steps_per_second": 0.05
|
| 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": 1.1,
|
| 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": 0.00035568679777363087,
|
| 4 |
+
"train_runtime": 8958.7654,
|
| 5 |
+
"train_samples": 45000,
|
| 6 |
+
"train_samples_per_second": 1.206,
|
| 7 |
+
"train_steps_per_second": 0.05
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2967 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.24,
|
| 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": 267.57032012939453,
|
| 13 |
+
"epoch": 0.0010666666666666667,
|
| 14 |
+
"grad_norm": 0.21315283923773715,
|
| 15 |
+
"kl": 0.0,
|
| 16 |
+
"learning_rate": 1.4285714285714285e-07,
|
| 17 |
+
"loss": -0.0,
|
| 18 |
+
"reward": 0.05208333465270698,
|
| 19 |
+
"reward_std": 0.12187675526365638,
|
| 20 |
+
"rewards/equation_reward_func": 0.0,
|
| 21 |
+
"rewards/format_reward_func": 0.05208333465270698,
|
| 22 |
+
"step": 2
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"completion_length": 258.91667556762695,
|
| 26 |
+
"epoch": 0.0021333333333333334,
|
| 27 |
+
"grad_norm": 0.39624029123072774,
|
| 28 |
+
"kl": 0.0005260705947875977,
|
| 29 |
+
"learning_rate": 2.857142857142857e-07,
|
| 30 |
+
"loss": 0.0,
|
| 31 |
+
"reward": 0.09114583535119891,
|
| 32 |
+
"reward_std": 0.2148200012743473,
|
| 33 |
+
"rewards/equation_reward_func": 0.0026041667442768812,
|
| 34 |
+
"rewards/format_reward_func": 0.08854166837409139,
|
| 35 |
+
"step": 4
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"completion_length": 254.1770896911621,
|
| 39 |
+
"epoch": 0.0032,
|
| 40 |
+
"grad_norm": 0.2789454187627699,
|
| 41 |
+
"kl": 0.0005211830139160156,
|
| 42 |
+
"learning_rate": 4.285714285714285e-07,
|
| 43 |
+
"loss": 0.0,
|
| 44 |
+
"reward": 0.08333333465270698,
|
| 45 |
+
"reward_std": 0.20366852125152946,
|
| 46 |
+
"rewards/equation_reward_func": 0.0026041667442768812,
|
| 47 |
+
"rewards/format_reward_func": 0.08072916860692203,
|
| 48 |
+
"step": 6
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"completion_length": 266.03125858306885,
|
| 52 |
+
"epoch": 0.004266666666666667,
|
| 53 |
+
"grad_norm": 0.1672287489912653,
|
| 54 |
+
"kl": 0.0005216598510742188,
|
| 55 |
+
"learning_rate": 5.714285714285714e-07,
|
| 56 |
+
"loss": 0.0,
|
| 57 |
+
"reward": 0.059895834885537624,
|
| 58 |
+
"reward_std": 0.15923613868653774,
|
| 59 |
+
"rewards/equation_reward_func": 0.0,
|
| 60 |
+
"rewards/format_reward_func": 0.059895834885537624,
|
| 61 |
+
"step": 8
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"completion_length": 248.87240409851074,
|
| 65 |
+
"epoch": 0.005333333333333333,
|
| 66 |
+
"grad_norm": 0.3959884726972001,
|
| 67 |
+
"kl": 0.0006082057952880859,
|
| 68 |
+
"learning_rate": 7.142857142857143e-07,
|
| 69 |
+
"loss": 0.0,
|
| 70 |
+
"reward": 0.09635416883975267,
|
| 71 |
+
"reward_std": 0.21937653236091137,
|
| 72 |
+
"rewards/equation_reward_func": 0.0052083334885537624,
|
| 73 |
+
"rewards/format_reward_func": 0.09114583488553762,
|
| 74 |
+
"step": 10
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"completion_length": 245.3203182220459,
|
| 78 |
+
"epoch": 0.0064,
|
| 79 |
+
"grad_norm": 0.28583532187086624,
|
| 80 |
+
"kl": 0.0007441043853759766,
|
| 81 |
+
"learning_rate": 8.57142857142857e-07,
|
| 82 |
+
"loss": 0.0,
|
| 83 |
+
"reward": 0.0781250016298145,
|
| 84 |
+
"reward_std": 0.2006211462430656,
|
| 85 |
+
"rewards/equation_reward_func": 0.0,
|
| 86 |
+
"rewards/format_reward_func": 0.0781250016298145,
|
| 87 |
+
"step": 12
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"completion_length": 232.69531726837158,
|
| 91 |
+
"epoch": 0.007466666666666667,
|
| 92 |
+
"grad_norm": 0.30125607821086126,
|
| 93 |
+
"kl": 0.0013909339904785156,
|
| 94 |
+
"learning_rate": 1e-06,
|
| 95 |
+
"loss": 0.0,
|
| 96 |
+
"reward": 0.1015625016298145,
|
| 97 |
+
"reward_std": 0.22314835852012038,
|
| 98 |
+
"rewards/equation_reward_func": 0.0026041667442768812,
|
| 99 |
+
"rewards/format_reward_func": 0.09895833465270698,
|
| 100 |
+
"step": 14
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"completion_length": 226.94271278381348,
|
| 104 |
+
"epoch": 0.008533333333333334,
|
| 105 |
+
"grad_norm": 0.3849252255439341,
|
| 106 |
+
"kl": 0.00328826904296875,
|
| 107 |
+
"learning_rate": 9.999480818449865e-07,
|
| 108 |
+
"loss": 0.0,
|
| 109 |
+
"reward": 0.14322916930541396,
|
| 110 |
+
"reward_std": 0.2906528012827039,
|
| 111 |
+
"rewards/equation_reward_func": 0.0052083334885537624,
|
| 112 |
+
"rewards/format_reward_func": 0.1380208362825215,
|
| 113 |
+
"step": 16
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"completion_length": 185.41146278381348,
|
| 117 |
+
"epoch": 0.0096,
|
| 118 |
+
"grad_norm": 0.4848458043128939,
|
| 119 |
+
"kl": 0.008299827575683594,
|
| 120 |
+
"learning_rate": 9.997923381619255e-07,
|
| 121 |
+
"loss": 0.0,
|
| 122 |
+
"reward": 0.22395834000781178,
|
| 123 |
+
"reward_std": 0.37577595096081495,
|
| 124 |
+
"rewards/equation_reward_func": 0.0052083334885537624,
|
| 125 |
+
"rewards/format_reward_func": 0.21875000419095159,
|
| 126 |
+
"step": 18
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"completion_length": 127.20573091506958,
|
| 130 |
+
"epoch": 0.010666666666666666,
|
| 131 |
+
"grad_norm": 0.5597326437314926,
|
| 132 |
+
"kl": 0.0244293212890625,
|
| 133 |
+
"learning_rate": 9.995328012945157e-07,
|
| 134 |
+
"loss": 0.0,
|
| 135 |
+
"reward": 0.4505208469927311,
|
| 136 |
+
"reward_std": 0.4992805514484644,
|
| 137 |
+
"rewards/equation_reward_func": 0.01822916720993817,
|
| 138 |
+
"rewards/format_reward_func": 0.4322916753590107,
|
| 139 |
+
"step": 20
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"completion_length": 93.7682318687439,
|
| 143 |
+
"epoch": 0.011733333333333333,
|
| 144 |
+
"grad_norm": 0.5475672226425161,
|
| 145 |
+
"kl": 0.048309326171875,
|
| 146 |
+
"learning_rate": 9.991695251414583e-07,
|
| 147 |
+
"loss": 0.0,
|
| 148 |
+
"reward": 0.5755208488553762,
|
| 149 |
+
"reward_std": 0.4716507475823164,
|
| 150 |
+
"rewards/equation_reward_func": 0.007812500232830644,
|
| 151 |
+
"rewards/format_reward_func": 0.5677083469927311,
|
| 152 |
+
"step": 22
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"completion_length": 84.74739789962769,
|
| 156 |
+
"epoch": 0.0128,
|
| 157 |
+
"grad_norm": 0.6354342945220236,
|
| 158 |
+
"kl": 0.0808868408203125,
|
| 159 |
+
"learning_rate": 9.987025851452636e-07,
|
| 160 |
+
"loss": 0.0001,
|
| 161 |
+
"reward": 0.7291666865348816,
|
| 162 |
+
"reward_std": 0.42415964510291815,
|
| 163 |
+
"rewards/equation_reward_func": 0.0052083334885537624,
|
| 164 |
+
"rewards/format_reward_func": 0.7239583507180214,
|
| 165 |
+
"step": 24
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"completion_length": 82.8072943687439,
|
| 169 |
+
"epoch": 0.013866666666666666,
|
| 170 |
+
"grad_norm": 0.5155554110023303,
|
| 171 |
+
"kl": 0.128814697265625,
|
| 172 |
+
"learning_rate": 9.981320782765846e-07,
|
| 173 |
+
"loss": 0.0001,
|
| 174 |
+
"reward": 0.8750000223517418,
|
| 175 |
+
"reward_std": 0.270578539930284,
|
| 176 |
+
"rewards/equation_reward_func": 0.0026041667442768812,
|
| 177 |
+
"rewards/format_reward_func": 0.8723958544433117,
|
| 178 |
+
"step": 26
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"completion_length": 64.10937762260437,
|
| 182 |
+
"epoch": 0.014933333333333333,
|
| 183 |
+
"grad_norm": 0.3480478612027264,
|
| 184 |
+
"kl": 0.24591064453125,
|
| 185 |
+
"learning_rate": 9.974581230140768e-07,
|
| 186 |
+
"loss": 0.0002,
|
| 187 |
+
"reward": 0.9713541902601719,
|
| 188 |
+
"reward_std": 0.12162131909281015,
|
| 189 |
+
"rewards/equation_reward_func": 0.007812500232830644,
|
| 190 |
+
"rewards/format_reward_func": 0.9635416939854622,
|
| 191 |
+
"step": 28
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"completion_length": 60.281251668930054,
|
| 195 |
+
"epoch": 0.016,
|
| 196 |
+
"grad_norm": 0.005562791609333587,
|
| 197 |
+
"kl": 0.3465576171875,
|
| 198 |
+
"learning_rate": 9.966808593197956e-07,
|
| 199 |
+
"loss": 0.0003,
|
| 200 |
+
"reward": 0.9869791753590107,
|
| 201 |
+
"reward_std": 0.03682847833260894,
|
| 202 |
+
"rewards/equation_reward_func": 0.0,
|
| 203 |
+
"rewards/format_reward_func": 0.9869791753590107,
|
| 204 |
+
"step": 30
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"completion_length": 45.51562637090683,
|
| 208 |
+
"epoch": 0.017066666666666667,
|
| 209 |
+
"grad_norm": 0.03097813949935321,
|
| 210 |
+
"kl": 0.49072265625,
|
| 211 |
+
"learning_rate": 9.958004486101293e-07,
|
| 212 |
+
"loss": 0.0005,
|
| 213 |
+
"reward": 0.9947916716337204,
|
| 214 |
+
"reward_std": 0.014731391333043575,
|
| 215 |
+
"rewards/equation_reward_func": 0.0,
|
| 216 |
+
"rewards/format_reward_func": 0.9947916716337204,
|
| 217 |
+
"step": 32
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"completion_length": 69.03385561704636,
|
| 221 |
+
"epoch": 0.018133333333333335,
|
| 222 |
+
"grad_norm": 0.006598716445330455,
|
| 223 |
+
"kl": 0.3668212890625,
|
| 224 |
+
"learning_rate": 9.948170737222762e-07,
|
| 225 |
+
"loss": 0.0004,
|
| 226 |
+
"reward": 1.0,
|
| 227 |
+
"reward_std": 0.0,
|
| 228 |
+
"rewards/equation_reward_func": 0.0,
|
| 229 |
+
"rewards/format_reward_func": 1.0,
|
| 230 |
+
"step": 34
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"completion_length": 98.88541954755783,
|
| 234 |
+
"epoch": 0.0192,
|
| 235 |
+
"grad_norm": 0.00733244130846419,
|
| 236 |
+
"kl": 0.3455810546875,
|
| 237 |
+
"learning_rate": 9.937309388762758e-07,
|
| 238 |
+
"loss": 0.0003,
|
| 239 |
+
"reward": 1.0,
|
| 240 |
+
"reward_std": 0.0,
|
| 241 |
+
"rewards/equation_reward_func": 0.0,
|
| 242 |
+
"rewards/format_reward_func": 1.0,
|
| 243 |
+
"step": 36
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"completion_length": 79.19531428813934,
|
| 247 |
+
"epoch": 0.020266666666666665,
|
| 248 |
+
"grad_norm": 0.002140189477159276,
|
| 249 |
+
"kl": 0.337646484375,
|
| 250 |
+
"learning_rate": 9.925422696325974e-07,
|
| 251 |
+
"loss": 0.0003,
|
| 252 |
+
"reward": 0.9973958358168602,
|
| 253 |
+
"reward_std": 0.007365695666521788,
|
| 254 |
+
"rewards/equation_reward_func": 0.0,
|
| 255 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 256 |
+
"step": 38
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"completion_length": 58.98437696695328,
|
| 260 |
+
"epoch": 0.021333333333333333,
|
| 261 |
+
"grad_norm": 0.005041926134939414,
|
| 262 |
+
"kl": 0.3271484375,
|
| 263 |
+
"learning_rate": 9.912513128452973e-07,
|
| 264 |
+
"loss": 0.0003,
|
| 265 |
+
"reward": 1.0,
|
| 266 |
+
"reward_std": 0.0,
|
| 267 |
+
"rewards/equation_reward_func": 0.0,
|
| 268 |
+
"rewards/format_reward_func": 1.0,
|
| 269 |
+
"step": 40
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"completion_length": 83.58854389190674,
|
| 273 |
+
"epoch": 0.0224,
|
| 274 |
+
"grad_norm": 0.003080339075344411,
|
| 275 |
+
"kl": 0.3336181640625,
|
| 276 |
+
"learning_rate": 9.898583366107536e-07,
|
| 277 |
+
"loss": 0.0003,
|
| 278 |
+
"reward": 1.0,
|
| 279 |
+
"reward_std": 0.0,
|
| 280 |
+
"rewards/equation_reward_func": 0.0,
|
| 281 |
+
"rewards/format_reward_func": 1.0,
|
| 282 |
+
"step": 42
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"completion_length": 68.85677361488342,
|
| 286 |
+
"epoch": 0.023466666666666667,
|
| 287 |
+
"grad_norm": 0.005847596515205166,
|
| 288 |
+
"kl": 0.355712890625,
|
| 289 |
+
"learning_rate": 9.88363630211991e-07,
|
| 290 |
+
"loss": 0.0004,
|
| 291 |
+
"reward": 1.0,
|
| 292 |
+
"reward_std": 0.0,
|
| 293 |
+
"rewards/equation_reward_func": 0.0,
|
| 294 |
+
"rewards/format_reward_func": 1.0,
|
| 295 |
+
"step": 44
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"completion_length": 81.716148853302,
|
| 299 |
+
"epoch": 0.024533333333333334,
|
| 300 |
+
"grad_norm": 0.006129286662886568,
|
| 301 |
+
"kl": 0.351806640625,
|
| 302 |
+
"learning_rate": 9.867675040586033e-07,
|
| 303 |
+
"loss": 0.0004,
|
| 304 |
+
"reward": 1.0,
|
| 305 |
+
"reward_std": 0.0,
|
| 306 |
+
"rewards/equation_reward_func": 0.0,
|
| 307 |
+
"rewards/format_reward_func": 1.0,
|
| 308 |
+
"step": 46
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"completion_length": 62.87239807844162,
|
| 312 |
+
"epoch": 0.0256,
|
| 313 |
+
"grad_norm": 0.0022686471037775217,
|
| 314 |
+
"kl": 0.3529052734375,
|
| 315 |
+
"learning_rate": 9.850702896222908e-07,
|
| 316 |
+
"loss": 0.0004,
|
| 317 |
+
"reward": 1.0,
|
| 318 |
+
"reward_std": 0.0,
|
| 319 |
+
"rewards/equation_reward_func": 0.0,
|
| 320 |
+
"rewards/format_reward_func": 1.0,
|
| 321 |
+
"step": 48
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"completion_length": 84.13802230358124,
|
| 325 |
+
"epoch": 0.02666666666666667,
|
| 326 |
+
"grad_norm": 0.001575588022391806,
|
| 327 |
+
"kl": 0.3443603515625,
|
| 328 |
+
"learning_rate": 9.83272339368022e-07,
|
| 329 |
+
"loss": 0.0003,
|
| 330 |
+
"reward": 1.0,
|
| 331 |
+
"reward_std": 0.0,
|
| 332 |
+
"rewards/equation_reward_func": 0.0,
|
| 333 |
+
"rewards/format_reward_func": 1.0,
|
| 334 |
+
"step": 50
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"completion_length": 70.85937684774399,
|
| 338 |
+
"epoch": 0.027733333333333332,
|
| 339 |
+
"grad_norm": 0.004893372014791935,
|
| 340 |
+
"kl": 0.3424072265625,
|
| 341 |
+
"learning_rate": 9.813740266808373e-07,
|
| 342 |
+
"loss": 0.0003,
|
| 343 |
+
"reward": 1.0,
|
| 344 |
+
"reward_std": 0.0,
|
| 345 |
+
"rewards/equation_reward_func": 0.0,
|
| 346 |
+
"rewards/format_reward_func": 1.0,
|
| 347 |
+
"step": 52
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"completion_length": 73.53125238418579,
|
| 351 |
+
"epoch": 0.0288,
|
| 352 |
+
"grad_norm": 0.0020037829167036995,
|
| 353 |
+
"kl": 0.3336181640625,
|
| 354 |
+
"learning_rate": 9.793757457883061e-07,
|
| 355 |
+
"loss": 0.0003,
|
| 356 |
+
"reward": 1.0,
|
| 357 |
+
"reward_std": 0.0,
|
| 358 |
+
"rewards/equation_reward_func": 0.0,
|
| 359 |
+
"rewards/format_reward_func": 1.0,
|
| 360 |
+
"step": 54
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"completion_length": 85.58073115348816,
|
| 364 |
+
"epoch": 0.029866666666666666,
|
| 365 |
+
"grad_norm": 0.0052068637869011935,
|
| 366 |
+
"kl": 0.3441162109375,
|
| 367 |
+
"learning_rate": 9.772779116786567e-07,
|
| 368 |
+
"loss": 0.0003,
|
| 369 |
+
"reward": 1.0,
|
| 370 |
+
"reward_std": 0.0,
|
| 371 |
+
"rewards/equation_reward_func": 0.0,
|
| 372 |
+
"rewards/format_reward_func": 1.0,
|
| 373 |
+
"step": 56
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"completion_length": 62.17968964576721,
|
| 377 |
+
"epoch": 0.030933333333333334,
|
| 378 |
+
"grad_norm": 0.004204176702325925,
|
| 379 |
+
"kl": 0.3392333984375,
|
| 380 |
+
"learning_rate": 9.750809600145952e-07,
|
| 381 |
+
"loss": 0.0003,
|
| 382 |
+
"reward": 1.0,
|
| 383 |
+
"reward_std": 0.0,
|
| 384 |
+
"rewards/equation_reward_func": 0.0,
|
| 385 |
+
"rewards/format_reward_func": 1.0,
|
| 386 |
+
"step": 58
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"completion_length": 60.56250220537186,
|
| 390 |
+
"epoch": 0.032,
|
| 391 |
+
"grad_norm": 0.0015647291182938304,
|
| 392 |
+
"kl": 0.35205078125,
|
| 393 |
+
"learning_rate": 9.7278534704283e-07,
|
| 394 |
+
"loss": 0.0004,
|
| 395 |
+
"reward": 1.0,
|
| 396 |
+
"reward_std": 0.0,
|
| 397 |
+
"rewards/equation_reward_func": 0.0,
|
| 398 |
+
"rewards/format_reward_func": 1.0,
|
| 399 |
+
"step": 60
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"completion_length": 63.45312720537186,
|
| 403 |
+
"epoch": 0.03306666666666667,
|
| 404 |
+
"grad_norm": 0.017107550850573614,
|
| 405 |
+
"kl": 0.3740234375,
|
| 406 |
+
"learning_rate": 9.703915494993213e-07,
|
| 407 |
+
"loss": 0.0004,
|
| 408 |
+
"reward": 1.0,
|
| 409 |
+
"reward_std": 0.0,
|
| 410 |
+
"rewards/equation_reward_func": 0.0,
|
| 411 |
+
"rewards/format_reward_func": 1.0,
|
| 412 |
+
"step": 62
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"completion_length": 59.20312708616257,
|
| 416 |
+
"epoch": 0.034133333333333335,
|
| 417 |
+
"grad_norm": 0.002029246366701105,
|
| 418 |
+
"kl": 0.344482421875,
|
| 419 |
+
"learning_rate": 9.67900064510277e-07,
|
| 420 |
+
"loss": 0.0003,
|
| 421 |
+
"reward": 1.0,
|
| 422 |
+
"reward_std": 0.0,
|
| 423 |
+
"rewards/equation_reward_func": 0.0,
|
| 424 |
+
"rewards/format_reward_func": 1.0,
|
| 425 |
+
"step": 64
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"completion_length": 49.76041799783707,
|
| 429 |
+
"epoch": 0.0352,
|
| 430 |
+
"grad_norm": 0.1707108937849788,
|
| 431 |
+
"kl": 0.3961181640625,
|
| 432 |
+
"learning_rate": 9.653114094889126e-07,
|
| 433 |
+
"loss": 0.0004,
|
| 434 |
+
"reward": 0.9973958358168602,
|
| 435 |
+
"reward_std": 0.007365695666521788,
|
| 436 |
+
"rewards/equation_reward_func": 0.0,
|
| 437 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 438 |
+
"step": 66
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"completion_length": 66.42448157072067,
|
| 442 |
+
"epoch": 0.03626666666666667,
|
| 443 |
+
"grad_norm": 0.004363766057327376,
|
| 444 |
+
"kl": 0.3525390625,
|
| 445 |
+
"learning_rate": 9.626261220279987e-07,
|
| 446 |
+
"loss": 0.0004,
|
| 447 |
+
"reward": 1.0,
|
| 448 |
+
"reward_std": 0.0,
|
| 449 |
+
"rewards/equation_reward_func": 0.0,
|
| 450 |
+
"rewards/format_reward_func": 1.0,
|
| 451 |
+
"step": 68
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"completion_length": 79.07031440734863,
|
| 455 |
+
"epoch": 0.037333333333333336,
|
| 456 |
+
"grad_norm": 0.002295566859094707,
|
| 457 |
+
"kl": 0.3387451171875,
|
| 458 |
+
"learning_rate": 9.598447597882179e-07,
|
| 459 |
+
"loss": 0.0003,
|
| 460 |
+
"reward": 1.0,
|
| 461 |
+
"reward_std": 0.0,
|
| 462 |
+
"rewards/equation_reward_func": 0.0,
|
| 463 |
+
"rewards/format_reward_func": 1.0,
|
| 464 |
+
"step": 70
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"completion_length": 85.9947943687439,
|
| 468 |
+
"epoch": 0.0384,
|
| 469 |
+
"grad_norm": 0.0028772515347710507,
|
| 470 |
+
"kl": 0.3404541015625,
|
| 471 |
+
"learning_rate": 9.56967900382354e-07,
|
| 472 |
+
"loss": 0.0003,
|
| 473 |
+
"reward": 1.0,
|
| 474 |
+
"reward_std": 0.0,
|
| 475 |
+
"rewards/equation_reward_func": 0.0,
|
| 476 |
+
"rewards/format_reward_func": 1.0,
|
| 477 |
+
"step": 72
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"completion_length": 118.54166984558105,
|
| 481 |
+
"epoch": 0.039466666666666664,
|
| 482 |
+
"grad_norm": 0.003804107622681421,
|
| 483 |
+
"kl": 0.35693359375,
|
| 484 |
+
"learning_rate": 9.539961412553374e-07,
|
| 485 |
+
"loss": 0.0004,
|
| 486 |
+
"reward": 1.0,
|
| 487 |
+
"reward_std": 0.0,
|
| 488 |
+
"rewards/equation_reward_func": 0.0,
|
| 489 |
+
"rewards/format_reward_func": 1.0,
|
| 490 |
+
"step": 74
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"completion_length": 132.24739825725555,
|
| 494 |
+
"epoch": 0.04053333333333333,
|
| 495 |
+
"grad_norm": 0.0026976286800890377,
|
| 496 |
+
"kl": 0.35357666015625,
|
| 497 |
+
"learning_rate": 9.509300995601719e-07,
|
| 498 |
+
"loss": 0.0004,
|
| 499 |
+
"reward": 1.0,
|
| 500 |
+
"reward_std": 0.0,
|
| 501 |
+
"rewards/equation_reward_func": 0.0,
|
| 502 |
+
"rewards/format_reward_func": 1.0,
|
| 503 |
+
"step": 76
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"completion_length": 150.5286495089531,
|
| 507 |
+
"epoch": 0.0416,
|
| 508 |
+
"grad_norm": 0.0029703461859595446,
|
| 509 |
+
"kl": 0.3365478515625,
|
| 510 |
+
"learning_rate": 9.477704120297696e-07,
|
| 511 |
+
"loss": 0.0003,
|
| 512 |
+
"reward": 1.0,
|
| 513 |
+
"reward_std": 0.0,
|
| 514 |
+
"rewards/equation_reward_func": 0.0,
|
| 515 |
+
"rewards/format_reward_func": 1.0,
|
| 516 |
+
"step": 78
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"completion_length": 132.16927361488342,
|
| 520 |
+
"epoch": 0.042666666666666665,
|
| 521 |
+
"grad_norm": 0.0042900161521961236,
|
| 522 |
+
"kl": 0.347900390625,
|
| 523 |
+
"learning_rate": 9.445177348447186e-07,
|
| 524 |
+
"loss": 0.0003,
|
| 525 |
+
"reward": 1.0,
|
| 526 |
+
"reward_std": 0.0,
|
| 527 |
+
"rewards/equation_reward_func": 0.0,
|
| 528 |
+
"rewards/format_reward_func": 1.0,
|
| 529 |
+
"step": 80
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"completion_length": 148.73958611488342,
|
| 533 |
+
"epoch": 0.04373333333333333,
|
| 534 |
+
"grad_norm": 0.009973192234001046,
|
| 535 |
+
"kl": 0.3470458984375,
|
| 536 |
+
"learning_rate": 9.41172743497012e-07,
|
| 537 |
+
"loss": 0.0003,
|
| 538 |
+
"reward": 0.9973958358168602,
|
| 539 |
+
"reward_std": 0.007365695666521788,
|
| 540 |
+
"rewards/equation_reward_func": 0.0,
|
| 541 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 542 |
+
"step": 82
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"completion_length": 160.5599009990692,
|
| 546 |
+
"epoch": 0.0448,
|
| 547 |
+
"grad_norm": 0.021317337320502894,
|
| 548 |
+
"kl": 0.34527587890625,
|
| 549 |
+
"learning_rate": 9.377361326497673e-07,
|
| 550 |
+
"loss": 0.0003,
|
| 551 |
+
"reward": 1.0,
|
| 552 |
+
"reward_std": 0.0,
|
| 553 |
+
"rewards/equation_reward_func": 0.0,
|
| 554 |
+
"rewards/format_reward_func": 1.0,
|
| 555 |
+
"step": 84
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"completion_length": 214.65365314483643,
|
| 559 |
+
"epoch": 0.04586666666666667,
|
| 560 |
+
"grad_norm": 0.004309834843107017,
|
| 561 |
+
"kl": 0.30181884765625,
|
| 562 |
+
"learning_rate": 9.342086159929629e-07,
|
| 563 |
+
"loss": 0.0003,
|
| 564 |
+
"reward": 1.0,
|
| 565 |
+
"reward_std": 0.0,
|
| 566 |
+
"rewards/equation_reward_func": 0.0,
|
| 567 |
+
"rewards/format_reward_func": 1.0,
|
| 568 |
+
"step": 86
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"completion_length": 297.28125953674316,
|
| 572 |
+
"epoch": 0.046933333333333334,
|
| 573 |
+
"grad_norm": 0.004971116117055498,
|
| 574 |
+
"kl": 0.263671875,
|
| 575 |
+
"learning_rate": 9.305909260952254e-07,
|
| 576 |
+
"loss": 0.0003,
|
| 577 |
+
"reward": 1.0,
|
| 578 |
+
"reward_std": 0.0,
|
| 579 |
+
"rewards/equation_reward_func": 0.0,
|
| 580 |
+
"rewards/format_reward_func": 1.0,
|
| 581 |
+
"step": 88
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"completion_length": 306.1666736602783,
|
| 585 |
+
"epoch": 0.048,
|
| 586 |
+
"grad_norm": 0.00330346016739013,
|
| 587 |
+
"kl": 0.24566650390625,
|
| 588 |
+
"learning_rate": 9.268838142516943e-07,
|
| 589 |
+
"loss": 0.0002,
|
| 590 |
+
"reward": 1.0,
|
| 591 |
+
"reward_std": 0.0,
|
| 592 |
+
"rewards/equation_reward_func": 0.0,
|
| 593 |
+
"rewards/format_reward_func": 1.0,
|
| 594 |
+
"step": 90
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"completion_length": 279.96355056762695,
|
| 598 |
+
"epoch": 0.04906666666666667,
|
| 599 |
+
"grad_norm": 0.00318035450007201,
|
| 600 |
+
"kl": 0.2532958984375,
|
| 601 |
+
"learning_rate": 9.23088050327999e-07,
|
| 602 |
+
"loss": 0.0003,
|
| 603 |
+
"reward": 1.0,
|
| 604 |
+
"reward_std": 0.0,
|
| 605 |
+
"rewards/equation_reward_func": 0.0,
|
| 606 |
+
"rewards/format_reward_func": 1.0,
|
| 607 |
+
"step": 92
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"completion_length": 261.7994842529297,
|
| 611 |
+
"epoch": 0.050133333333333335,
|
| 612 |
+
"grad_norm": 0.0036946086220320164,
|
| 613 |
+
"kl": 0.273681640625,
|
| 614 |
+
"learning_rate": 9.192044226003788e-07,
|
| 615 |
+
"loss": 0.0003,
|
| 616 |
+
"reward": 1.0,
|
| 617 |
+
"reward_std": 0.0,
|
| 618 |
+
"rewards/equation_reward_func": 0.0,
|
| 619 |
+
"rewards/format_reward_func": 1.0,
|
| 620 |
+
"step": 94
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"completion_length": 275.30209159851074,
|
| 624 |
+
"epoch": 0.0512,
|
| 625 |
+
"grad_norm": 0.004650714378941982,
|
| 626 |
+
"kl": 0.27423095703125,
|
| 627 |
+
"learning_rate": 9.15233737591979e-07,
|
| 628 |
+
"loss": 0.0003,
|
| 629 |
+
"reward": 1.0,
|
| 630 |
+
"reward_std": 0.0,
|
| 631 |
+
"rewards/equation_reward_func": 0.0,
|
| 632 |
+
"rewards/format_reward_func": 1.0,
|
| 633 |
+
"step": 96
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"completion_length": 275.5442762374878,
|
| 637 |
+
"epoch": 0.05226666666666667,
|
| 638 |
+
"grad_norm": 0.005874752003273854,
|
| 639 |
+
"kl": 0.283935546875,
|
| 640 |
+
"learning_rate": 9.111768199053586e-07,
|
| 641 |
+
"loss": 0.0003,
|
| 642 |
+
"reward": 1.0,
|
| 643 |
+
"reward_std": 0.0,
|
| 644 |
+
"rewards/equation_reward_func": 0.0,
|
| 645 |
+
"rewards/format_reward_func": 1.0,
|
| 646 |
+
"step": 98
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"completion_length": 224.13542079925537,
|
| 650 |
+
"epoch": 0.05333333333333334,
|
| 651 |
+
"grad_norm": 0.003767361371489657,
|
| 652 |
+
"kl": 0.29339599609375,
|
| 653 |
+
"learning_rate": 9.070345120512435e-07,
|
| 654 |
+
"loss": 0.0003,
|
| 655 |
+
"reward": 1.0,
|
| 656 |
+
"reward_std": 0.0,
|
| 657 |
+
"rewards/equation_reward_func": 0.0,
|
| 658 |
+
"rewards/format_reward_func": 1.0,
|
| 659 |
+
"step": 100
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"completion_length": 247.3671932220459,
|
| 663 |
+
"epoch": 0.0544,
|
| 664 |
+
"grad_norm": 0.21204929441134815,
|
| 665 |
+
"kl": 0.5673828125,
|
| 666 |
+
"learning_rate": 9.028076742735582e-07,
|
| 667 |
+
"loss": 0.0006,
|
| 668 |
+
"reward": 1.0,
|
| 669 |
+
"reward_std": 0.0,
|
| 670 |
+
"rewards/equation_reward_func": 0.0,
|
| 671 |
+
"rewards/format_reward_func": 1.0,
|
| 672 |
+
"step": 102
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"completion_length": 213.47917222976685,
|
| 676 |
+
"epoch": 0.055466666666666664,
|
| 677 |
+
"grad_norm": 0.003470673455225982,
|
| 678 |
+
"kl": 0.3101806640625,
|
| 679 |
+
"learning_rate": 8.984971843707787e-07,
|
| 680 |
+
"loss": 0.0003,
|
| 681 |
+
"reward": 1.0,
|
| 682 |
+
"reward_std": 0.0,
|
| 683 |
+
"rewards/equation_reward_func": 0.0,
|
| 684 |
+
"rewards/format_reward_func": 1.0,
|
| 685 |
+
"step": 104
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"completion_length": 175.9947967529297,
|
| 689 |
+
"epoch": 0.05653333333333333,
|
| 690 |
+
"grad_norm": 0.005812568326490043,
|
| 691 |
+
"kl": 0.32379150390625,
|
| 692 |
+
"learning_rate": 8.94103937513637e-07,
|
| 693 |
+
"loss": 0.0003,
|
| 694 |
+
"reward": 1.0,
|
| 695 |
+
"reward_std": 0.0,
|
| 696 |
+
"rewards/equation_reward_func": 0.0,
|
| 697 |
+
"rewards/format_reward_func": 1.0,
|
| 698 |
+
"step": 106
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"completion_length": 117.411461353302,
|
| 702 |
+
"epoch": 0.0576,
|
| 703 |
+
"grad_norm": 0.00354722170703675,
|
| 704 |
+
"kl": 0.3551025390625,
|
| 705 |
+
"learning_rate": 8.896288460592185e-07,
|
| 706 |
+
"loss": 0.0004,
|
| 707 |
+
"reward": 1.0,
|
| 708 |
+
"reward_std": 0.0,
|
| 709 |
+
"rewards/equation_reward_func": 0.0,
|
| 710 |
+
"rewards/format_reward_func": 1.0,
|
| 711 |
+
"step": 108
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"completion_length": 110.28125357627869,
|
| 715 |
+
"epoch": 0.058666666666666666,
|
| 716 |
+
"grad_norm": 0.003013321323384688,
|
| 717 |
+
"kl": 0.339599609375,
|
| 718 |
+
"learning_rate": 8.850728393614901e-07,
|
| 719 |
+
"loss": 0.0003,
|
| 720 |
+
"reward": 1.0,
|
| 721 |
+
"reward_std": 0.0,
|
| 722 |
+
"rewards/equation_reward_func": 0.0,
|
| 723 |
+
"rewards/format_reward_func": 1.0,
|
| 724 |
+
"step": 110
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"completion_length": 106.96875274181366,
|
| 728 |
+
"epoch": 0.05973333333333333,
|
| 729 |
+
"grad_norm": 0.003463116039679215,
|
| 730 |
+
"kl": 0.35595703125,
|
| 731 |
+
"learning_rate": 8.804368635783002e-07,
|
| 732 |
+
"loss": 0.0004,
|
| 733 |
+
"reward": 1.0,
|
| 734 |
+
"reward_std": 0.0,
|
| 735 |
+
"rewards/equation_reward_func": 0.0,
|
| 736 |
+
"rewards/format_reward_func": 1.0,
|
| 737 |
+
"step": 112
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"completion_length": 85.82552325725555,
|
| 741 |
+
"epoch": 0.0608,
|
| 742 |
+
"grad_norm": 0.03677716489910903,
|
| 743 |
+
"kl": 0.3817138671875,
|
| 744 |
+
"learning_rate": 8.75721881474886e-07,
|
| 745 |
+
"loss": 0.0004,
|
| 746 |
+
"reward": 0.9973958358168602,
|
| 747 |
+
"reward_std": 0.007365695666521788,
|
| 748 |
+
"rewards/equation_reward_func": 0.0,
|
| 749 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 750 |
+
"step": 114
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"completion_length": 69.90885663032532,
|
| 754 |
+
"epoch": 0.06186666666666667,
|
| 755 |
+
"grad_norm": 0.014502649071450845,
|
| 756 |
+
"kl": 0.4462890625,
|
| 757 |
+
"learning_rate": 8.709288722239342e-07,
|
| 758 |
+
"loss": 0.0004,
|
| 759 |
+
"reward": 1.0,
|
| 760 |
+
"reward_std": 0.0,
|
| 761 |
+
"rewards/equation_reward_func": 0.0,
|
| 762 |
+
"rewards/format_reward_func": 1.0,
|
| 763 |
+
"step": 116
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"completion_length": 60.21093875169754,
|
| 767 |
+
"epoch": 0.06293333333333333,
|
| 768 |
+
"grad_norm": 0.018319766426722457,
|
| 769 |
+
"kl": 0.4246826171875,
|
| 770 |
+
"learning_rate": 8.660588312022343e-07,
|
| 771 |
+
"loss": 0.0004,
|
| 772 |
+
"reward": 0.9973958358168602,
|
| 773 |
+
"reward_std": 0.007365695666521788,
|
| 774 |
+
"rewards/equation_reward_func": 0.0,
|
| 775 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 776 |
+
"step": 118
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"completion_length": 48.75000184774399,
|
| 780 |
+
"epoch": 0.064,
|
| 781 |
+
"grad_norm": 0.002819178233147015,
|
| 782 |
+
"kl": 0.3572998046875,
|
| 783 |
+
"learning_rate": 8.611127697839647e-07,
|
| 784 |
+
"loss": 0.0004,
|
| 785 |
+
"reward": 0.9973958358168602,
|
| 786 |
+
"reward_std": 0.007365695666521788,
|
| 787 |
+
"rewards/equation_reward_func": 0.0,
|
| 788 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 789 |
+
"step": 120
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"completion_length": 29.817708730697632,
|
| 793 |
+
"epoch": 0.06506666666666666,
|
| 794 |
+
"grad_norm": 0.048219241978314446,
|
| 795 |
+
"kl": 0.6165771484375,
|
| 796 |
+
"learning_rate": 8.560917151306592e-07,
|
| 797 |
+
"loss": 0.0006,
|
| 798 |
+
"reward": 1.0,
|
| 799 |
+
"reward_std": 0.0,
|
| 800 |
+
"rewards/equation_reward_func": 0.0,
|
| 801 |
+
"rewards/format_reward_func": 1.0,
|
| 802 |
+
"step": 122
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"completion_length": 29.747396767139435,
|
| 806 |
+
"epoch": 0.06613333333333334,
|
| 807 |
+
"grad_norm": 0.002491973715723316,
|
| 808 |
+
"kl": 0.394287109375,
|
| 809 |
+
"learning_rate": 8.509967099778933e-07,
|
| 810 |
+
"loss": 0.0004,
|
| 811 |
+
"reward": 0.9973958358168602,
|
| 812 |
+
"reward_std": 0.007365695666521788,
|
| 813 |
+
"rewards/equation_reward_func": 0.0,
|
| 814 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 815 |
+
"step": 124
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"completion_length": 22.74479252099991,
|
| 819 |
+
"epoch": 0.0672,
|
| 820 |
+
"grad_norm": 0.03159552463178033,
|
| 821 |
+
"kl": 0.519287109375,
|
| 822 |
+
"learning_rate": 8.458288124187358e-07,
|
| 823 |
+
"loss": 0.0005,
|
| 824 |
+
"reward": 1.0,
|
| 825 |
+
"reward_std": 0.0,
|
| 826 |
+
"rewards/equation_reward_func": 0.0,
|
| 827 |
+
"rewards/format_reward_func": 1.0,
|
| 828 |
+
"step": 126
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"completion_length": 18.752604603767395,
|
| 832 |
+
"epoch": 0.06826666666666667,
|
| 833 |
+
"grad_norm": 0.0047626328583877,
|
| 834 |
+
"kl": 0.4564208984375,
|
| 835 |
+
"learning_rate": 8.405890956840135e-07,
|
| 836 |
+
"loss": 0.0005,
|
| 837 |
+
"reward": 0.9973958358168602,
|
| 838 |
+
"reward_std": 0.007365695666521788,
|
| 839 |
+
"rewards/equation_reward_func": 0.0,
|
| 840 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 841 |
+
"step": 128
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"completion_length": 25.395833671092987,
|
| 845 |
+
"epoch": 0.06933333333333333,
|
| 846 |
+
"grad_norm": 0.007120551314141992,
|
| 847 |
+
"kl": 0.4727783203125,
|
| 848 |
+
"learning_rate": 8.352786479194287e-07,
|
| 849 |
+
"loss": 0.0005,
|
| 850 |
+
"reward": 1.0,
|
| 851 |
+
"reward_std": 0.0,
|
| 852 |
+
"rewards/equation_reward_func": 0.0,
|
| 853 |
+
"rewards/format_reward_func": 1.0,
|
| 854 |
+
"step": 130
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"completion_length": 18.677083611488342,
|
| 858 |
+
"epoch": 0.0704,
|
| 859 |
+
"grad_norm": 0.007749349171876843,
|
| 860 |
+
"kl": 0.4617919921875,
|
| 861 |
+
"learning_rate": 8.298985719595823e-07,
|
| 862 |
+
"loss": 0.0005,
|
| 863 |
+
"reward": 1.0,
|
| 864 |
+
"reward_std": 0.0,
|
| 865 |
+
"rewards/equation_reward_func": 0.0,
|
| 866 |
+
"rewards/format_reward_func": 1.0,
|
| 867 |
+
"step": 132
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"completion_length": 28.72395944595337,
|
| 871 |
+
"epoch": 0.07146666666666666,
|
| 872 |
+
"grad_norm": 0.01725236239408364,
|
| 873 |
+
"kl": 0.512451171875,
|
| 874 |
+
"learning_rate": 8.244499850989451e-07,
|
| 875 |
+
"loss": 0.0005,
|
| 876 |
+
"reward": 1.0,
|
| 877 |
+
"reward_std": 0.0,
|
| 878 |
+
"rewards/equation_reward_func": 0.0,
|
| 879 |
+
"rewards/format_reward_func": 1.0,
|
| 880 |
+
"step": 134
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"completion_length": 32.2265630364418,
|
| 884 |
+
"epoch": 0.07253333333333334,
|
| 885 |
+
"grad_norm": 0.05304546123188986,
|
| 886 |
+
"kl": 0.5926513671875,
|
| 887 |
+
"learning_rate": 8.189340188598262e-07,
|
| 888 |
+
"loss": 0.0006,
|
| 889 |
+
"reward": 0.9973958358168602,
|
| 890 |
+
"reward_std": 0.007365695666521788,
|
| 891 |
+
"rewards/equation_reward_func": 0.0,
|
| 892 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 893 |
+
"step": 136
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"completion_length": 23.41145896911621,
|
| 897 |
+
"epoch": 0.0736,
|
| 898 |
+
"grad_norm": 0.013190548453262672,
|
| 899 |
+
"kl": 0.4632568359375,
|
| 900 |
+
"learning_rate": 8.133518187573862e-07,
|
| 901 |
+
"loss": 0.0005,
|
| 902 |
+
"reward": 1.0,
|
| 903 |
+
"reward_std": 0.0,
|
| 904 |
+
"rewards/equation_reward_func": 0.0,
|
| 905 |
+
"rewards/format_reward_func": 1.0,
|
| 906 |
+
"step": 138
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"completion_length": 24.82552134990692,
|
| 910 |
+
"epoch": 0.07466666666666667,
|
| 911 |
+
"grad_norm": 0.012046997929078053,
|
| 912 |
+
"kl": 0.4591064453125,
|
| 913 |
+
"learning_rate": 8.077045440617464e-07,
|
| 914 |
+
"loss": 0.0005,
|
| 915 |
+
"reward": 1.0,
|
| 916 |
+
"reward_std": 0.0,
|
| 917 |
+
"rewards/equation_reward_func": 0.0,
|
| 918 |
+
"rewards/format_reward_func": 1.0,
|
| 919 |
+
"step": 140
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"completion_length": 17.638021290302277,
|
| 923 |
+
"epoch": 0.07573333333333333,
|
| 924 |
+
"grad_norm": 0.0070439803543669265,
|
| 925 |
+
"kl": 0.425048828125,
|
| 926 |
+
"learning_rate": 8.019933675572388e-07,
|
| 927 |
+
"loss": 0.0004,
|
| 928 |
+
"reward": 1.0,
|
| 929 |
+
"reward_std": 0.0,
|
| 930 |
+
"rewards/equation_reward_func": 0.0,
|
| 931 |
+
"rewards/format_reward_func": 1.0,
|
| 932 |
+
"step": 142
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"completion_length": 19.304688096046448,
|
| 936 |
+
"epoch": 0.0768,
|
| 937 |
+
"grad_norm": 0.007119537656182656,
|
| 938 |
+
"kl": 0.952880859375,
|
| 939 |
+
"learning_rate": 7.962194752988518e-07,
|
| 940 |
+
"loss": 0.001,
|
| 941 |
+
"reward": 0.9973958358168602,
|
| 942 |
+
"reward_std": 0.007365695666521788,
|
| 943 |
+
"rewards/equation_reward_func": 0.0,
|
| 944 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 945 |
+
"step": 144
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"completion_length": 23.434896171092987,
|
| 949 |
+
"epoch": 0.07786666666666667,
|
| 950 |
+
"grad_norm": 0.004351846417110043,
|
| 951 |
+
"kl": 0.39892578125,
|
| 952 |
+
"learning_rate": 7.903840663659184e-07,
|
| 953 |
+
"loss": 0.0004,
|
| 954 |
+
"reward": 1.0,
|
| 955 |
+
"reward_std": 0.0,
|
| 956 |
+
"rewards/equation_reward_func": 0.0,
|
| 957 |
+
"rewards/format_reward_func": 1.0,
|
| 958 |
+
"step": 146
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"completion_length": 17.36458384990692,
|
| 962 |
+
"epoch": 0.07893333333333333,
|
| 963 |
+
"grad_norm": 0.014464883726819155,
|
| 964 |
+
"kl": 0.427490234375,
|
| 965 |
+
"learning_rate": 7.844883526131013e-07,
|
| 966 |
+
"loss": 0.0004,
|
| 967 |
+
"reward": 1.0,
|
| 968 |
+
"reward_std": 0.0,
|
| 969 |
+
"rewards/equation_reward_func": 0.0,
|
| 970 |
+
"rewards/format_reward_func": 1.0,
|
| 971 |
+
"step": 148
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"completion_length": 23.164062976837158,
|
| 975 |
+
"epoch": 0.08,
|
| 976 |
+
"grad_norm": 0.003214313828585045,
|
| 977 |
+
"kl": 0.38671875,
|
| 978 |
+
"learning_rate": 7.785335584187219e-07,
|
| 979 |
+
"loss": 0.0004,
|
| 980 |
+
"reward": 1.0,
|
| 981 |
+
"reward_std": 0.0,
|
| 982 |
+
"rewards/equation_reward_func": 0.0,
|
| 983 |
+
"rewards/format_reward_func": 1.0,
|
| 984 |
+
"step": 150
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"completion_length": 17.333333790302277,
|
| 988 |
+
"epoch": 0.08106666666666666,
|
| 989 |
+
"grad_norm": 0.005708191232659279,
|
| 990 |
+
"kl": 0.44482421875,
|
| 991 |
+
"learning_rate": 7.725209204304928e-07,
|
| 992 |
+
"loss": 0.0004,
|
| 993 |
+
"reward": 0.9973958358168602,
|
| 994 |
+
"reward_std": 0.007365695666521788,
|
| 995 |
+
"rewards/equation_reward_func": 0.0,
|
| 996 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 997 |
+
"step": 152
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"completion_length": 18.481771171092987,
|
| 1001 |
+
"epoch": 0.08213333333333334,
|
| 1002 |
+
"grad_norm": 0.18097704136374604,
|
| 1003 |
+
"kl": 0.6190185546875,
|
| 1004 |
+
"learning_rate": 7.664516873086987e-07,
|
| 1005 |
+
"loss": 0.0006,
|
| 1006 |
+
"reward": 0.9947916716337204,
|
| 1007 |
+
"reward_std": 0.014731391333043575,
|
| 1008 |
+
"rewards/equation_reward_func": 0.0,
|
| 1009 |
+
"rewards/format_reward_func": 0.9947916716337204,
|
| 1010 |
+
"step": 154
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"completion_length": 20.242188155651093,
|
| 1014 |
+
"epoch": 0.0832,
|
| 1015 |
+
"grad_norm": 0.004715553205403204,
|
| 1016 |
+
"kl": 0.42333984375,
|
| 1017 |
+
"learning_rate": 7.603271194668835e-07,
|
| 1018 |
+
"loss": 0.0004,
|
| 1019 |
+
"reward": 1.0,
|
| 1020 |
+
"reward_std": 0.0,
|
| 1021 |
+
"rewards/equation_reward_func": 0.0,
|
| 1022 |
+
"rewards/format_reward_func": 1.0,
|
| 1023 |
+
"step": 156
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"completion_length": 18.898438036441803,
|
| 1027 |
+
"epoch": 0.08426666666666667,
|
| 1028 |
+
"grad_norm": 0.003966040759048578,
|
| 1029 |
+
"kl": 0.403076171875,
|
| 1030 |
+
"learning_rate": 7.541484888100973e-07,
|
| 1031 |
+
"loss": 0.0004,
|
| 1032 |
+
"reward": 1.0,
|
| 1033 |
+
"reward_std": 0.0,
|
| 1034 |
+
"rewards/equation_reward_func": 0.0,
|
| 1035 |
+
"rewards/format_reward_func": 1.0,
|
| 1036 |
+
"step": 158
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"completion_length": 16.60416716337204,
|
| 1040 |
+
"epoch": 0.08533333333333333,
|
| 1041 |
+
"grad_norm": 0.002003261027905268,
|
| 1042 |
+
"kl": 0.34228515625,
|
| 1043 |
+
"learning_rate": 7.479170784707574e-07,
|
| 1044 |
+
"loss": 0.0003,
|
| 1045 |
+
"reward": 0.9973958358168602,
|
| 1046 |
+
"reward_std": 0.007365695666521788,
|
| 1047 |
+
"rewards/equation_reward_func": 0.0,
|
| 1048 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1049 |
+
"step": 160
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"completion_length": 24.88802134990692,
|
| 1053 |
+
"epoch": 0.0864,
|
| 1054 |
+
"grad_norm": 0.009211378308009072,
|
| 1055 |
+
"kl": 0.384521484375,
|
| 1056 |
+
"learning_rate": 7.416341825421753e-07,
|
| 1057 |
+
"loss": 0.0004,
|
| 1058 |
+
"reward": 1.0,
|
| 1059 |
+
"reward_std": 0.0,
|
| 1060 |
+
"rewards/equation_reward_func": 0.0,
|
| 1061 |
+
"rewards/format_reward_func": 1.0,
|
| 1062 |
+
"step": 162
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"completion_length": 16.401042103767395,
|
| 1066 |
+
"epoch": 0.08746666666666666,
|
| 1067 |
+
"grad_norm": 0.005184187636080286,
|
| 1068 |
+
"kl": 0.3958740234375,
|
| 1069 |
+
"learning_rate": 7.353011058098103e-07,
|
| 1070 |
+
"loss": 0.0004,
|
| 1071 |
+
"reward": 1.0,
|
| 1072 |
+
"reward_std": 0.0,
|
| 1073 |
+
"rewards/equation_reward_func": 0.0,
|
| 1074 |
+
"rewards/format_reward_func": 1.0,
|
| 1075 |
+
"step": 164
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"completion_length": 21.299479722976685,
|
| 1079 |
+
"epoch": 0.08853333333333334,
|
| 1080 |
+
"grad_norm": 0.07832411386720319,
|
| 1081 |
+
"kl": 0.3946533203125,
|
| 1082 |
+
"learning_rate": 7.289191634803002e-07,
|
| 1083 |
+
"loss": 0.0004,
|
| 1084 |
+
"reward": 0.9947916716337204,
|
| 1085 |
+
"reward_std": 0.014731391333043575,
|
| 1086 |
+
"rewards/equation_reward_func": 0.0,
|
| 1087 |
+
"rewards/format_reward_func": 0.9947916716337204,
|
| 1088 |
+
"step": 166
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"completion_length": 21.127604603767395,
|
| 1092 |
+
"epoch": 0.0896,
|
| 1093 |
+
"grad_norm": 0.006189211303564496,
|
| 1094 |
+
"kl": 0.3651123046875,
|
| 1095 |
+
"learning_rate": 7.224896809083297e-07,
|
| 1096 |
+
"loss": 0.0004,
|
| 1097 |
+
"reward": 1.0,
|
| 1098 |
+
"reward_std": 0.0,
|
| 1099 |
+
"rewards/equation_reward_func": 0.0,
|
| 1100 |
+
"rewards/format_reward_func": 1.0,
|
| 1101 |
+
"step": 168
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"completion_length": 18.304687917232513,
|
| 1105 |
+
"epoch": 0.09066666666666667,
|
| 1106 |
+
"grad_norm": 0.002963052569393672,
|
| 1107 |
+
"kl": 0.345458984375,
|
| 1108 |
+
"learning_rate": 7.160139933213898e-07,
|
| 1109 |
+
"loss": 0.0003,
|
| 1110 |
+
"reward": 1.0,
|
| 1111 |
+
"reward_std": 0.0,
|
| 1112 |
+
"rewards/equation_reward_func": 0.0,
|
| 1113 |
+
"rewards/format_reward_func": 1.0,
|
| 1114 |
+
"step": 170
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"completion_length": 24.544271409511566,
|
| 1118 |
+
"epoch": 0.09173333333333333,
|
| 1119 |
+
"grad_norm": 0.006809551131670742,
|
| 1120 |
+
"kl": 0.376708984375,
|
| 1121 |
+
"learning_rate": 7.094934455424888e-07,
|
| 1122 |
+
"loss": 0.0004,
|
| 1123 |
+
"reward": 1.0,
|
| 1124 |
+
"reward_std": 0.0,
|
| 1125 |
+
"rewards/equation_reward_func": 0.0,
|
| 1126 |
+
"rewards/format_reward_func": 1.0,
|
| 1127 |
+
"step": 172
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"completion_length": 23.06510454416275,
|
| 1131 |
+
"epoch": 0.0928,
|
| 1132 |
+
"grad_norm": 0.007169853628551621,
|
| 1133 |
+
"kl": 0.377685546875,
|
| 1134 |
+
"learning_rate": 7.029293917108677e-07,
|
| 1135 |
+
"loss": 0.0004,
|
| 1136 |
+
"reward": 1.0,
|
| 1137 |
+
"reward_std": 0.0,
|
| 1138 |
+
"rewards/equation_reward_func": 0.0,
|
| 1139 |
+
"rewards/format_reward_func": 1.0,
|
| 1140 |
+
"step": 174
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"completion_length": 29.372397005558014,
|
| 1144 |
+
"epoch": 0.09386666666666667,
|
| 1145 |
+
"grad_norm": 0.0049609216072808585,
|
| 1146 |
+
"kl": 0.38330078125,
|
| 1147 |
+
"learning_rate": 6.963231950007844e-07,
|
| 1148 |
+
"loss": 0.0004,
|
| 1149 |
+
"reward": 1.0,
|
| 1150 |
+
"reward_std": 0.0,
|
| 1151 |
+
"rewards/equation_reward_func": 0.0,
|
| 1152 |
+
"rewards/format_reward_func": 1.0,
|
| 1153 |
+
"step": 176
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"completion_length": 22.596354842185974,
|
| 1157 |
+
"epoch": 0.09493333333333333,
|
| 1158 |
+
"grad_norm": 0.0033493670757974436,
|
| 1159 |
+
"kl": 0.3634033203125,
|
| 1160 |
+
"learning_rate": 6.896762273384178e-07,
|
| 1161 |
+
"loss": 0.0004,
|
| 1162 |
+
"reward": 1.0,
|
| 1163 |
+
"reward_std": 0.0,
|
| 1164 |
+
"rewards/equation_reward_func": 0.0,
|
| 1165 |
+
"rewards/format_reward_func": 1.0,
|
| 1166 |
+
"step": 178
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"completion_length": 21.070313334465027,
|
| 1170 |
+
"epoch": 0.096,
|
| 1171 |
+
"grad_norm": 0.007358426048838388,
|
| 1172 |
+
"kl": 0.374267578125,
|
| 1173 |
+
"learning_rate": 6.829898691169579e-07,
|
| 1174 |
+
"loss": 0.0004,
|
| 1175 |
+
"reward": 1.0,
|
| 1176 |
+
"reward_std": 0.0,
|
| 1177 |
+
"rewards/equation_reward_func": 0.0,
|
| 1178 |
+
"rewards/format_reward_func": 1.0,
|
| 1179 |
+
"step": 180
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"completion_length": 25.106771528720856,
|
| 1183 |
+
"epoch": 0.09706666666666666,
|
| 1184 |
+
"grad_norm": 0.002673392643388894,
|
| 1185 |
+
"kl": 0.3927001953125,
|
| 1186 |
+
"learning_rate": 6.762655089099353e-07,
|
| 1187 |
+
"loss": 0.0004,
|
| 1188 |
+
"reward": 1.0,
|
| 1189 |
+
"reward_std": 0.0,
|
| 1190 |
+
"rewards/equation_reward_func": 0.0,
|
| 1191 |
+
"rewards/format_reward_func": 1.0,
|
| 1192 |
+
"step": 182
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"completion_length": 23.947917222976685,
|
| 1196 |
+
"epoch": 0.09813333333333334,
|
| 1197 |
+
"grad_norm": 0.009888137802143636,
|
| 1198 |
+
"kl": 0.4693603515625,
|
| 1199 |
+
"learning_rate": 6.695045431828524e-07,
|
| 1200 |
+
"loss": 0.0005,
|
| 1201 |
+
"reward": 1.0,
|
| 1202 |
+
"reward_std": 0.0,
|
| 1203 |
+
"rewards/equation_reward_func": 0.0,
|
| 1204 |
+
"rewards/format_reward_func": 1.0,
|
| 1205 |
+
"step": 184
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"completion_length": 20.893229484558105,
|
| 1209 |
+
"epoch": 0.0992,
|
| 1210 |
+
"grad_norm": 0.0022607662791007638,
|
| 1211 |
+
"kl": 0.3218994140625,
|
| 1212 |
+
"learning_rate": 6.627083760031754e-07,
|
| 1213 |
+
"loss": 0.0003,
|
| 1214 |
+
"reward": 1.0,
|
| 1215 |
+
"reward_std": 0.0,
|
| 1216 |
+
"rewards/equation_reward_func": 0.0,
|
| 1217 |
+
"rewards/format_reward_func": 1.0,
|
| 1218 |
+
"step": 186
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"completion_length": 19.361979722976685,
|
| 1222 |
+
"epoch": 0.10026666666666667,
|
| 1223 |
+
"grad_norm": 0.0032574710279835964,
|
| 1224 |
+
"kl": 0.3485107421875,
|
| 1225 |
+
"learning_rate": 6.558784187487494e-07,
|
| 1226 |
+
"loss": 0.0003,
|
| 1227 |
+
"reward": 1.0,
|
| 1228 |
+
"reward_std": 0.0,
|
| 1229 |
+
"rewards/equation_reward_func": 0.0,
|
| 1230 |
+
"rewards/format_reward_func": 1.0,
|
| 1231 |
+
"step": 188
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"completion_length": 20.19270896911621,
|
| 1235 |
+
"epoch": 0.10133333333333333,
|
| 1236 |
+
"grad_norm": 0.003981197482172737,
|
| 1237 |
+
"kl": 0.3470458984375,
|
| 1238 |
+
"learning_rate": 6.490160898146918e-07,
|
| 1239 |
+
"loss": 0.0003,
|
| 1240 |
+
"reward": 1.0,
|
| 1241 |
+
"reward_std": 0.0,
|
| 1242 |
+
"rewards/equation_reward_func": 0.0,
|
| 1243 |
+
"rewards/format_reward_func": 1.0,
|
| 1244 |
+
"step": 190
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"completion_length": 19.901042342185974,
|
| 1248 |
+
"epoch": 0.1024,
|
| 1249 |
+
"grad_norm": 0.002514384073907443,
|
| 1250 |
+
"kl": 0.3548583984375,
|
| 1251 |
+
"learning_rate": 6.421228143188324e-07,
|
| 1252 |
+
"loss": 0.0004,
|
| 1253 |
+
"reward": 1.0,
|
| 1254 |
+
"reward_std": 0.0,
|
| 1255 |
+
"rewards/equation_reward_func": 0.0,
|
| 1256 |
+
"rewards/format_reward_func": 1.0,
|
| 1257 |
+
"step": 192
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"completion_length": 24.843750834465027,
|
| 1261 |
+
"epoch": 0.10346666666666667,
|
| 1262 |
+
"grad_norm": 0.007842781472841063,
|
| 1263 |
+
"kl": 0.3717041015625,
|
| 1264 |
+
"learning_rate": 6.352000238057539e-07,
|
| 1265 |
+
"loss": 0.0004,
|
| 1266 |
+
"reward": 1.0,
|
| 1267 |
+
"reward_std": 0.0,
|
| 1268 |
+
"rewards/equation_reward_func": 0.0,
|
| 1269 |
+
"rewards/format_reward_func": 1.0,
|
| 1270 |
+
"step": 194
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"completion_length": 21.859375715255737,
|
| 1274 |
+
"epoch": 0.10453333333333334,
|
| 1275 |
+
"grad_norm": 0.0060397124228908665,
|
| 1276 |
+
"kl": 0.357666015625,
|
| 1277 |
+
"learning_rate": 6.282491559495004e-07,
|
| 1278 |
+
"loss": 0.0004,
|
| 1279 |
+
"reward": 1.0,
|
| 1280 |
+
"reward_std": 0.0,
|
| 1281 |
+
"rewards/equation_reward_func": 0.0,
|
| 1282 |
+
"rewards/format_reward_func": 1.0,
|
| 1283 |
+
"step": 196
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"completion_length": 22.700521767139435,
|
| 1287 |
+
"epoch": 0.1056,
|
| 1288 |
+
"grad_norm": 0.005604073044775416,
|
| 1289 |
+
"kl": 0.4561767578125,
|
| 1290 |
+
"learning_rate": 6.212716542550112e-07,
|
| 1291 |
+
"loss": 0.0005,
|
| 1292 |
+
"reward": 1.0,
|
| 1293 |
+
"reward_std": 0.0,
|
| 1294 |
+
"rewards/equation_reward_func": 0.0,
|
| 1295 |
+
"rewards/format_reward_func": 1.0,
|
| 1296 |
+
"step": 198
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"completion_length": 22.901042222976685,
|
| 1300 |
+
"epoch": 0.10666666666666667,
|
| 1301 |
+
"grad_norm": 0.0026562431855837636,
|
| 1302 |
+
"kl": 0.3526611328125,
|
| 1303 |
+
"learning_rate": 6.142689677583445e-07,
|
| 1304 |
+
"loss": 0.0004,
|
| 1305 |
+
"reward": 1.0,
|
| 1306 |
+
"reward_std": 0.0,
|
| 1307 |
+
"rewards/equation_reward_func": 0.0,
|
| 1308 |
+
"rewards/format_reward_func": 1.0,
|
| 1309 |
+
"step": 200
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"completion_length": 24.648437798023224,
|
| 1313 |
+
"epoch": 0.10773333333333333,
|
| 1314 |
+
"grad_norm": 0.002588639622294546,
|
| 1315 |
+
"kl": 0.3214111328125,
|
| 1316 |
+
"learning_rate": 6.072425507257527e-07,
|
| 1317 |
+
"loss": 0.0003,
|
| 1318 |
+
"reward": 1.0,
|
| 1319 |
+
"reward_std": 0.0,
|
| 1320 |
+
"rewards/equation_reward_func": 0.0,
|
| 1321 |
+
"rewards/format_reward_func": 1.0,
|
| 1322 |
+
"step": 202
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"completion_length": 22.382813036441803,
|
| 1326 |
+
"epoch": 0.1088,
|
| 1327 |
+
"grad_norm": 0.00323664281124744,
|
| 1328 |
+
"kl": 0.36767578125,
|
| 1329 |
+
"learning_rate": 6.001938623516705e-07,
|
| 1330 |
+
"loss": 0.0004,
|
| 1331 |
+
"reward": 1.0,
|
| 1332 |
+
"reward_std": 0.0,
|
| 1333 |
+
"rewards/equation_reward_func": 0.0,
|
| 1334 |
+
"rewards/format_reward_func": 1.0,
|
| 1335 |
+
"step": 204
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"completion_length": 23.174479961395264,
|
| 1339 |
+
"epoch": 0.10986666666666667,
|
| 1340 |
+
"grad_norm": 0.004194373433624107,
|
| 1341 |
+
"kl": 0.361572265625,
|
| 1342 |
+
"learning_rate": 5.931243664556802e-07,
|
| 1343 |
+
"loss": 0.0004,
|
| 1344 |
+
"reward": 1.0,
|
| 1345 |
+
"reward_std": 0.0,
|
| 1346 |
+
"rewards/equation_reward_func": 0.0,
|
| 1347 |
+
"rewards/format_reward_func": 1.0,
|
| 1348 |
+
"step": 206
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"completion_length": 22.494792342185974,
|
| 1352 |
+
"epoch": 0.11093333333333333,
|
| 1353 |
+
"grad_norm": 0.0635254172279242,
|
| 1354 |
+
"kl": 0.389404296875,
|
| 1355 |
+
"learning_rate": 5.860355311785175e-07,
|
| 1356 |
+
"loss": 0.0004,
|
| 1357 |
+
"reward": 0.9973958358168602,
|
| 1358 |
+
"reward_std": 0.007365695666521788,
|
| 1359 |
+
"rewards/equation_reward_func": 0.0,
|
| 1360 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1361 |
+
"step": 208
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"completion_length": 24.593750715255737,
|
| 1365 |
+
"epoch": 0.112,
|
| 1366 |
+
"grad_norm": 0.0040160843770763,
|
| 1367 |
+
"kl": 0.381103515625,
|
| 1368 |
+
"learning_rate": 5.78928828677177e-07,
|
| 1369 |
+
"loss": 0.0004,
|
| 1370 |
+
"reward": 1.0,
|
| 1371 |
+
"reward_std": 0.0,
|
| 1372 |
+
"rewards/equation_reward_func": 0.0,
|
| 1373 |
+
"rewards/format_reward_func": 1.0,
|
| 1374 |
+
"step": 210
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"completion_length": 19.640625715255737,
|
| 1378 |
+
"epoch": 0.11306666666666666,
|
| 1379 |
+
"grad_norm": 0.004056894485336028,
|
| 1380 |
+
"kl": 0.3604736328125,
|
| 1381 |
+
"learning_rate": 5.718057348191874e-07,
|
| 1382 |
+
"loss": 0.0004,
|
| 1383 |
+
"reward": 1.0,
|
| 1384 |
+
"reward_std": 0.0,
|
| 1385 |
+
"rewards/equation_reward_func": 0.0,
|
| 1386 |
+
"rewards/format_reward_func": 1.0,
|
| 1387 |
+
"step": 212
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"completion_length": 26.078125655651093,
|
| 1391 |
+
"epoch": 0.11413333333333334,
|
| 1392 |
+
"grad_norm": 0.001426529705853739,
|
| 1393 |
+
"kl": 0.3330078125,
|
| 1394 |
+
"learning_rate": 5.646677288761132e-07,
|
| 1395 |
+
"loss": 0.0003,
|
| 1396 |
+
"reward": 0.9973958358168602,
|
| 1397 |
+
"reward_std": 0.007365695666521788,
|
| 1398 |
+
"rewards/equation_reward_func": 0.0,
|
| 1399 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1400 |
+
"step": 214
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"completion_length": 18.17447954416275,
|
| 1404 |
+
"epoch": 0.1152,
|
| 1405 |
+
"grad_norm": 0.00410254063002437,
|
| 1406 |
+
"kl": 0.360107421875,
|
| 1407 |
+
"learning_rate": 5.575162932163501e-07,
|
| 1408 |
+
"loss": 0.0004,
|
| 1409 |
+
"reward": 1.0,
|
| 1410 |
+
"reward_std": 0.0,
|
| 1411 |
+
"rewards/equation_reward_func": 0.0,
|
| 1412 |
+
"rewards/format_reward_func": 1.0,
|
| 1413 |
+
"step": 216
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"completion_length": 19.231771647930145,
|
| 1417 |
+
"epoch": 0.11626666666666667,
|
| 1418 |
+
"grad_norm": 0.004454685182350467,
|
| 1419 |
+
"kl": 0.3358154296875,
|
| 1420 |
+
"learning_rate": 5.503529129972792e-07,
|
| 1421 |
+
"loss": 0.0003,
|
| 1422 |
+
"reward": 1.0,
|
| 1423 |
+
"reward_std": 0.0,
|
| 1424 |
+
"rewards/equation_reward_func": 0.0,
|
| 1425 |
+
"rewards/format_reward_func": 1.0,
|
| 1426 |
+
"step": 218
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"completion_length": 22.81250023841858,
|
| 1430 |
+
"epoch": 0.11733333333333333,
|
| 1431 |
+
"grad_norm": 0.0033331903371956913,
|
| 1432 |
+
"kl": 0.3734130859375,
|
| 1433 |
+
"learning_rate": 5.431790758568388e-07,
|
| 1434 |
+
"loss": 0.0004,
|
| 1435 |
+
"reward": 1.0,
|
| 1436 |
+
"reward_std": 0.0,
|
| 1437 |
+
"rewards/equation_reward_func": 0.0,
|
| 1438 |
+
"rewards/format_reward_func": 1.0,
|
| 1439 |
+
"step": 220
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"completion_length": 33.567709386348724,
|
| 1443 |
+
"epoch": 0.1184,
|
| 1444 |
+
"grad_norm": 0.15156820624659995,
|
| 1445 |
+
"kl": 0.3978271484375,
|
| 1446 |
+
"learning_rate": 5.359962716045835e-07,
|
| 1447 |
+
"loss": 0.0004,
|
| 1448 |
+
"reward": 0.9973958358168602,
|
| 1449 |
+
"reward_std": 0.007365695666521788,
|
| 1450 |
+
"rewards/equation_reward_func": 0.0,
|
| 1451 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1452 |
+
"step": 222
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"completion_length": 23.721354722976685,
|
| 1456 |
+
"epoch": 0.11946666666666667,
|
| 1457 |
+
"grad_norm": 0.1302812080137746,
|
| 1458 |
+
"kl": 0.3497314453125,
|
| 1459 |
+
"learning_rate": 5.288059919122921e-07,
|
| 1460 |
+
"loss": 0.0003,
|
| 1461 |
+
"reward": 0.9973958358168602,
|
| 1462 |
+
"reward_std": 0.007365695666521788,
|
| 1463 |
+
"rewards/equation_reward_func": 0.0,
|
| 1464 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1465 |
+
"step": 224
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"completion_length": 23.40104216337204,
|
| 1469 |
+
"epoch": 0.12053333333333334,
|
| 1470 |
+
"grad_norm": 0.10213989334337155,
|
| 1471 |
+
"kl": 0.3587646484375,
|
| 1472 |
+
"learning_rate": 5.216097300041869e-07,
|
| 1473 |
+
"loss": 0.0004,
|
| 1474 |
+
"reward": 0.9973958358168602,
|
| 1475 |
+
"reward_std": 0.007365695666521788,
|
| 1476 |
+
"rewards/equation_reward_func": 0.0,
|
| 1477 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1478 |
+
"step": 226
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"completion_length": 28.752604722976685,
|
| 1482 |
+
"epoch": 0.1216,
|
| 1483 |
+
"grad_norm": 0.00405575423941108,
|
| 1484 |
+
"kl": 0.356689453125,
|
| 1485 |
+
"learning_rate": 5.144089803468332e-07,
|
| 1486 |
+
"loss": 0.0004,
|
| 1487 |
+
"reward": 1.0,
|
| 1488 |
+
"reward_std": 0.0,
|
| 1489 |
+
"rewards/equation_reward_func": 0.0,
|
| 1490 |
+
"rewards/format_reward_func": 1.0,
|
| 1491 |
+
"step": 228
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"completion_length": 21.56770896911621,
|
| 1495 |
+
"epoch": 0.12266666666666666,
|
| 1496 |
+
"grad_norm": 0.006688124816156981,
|
| 1497 |
+
"kl": 0.38134765625,
|
| 1498 |
+
"learning_rate": 5.072052383387786e-07,
|
| 1499 |
+
"loss": 0.0004,
|
| 1500 |
+
"reward": 1.0,
|
| 1501 |
+
"reward_std": 0.0,
|
| 1502 |
+
"rewards/equation_reward_func": 0.0,
|
| 1503 |
+
"rewards/format_reward_func": 1.0,
|
| 1504 |
+
"step": 230
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"completion_length": 30.83333396911621,
|
| 1508 |
+
"epoch": 0.12373333333333333,
|
| 1509 |
+
"grad_norm": 0.003894854329196007,
|
| 1510 |
+
"kl": 0.36962890625,
|
| 1511 |
+
"learning_rate": 5e-07,
|
| 1512 |
+
"loss": 0.0004,
|
| 1513 |
+
"reward": 1.0,
|
| 1514 |
+
"reward_std": 0.0,
|
| 1515 |
+
"rewards/equation_reward_func": 0.0,
|
| 1516 |
+
"rewards/format_reward_func": 1.0,
|
| 1517 |
+
"step": 232
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"completion_length": 38.3385426402092,
|
| 1521 |
+
"epoch": 0.1248,
|
| 1522 |
+
"grad_norm": 0.003321341352419865,
|
| 1523 |
+
"kl": 0.339111328125,
|
| 1524 |
+
"learning_rate": 4.927947616612215e-07,
|
| 1525 |
+
"loss": 0.0003,
|
| 1526 |
+
"reward": 1.0,
|
| 1527 |
+
"reward_std": 0.0,
|
| 1528 |
+
"rewards/equation_reward_func": 0.0,
|
| 1529 |
+
"rewards/format_reward_func": 1.0,
|
| 1530 |
+
"step": 234
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"completion_length": 25.40104252099991,
|
| 1534 |
+
"epoch": 0.12586666666666665,
|
| 1535 |
+
"grad_norm": 0.003665106813183815,
|
| 1536 |
+
"kl": 0.347900390625,
|
| 1537 |
+
"learning_rate": 4.855910196531669e-07,
|
| 1538 |
+
"loss": 0.0003,
|
| 1539 |
+
"reward": 1.0,
|
| 1540 |
+
"reward_std": 0.0,
|
| 1541 |
+
"rewards/equation_reward_func": 0.0,
|
| 1542 |
+
"rewards/format_reward_func": 1.0,
|
| 1543 |
+
"step": 236
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"completion_length": 32.97656351327896,
|
| 1547 |
+
"epoch": 0.12693333333333334,
|
| 1548 |
+
"grad_norm": 0.0029113458934489096,
|
| 1549 |
+
"kl": 0.3333740234375,
|
| 1550 |
+
"learning_rate": 4.783902699958129e-07,
|
| 1551 |
+
"loss": 0.0003,
|
| 1552 |
+
"reward": 1.0,
|
| 1553 |
+
"reward_std": 0.0,
|
| 1554 |
+
"rewards/equation_reward_func": 0.0,
|
| 1555 |
+
"rewards/format_reward_func": 1.0,
|
| 1556 |
+
"step": 238
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"completion_length": 32.789063453674316,
|
| 1560 |
+
"epoch": 0.128,
|
| 1561 |
+
"grad_norm": 0.2144030619978656,
|
| 1562 |
+
"kl": 0.98974609375,
|
| 1563 |
+
"learning_rate": 4.711940080877079e-07,
|
| 1564 |
+
"loss": 0.001,
|
| 1565 |
+
"reward": 1.0,
|
| 1566 |
+
"reward_std": 0.0,
|
| 1567 |
+
"rewards/equation_reward_func": 0.0,
|
| 1568 |
+
"rewards/format_reward_func": 1.0,
|
| 1569 |
+
"step": 240
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"completion_length": 27.617188334465027,
|
| 1573 |
+
"epoch": 0.12906666666666666,
|
| 1574 |
+
"grad_norm": 0.004925288676087789,
|
| 1575 |
+
"kl": 0.355712890625,
|
| 1576 |
+
"learning_rate": 4.6400372839541647e-07,
|
| 1577 |
+
"loss": 0.0004,
|
| 1578 |
+
"reward": 1.0,
|
| 1579 |
+
"reward_std": 0.0,
|
| 1580 |
+
"rewards/equation_reward_func": 0.0,
|
| 1581 |
+
"rewards/format_reward_func": 1.0,
|
| 1582 |
+
"step": 242
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"completion_length": 27.859375417232513,
|
| 1586 |
+
"epoch": 0.13013333333333332,
|
| 1587 |
+
"grad_norm": 0.004078504522837724,
|
| 1588 |
+
"kl": 0.3455810546875,
|
| 1589 |
+
"learning_rate": 4.568209241431614e-07,
|
| 1590 |
+
"loss": 0.0003,
|
| 1591 |
+
"reward": 1.0,
|
| 1592 |
+
"reward_std": 0.0,
|
| 1593 |
+
"rewards/equation_reward_func": 0.0,
|
| 1594 |
+
"rewards/format_reward_func": 1.0,
|
| 1595 |
+
"step": 244
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"completion_length": 38.55729287862778,
|
| 1599 |
+
"epoch": 0.1312,
|
| 1600 |
+
"grad_norm": 0.003946730384617455,
|
| 1601 |
+
"kl": 0.3524169921875,
|
| 1602 |
+
"learning_rate": 4.4964708700272086e-07,
|
| 1603 |
+
"loss": 0.0004,
|
| 1604 |
+
"reward": 1.0,
|
| 1605 |
+
"reward_std": 0.0,
|
| 1606 |
+
"rewards/equation_reward_func": 0.0,
|
| 1607 |
+
"rewards/format_reward_func": 1.0,
|
| 1608 |
+
"step": 246
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"completion_length": 36.60937589406967,
|
| 1612 |
+
"epoch": 0.13226666666666667,
|
| 1613 |
+
"grad_norm": 0.0025911319683286625,
|
| 1614 |
+
"kl": 0.3311767578125,
|
| 1615 |
+
"learning_rate": 4.424837067836499e-07,
|
| 1616 |
+
"loss": 0.0003,
|
| 1617 |
+
"reward": 1.0,
|
| 1618 |
+
"reward_std": 0.0,
|
| 1619 |
+
"rewards/equation_reward_func": 0.0,
|
| 1620 |
+
"rewards/format_reward_func": 1.0,
|
| 1621 |
+
"step": 248
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"completion_length": 34.45312637090683,
|
| 1625 |
+
"epoch": 0.13333333333333333,
|
| 1626 |
+
"grad_norm": 0.02082559848620119,
|
| 1627 |
+
"kl": 0.3614501953125,
|
| 1628 |
+
"learning_rate": 4.353322711238869e-07,
|
| 1629 |
+
"loss": 0.0004,
|
| 1630 |
+
"reward": 1.0,
|
| 1631 |
+
"reward_std": 0.0,
|
| 1632 |
+
"rewards/equation_reward_func": 0.0,
|
| 1633 |
+
"rewards/format_reward_func": 1.0,
|
| 1634 |
+
"step": 250
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"completion_length": 44.54427218437195,
|
| 1638 |
+
"epoch": 0.1344,
|
| 1639 |
+
"grad_norm": 0.004583784786433975,
|
| 1640 |
+
"kl": 0.3564453125,
|
| 1641 |
+
"learning_rate": 4.2819426518081256e-07,
|
| 1642 |
+
"loss": 0.0004,
|
| 1643 |
+
"reward": 0.9973958358168602,
|
| 1644 |
+
"reward_std": 0.007365695666521788,
|
| 1645 |
+
"rewards/equation_reward_func": 0.0,
|
| 1646 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1647 |
+
"step": 252
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"completion_length": 32.47656333446503,
|
| 1651 |
+
"epoch": 0.13546666666666668,
|
| 1652 |
+
"grad_norm": 0.2516579496285295,
|
| 1653 |
+
"kl": 0.348388671875,
|
| 1654 |
+
"learning_rate": 4.21071171322823e-07,
|
| 1655 |
+
"loss": 0.0003,
|
| 1656 |
+
"reward": 0.9973958358168602,
|
| 1657 |
+
"reward_std": 0.007365695666521788,
|
| 1658 |
+
"rewards/equation_reward_func": 0.0,
|
| 1659 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1660 |
+
"step": 254
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"completion_length": 33.247396886348724,
|
| 1664 |
+
"epoch": 0.13653333333333334,
|
| 1665 |
+
"grad_norm": 0.003750844211497512,
|
| 1666 |
+
"kl": 0.336669921875,
|
| 1667 |
+
"learning_rate": 4.139644688214826e-07,
|
| 1668 |
+
"loss": 0.0003,
|
| 1669 |
+
"reward": 1.0,
|
| 1670 |
+
"reward_std": 0.0,
|
| 1671 |
+
"rewards/equation_reward_func": 0.0,
|
| 1672 |
+
"rewards/format_reward_func": 1.0,
|
| 1673 |
+
"step": 256
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"completion_length": 40.09375071525574,
|
| 1677 |
+
"epoch": 0.1376,
|
| 1678 |
+
"grad_norm": 0.0056285121180003365,
|
| 1679 |
+
"kl": 0.3736572265625,
|
| 1680 |
+
"learning_rate": 4.068756335443198e-07,
|
| 1681 |
+
"loss": 0.0004,
|
| 1682 |
+
"reward": 1.0,
|
| 1683 |
+
"reward_std": 0.0,
|
| 1684 |
+
"rewards/equation_reward_func": 0.0,
|
| 1685 |
+
"rewards/format_reward_func": 1.0,
|
| 1686 |
+
"step": 258
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"completion_length": 32.55468821525574,
|
| 1690 |
+
"epoch": 0.13866666666666666,
|
| 1691 |
+
"grad_norm": 0.19004745287506442,
|
| 1692 |
+
"kl": 0.3536376953125,
|
| 1693 |
+
"learning_rate": 3.998061376483297e-07,
|
| 1694 |
+
"loss": 0.0004,
|
| 1695 |
+
"reward": 0.9947916716337204,
|
| 1696 |
+
"reward_std": 0.014731391333043575,
|
| 1697 |
+
"rewards/equation_reward_func": 0.0,
|
| 1698 |
+
"rewards/format_reward_func": 0.9947916716337204,
|
| 1699 |
+
"step": 260
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"completion_length": 41.53125083446503,
|
| 1703 |
+
"epoch": 0.13973333333333332,
|
| 1704 |
+
"grad_norm": 0.005613467195859889,
|
| 1705 |
+
"kl": 0.3878173828125,
|
| 1706 |
+
"learning_rate": 3.9275744927424723e-07,
|
| 1707 |
+
"loss": 0.0004,
|
| 1708 |
+
"reward": 0.9973958358168602,
|
| 1709 |
+
"reward_std": 0.007365695666521788,
|
| 1710 |
+
"rewards/equation_reward_func": 0.0,
|
| 1711 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1712 |
+
"step": 262
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"completion_length": 41.14323008060455,
|
| 1716 |
+
"epoch": 0.1408,
|
| 1717 |
+
"grad_norm": 0.004079597325625193,
|
| 1718 |
+
"kl": 0.341552734375,
|
| 1719 |
+
"learning_rate": 3.8573103224165547e-07,
|
| 1720 |
+
"loss": 0.0003,
|
| 1721 |
+
"reward": 0.9973958358168602,
|
| 1722 |
+
"reward_std": 0.007365695666521788,
|
| 1723 |
+
"rewards/equation_reward_func": 0.0,
|
| 1724 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1725 |
+
"step": 264
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"completion_length": 45.320313930511475,
|
| 1729 |
+
"epoch": 0.14186666666666667,
|
| 1730 |
+
"grad_norm": 0.004450670587730382,
|
| 1731 |
+
"kl": 0.3409423828125,
|
| 1732 |
+
"learning_rate": 3.787283457449889e-07,
|
| 1733 |
+
"loss": 0.0003,
|
| 1734 |
+
"reward": 1.0,
|
| 1735 |
+
"reward_std": 0.0,
|
| 1736 |
+
"rewards/equation_reward_func": 0.0,
|
| 1737 |
+
"rewards/format_reward_func": 1.0,
|
| 1738 |
+
"step": 266
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"completion_length": 34.77083444595337,
|
| 1742 |
+
"epoch": 0.14293333333333333,
|
| 1743 |
+
"grad_norm": 0.005059899644254497,
|
| 1744 |
+
"kl": 0.3505859375,
|
| 1745 |
+
"learning_rate": 3.717508440504997e-07,
|
| 1746 |
+
"loss": 0.0004,
|
| 1747 |
+
"reward": 1.0,
|
| 1748 |
+
"reward_std": 0.0,
|
| 1749 |
+
"rewards/equation_reward_func": 0.0,
|
| 1750 |
+
"rewards/format_reward_func": 1.0,
|
| 1751 |
+
"step": 268
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"completion_length": 34.54948043823242,
|
| 1755 |
+
"epoch": 0.144,
|
| 1756 |
+
"grad_norm": 0.23991193127694865,
|
| 1757 |
+
"kl": 0.3560791015625,
|
| 1758 |
+
"learning_rate": 3.64799976194246e-07,
|
| 1759 |
+
"loss": 0.0004,
|
| 1760 |
+
"reward": 0.9973958358168602,
|
| 1761 |
+
"reward_std": 0.007365695666521788,
|
| 1762 |
+
"rewards/equation_reward_func": 0.0,
|
| 1763 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1764 |
+
"step": 270
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"completion_length": 25.007812976837158,
|
| 1768 |
+
"epoch": 0.14506666666666668,
|
| 1769 |
+
"grad_norm": 0.004586704746066067,
|
| 1770 |
+
"kl": 0.342529296875,
|
| 1771 |
+
"learning_rate": 3.5787718568116757e-07,
|
| 1772 |
+
"loss": 0.0003,
|
| 1773 |
+
"reward": 1.0,
|
| 1774 |
+
"reward_std": 0.0,
|
| 1775 |
+
"rewards/equation_reward_func": 0.0,
|
| 1776 |
+
"rewards/format_reward_func": 1.0,
|
| 1777 |
+
"step": 272
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"completion_length": 33.093751192092896,
|
| 1781 |
+
"epoch": 0.14613333333333334,
|
| 1782 |
+
"grad_norm": 0.07458441934892074,
|
| 1783 |
+
"kl": 0.3690185546875,
|
| 1784 |
+
"learning_rate": 3.5098391018530813e-07,
|
| 1785 |
+
"loss": 0.0004,
|
| 1786 |
+
"reward": 0.9973958358168602,
|
| 1787 |
+
"reward_std": 0.007365695666521788,
|
| 1788 |
+
"rewards/equation_reward_func": 0.0,
|
| 1789 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1790 |
+
"step": 274
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"completion_length": 26.036458909511566,
|
| 1794 |
+
"epoch": 0.1472,
|
| 1795 |
+
"grad_norm": 0.0018656461831637952,
|
| 1796 |
+
"kl": 0.3399658203125,
|
| 1797 |
+
"learning_rate": 3.4412158125125073e-07,
|
| 1798 |
+
"loss": 0.0003,
|
| 1799 |
+
"reward": 1.0,
|
| 1800 |
+
"reward_std": 0.0,
|
| 1801 |
+
"rewards/equation_reward_func": 0.0,
|
| 1802 |
+
"rewards/format_reward_func": 1.0,
|
| 1803 |
+
"step": 276
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"completion_length": 27.08854240179062,
|
| 1807 |
+
"epoch": 0.14826666666666666,
|
| 1808 |
+
"grad_norm": 0.008897084345492313,
|
| 1809 |
+
"kl": 0.3919677734375,
|
| 1810 |
+
"learning_rate": 3.372916239968245e-07,
|
| 1811 |
+
"loss": 0.0004,
|
| 1812 |
+
"reward": 1.0,
|
| 1813 |
+
"reward_std": 0.0,
|
| 1814 |
+
"rewards/equation_reward_func": 0.0,
|
| 1815 |
+
"rewards/format_reward_func": 1.0,
|
| 1816 |
+
"step": 278
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"completion_length": 29.460938274860382,
|
| 1820 |
+
"epoch": 0.14933333333333335,
|
| 1821 |
+
"grad_norm": 0.0038377037869509346,
|
| 1822 |
+
"kl": 0.3658447265625,
|
| 1823 |
+
"learning_rate": 3.3049545681714775e-07,
|
| 1824 |
+
"loss": 0.0004,
|
| 1825 |
+
"reward": 1.0,
|
| 1826 |
+
"reward_std": 0.0,
|
| 1827 |
+
"rewards/equation_reward_func": 0.0,
|
| 1828 |
+
"rewards/format_reward_func": 1.0,
|
| 1829 |
+
"step": 280
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"completion_length": 20.976563274860382,
|
| 1833 |
+
"epoch": 0.1504,
|
| 1834 |
+
"grad_norm": 0.005935196775123734,
|
| 1835 |
+
"kl": 0.365234375,
|
| 1836 |
+
"learning_rate": 3.2373449109006474e-07,
|
| 1837 |
+
"loss": 0.0004,
|
| 1838 |
+
"reward": 0.9973958358168602,
|
| 1839 |
+
"reward_std": 0.007365695666521788,
|
| 1840 |
+
"rewards/equation_reward_func": 0.0,
|
| 1841 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1842 |
+
"step": 282
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"completion_length": 28.750000536441803,
|
| 1846 |
+
"epoch": 0.15146666666666667,
|
| 1847 |
+
"grad_norm": 0.006185835967833735,
|
| 1848 |
+
"kl": 0.3516845703125,
|
| 1849 |
+
"learning_rate": 3.1701013088304206e-07,
|
| 1850 |
+
"loss": 0.0004,
|
| 1851 |
+
"reward": 0.9973958358168602,
|
| 1852 |
+
"reward_std": 0.007365695666521788,
|
| 1853 |
+
"rewards/equation_reward_func": 0.0,
|
| 1854 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1855 |
+
"step": 284
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"completion_length": 33.2838551402092,
|
| 1859 |
+
"epoch": 0.15253333333333333,
|
| 1860 |
+
"grad_norm": 0.004000349539922501,
|
| 1861 |
+
"kl": 0.3592529296875,
|
| 1862 |
+
"learning_rate": 3.1032377266158214e-07,
|
| 1863 |
+
"loss": 0.0004,
|
| 1864 |
+
"reward": 1.0,
|
| 1865 |
+
"reward_std": 0.0,
|
| 1866 |
+
"rewards/equation_reward_func": 0.0,
|
| 1867 |
+
"rewards/format_reward_func": 1.0,
|
| 1868 |
+
"step": 286
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"completion_length": 25.153646647930145,
|
| 1872 |
+
"epoch": 0.1536,
|
| 1873 |
+
"grad_norm": 0.006473305213296596,
|
| 1874 |
+
"kl": 0.3848876953125,
|
| 1875 |
+
"learning_rate": 3.036768049992157e-07,
|
| 1876 |
+
"loss": 0.0004,
|
| 1877 |
+
"reward": 1.0,
|
| 1878 |
+
"reward_std": 0.0,
|
| 1879 |
+
"rewards/equation_reward_func": 0.0,
|
| 1880 |
+
"rewards/format_reward_func": 1.0,
|
| 1881 |
+
"step": 288
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"completion_length": 29.770834028720856,
|
| 1885 |
+
"epoch": 0.15466666666666667,
|
| 1886 |
+
"grad_norm": 0.003973211467908967,
|
| 1887 |
+
"kl": 0.3634033203125,
|
| 1888 |
+
"learning_rate": 2.9707060828913224e-07,
|
| 1889 |
+
"loss": 0.0004,
|
| 1890 |
+
"reward": 1.0,
|
| 1891 |
+
"reward_std": 0.0,
|
| 1892 |
+
"rewards/equation_reward_func": 0.0,
|
| 1893 |
+
"rewards/format_reward_func": 1.0,
|
| 1894 |
+
"step": 290
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"completion_length": 22.49479216337204,
|
| 1898 |
+
"epoch": 0.15573333333333333,
|
| 1899 |
+
"grad_norm": 0.00485962798897511,
|
| 1900 |
+
"kl": 0.3812255859375,
|
| 1901 |
+
"learning_rate": 2.9050655445751137e-07,
|
| 1902 |
+
"loss": 0.0004,
|
| 1903 |
+
"reward": 1.0,
|
| 1904 |
+
"reward_std": 0.0,
|
| 1905 |
+
"rewards/equation_reward_func": 0.0,
|
| 1906 |
+
"rewards/format_reward_func": 1.0,
|
| 1907 |
+
"step": 292
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"completion_length": 20.432292222976685,
|
| 1911 |
+
"epoch": 0.1568,
|
| 1912 |
+
"grad_norm": 0.0763280002987439,
|
| 1913 |
+
"kl": 0.359619140625,
|
| 1914 |
+
"learning_rate": 2.839860066786103e-07,
|
| 1915 |
+
"loss": 0.0004,
|
| 1916 |
+
"reward": 0.9973958358168602,
|
| 1917 |
+
"reward_std": 0.007365695666521788,
|
| 1918 |
+
"rewards/equation_reward_func": 0.0,
|
| 1919 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 1920 |
+
"step": 294
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"completion_length": 22.606771171092987,
|
| 1924 |
+
"epoch": 0.15786666666666666,
|
| 1925 |
+
"grad_norm": 0.007131272098647573,
|
| 1926 |
+
"kl": 0.407958984375,
|
| 1927 |
+
"learning_rate": 2.7751031909167045e-07,
|
| 1928 |
+
"loss": 0.0004,
|
| 1929 |
+
"reward": 1.0,
|
| 1930 |
+
"reward_std": 0.0,
|
| 1931 |
+
"rewards/equation_reward_func": 0.0,
|
| 1932 |
+
"rewards/format_reward_func": 1.0,
|
| 1933 |
+
"step": 296
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"completion_length": 26.085938036441803,
|
| 1937 |
+
"epoch": 0.15893333333333334,
|
| 1938 |
+
"grad_norm": 0.006348176701408371,
|
| 1939 |
+
"kl": 0.374755859375,
|
| 1940 |
+
"learning_rate": 2.710808365197e-07,
|
| 1941 |
+
"loss": 0.0004,
|
| 1942 |
+
"reward": 1.0,
|
| 1943 |
+
"reward_std": 0.0,
|
| 1944 |
+
"rewards/equation_reward_func": 0.0,
|
| 1945 |
+
"rewards/format_reward_func": 1.0,
|
| 1946 |
+
"step": 298
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"completion_length": 28.302084028720856,
|
| 1950 |
+
"epoch": 0.16,
|
| 1951 |
+
"grad_norm": 0.00329994798018086,
|
| 1952 |
+
"kl": 0.357421875,
|
| 1953 |
+
"learning_rate": 2.646988941901898e-07,
|
| 1954 |
+
"loss": 0.0004,
|
| 1955 |
+
"reward": 1.0,
|
| 1956 |
+
"reward_std": 0.0,
|
| 1957 |
+
"rewards/equation_reward_func": 0.0,
|
| 1958 |
+
"rewards/format_reward_func": 1.0,
|
| 1959 |
+
"step": 300
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"completion_length": 24.33333396911621,
|
| 1963 |
+
"epoch": 0.16106666666666666,
|
| 1964 |
+
"grad_norm": 0.0041357777851463405,
|
| 1965 |
+
"kl": 0.3416748046875,
|
| 1966 |
+
"learning_rate": 2.583658174578247e-07,
|
| 1967 |
+
"loss": 0.0003,
|
| 1968 |
+
"reward": 1.0,
|
| 1969 |
+
"reward_std": 0.0,
|
| 1970 |
+
"rewards/equation_reward_func": 0.0,
|
| 1971 |
+
"rewards/format_reward_func": 1.0,
|
| 1972 |
+
"step": 302
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"completion_length": 21.351562976837158,
|
| 1976 |
+
"epoch": 0.16213333333333332,
|
| 1977 |
+
"grad_norm": 0.004744062593804727,
|
| 1978 |
+
"kl": 0.3682861328125,
|
| 1979 |
+
"learning_rate": 2.520829215292426e-07,
|
| 1980 |
+
"loss": 0.0004,
|
| 1981 |
+
"reward": 1.0,
|
| 1982 |
+
"reward_std": 0.0,
|
| 1983 |
+
"rewards/equation_reward_func": 0.0,
|
| 1984 |
+
"rewards/format_reward_func": 1.0,
|
| 1985 |
+
"step": 304
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"completion_length": 17.56510454416275,
|
| 1989 |
+
"epoch": 0.1632,
|
| 1990 |
+
"grad_norm": 0.004071588499335152,
|
| 1991 |
+
"kl": 0.3665771484375,
|
| 1992 |
+
"learning_rate": 2.4585151118990285e-07,
|
| 1993 |
+
"loss": 0.0004,
|
| 1994 |
+
"reward": 1.0,
|
| 1995 |
+
"reward_std": 0.0,
|
| 1996 |
+
"rewards/equation_reward_func": 0.0,
|
| 1997 |
+
"rewards/format_reward_func": 1.0,
|
| 1998 |
+
"step": 306
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"completion_length": 25.093750953674316,
|
| 2002 |
+
"epoch": 0.16426666666666667,
|
| 2003 |
+
"grad_norm": 0.003447178186189921,
|
| 2004 |
+
"kl": 0.3665771484375,
|
| 2005 |
+
"learning_rate": 2.396728805331167e-07,
|
| 2006 |
+
"loss": 0.0004,
|
| 2007 |
+
"reward": 1.0,
|
| 2008 |
+
"reward_std": 0.0,
|
| 2009 |
+
"rewards/equation_reward_func": 0.0,
|
| 2010 |
+
"rewards/format_reward_func": 1.0,
|
| 2011 |
+
"step": 308
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"completion_length": 39.61198019981384,
|
| 2015 |
+
"epoch": 0.16533333333333333,
|
| 2016 |
+
"grad_norm": 0.15778923141355317,
|
| 2017 |
+
"kl": 0.3824462890625,
|
| 2018 |
+
"learning_rate": 2.3354831269130132e-07,
|
| 2019 |
+
"loss": 0.0004,
|
| 2020 |
+
"reward": 0.9947916716337204,
|
| 2021 |
+
"reward_std": 0.014731391333043575,
|
| 2022 |
+
"rewards/equation_reward_func": 0.0,
|
| 2023 |
+
"rewards/format_reward_func": 0.9947916716337204,
|
| 2024 |
+
"step": 310
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"completion_length": 20.757813096046448,
|
| 2028 |
+
"epoch": 0.1664,
|
| 2029 |
+
"grad_norm": 0.006322854264740659,
|
| 2030 |
+
"kl": 0.3919677734375,
|
| 2031 |
+
"learning_rate": 2.2747907956950707e-07,
|
| 2032 |
+
"loss": 0.0004,
|
| 2033 |
+
"reward": 1.0,
|
| 2034 |
+
"reward_std": 0.0,
|
| 2035 |
+
"rewards/equation_reward_func": 0.0,
|
| 2036 |
+
"rewards/format_reward_func": 1.0,
|
| 2037 |
+
"step": 312
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"completion_length": 18.778646290302277,
|
| 2041 |
+
"epoch": 0.16746666666666668,
|
| 2042 |
+
"grad_norm": 0.004277715160452868,
|
| 2043 |
+
"kl": 0.3692626953125,
|
| 2044 |
+
"learning_rate": 2.2146644158127826e-07,
|
| 2045 |
+
"loss": 0.0004,
|
| 2046 |
+
"reward": 1.0,
|
| 2047 |
+
"reward_std": 0.0,
|
| 2048 |
+
"rewards/equation_reward_func": 0.0,
|
| 2049 |
+
"rewards/format_reward_func": 1.0,
|
| 2050 |
+
"step": 314
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"completion_length": 22.914063036441803,
|
| 2054 |
+
"epoch": 0.16853333333333334,
|
| 2055 |
+
"grad_norm": 0.005925839314806503,
|
| 2056 |
+
"kl": 0.3868408203125,
|
| 2057 |
+
"learning_rate": 2.1551164738689892e-07,
|
| 2058 |
+
"loss": 0.0004,
|
| 2059 |
+
"reward": 1.0,
|
| 2060 |
+
"reward_std": 0.0,
|
| 2061 |
+
"rewards/equation_reward_func": 0.0,
|
| 2062 |
+
"rewards/format_reward_func": 1.0,
|
| 2063 |
+
"step": 316
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"completion_length": 19.528646647930145,
|
| 2067 |
+
"epoch": 0.1696,
|
| 2068 |
+
"grad_norm": 0.004034463075140326,
|
| 2069 |
+
"kl": 0.3629150390625,
|
| 2070 |
+
"learning_rate": 2.0961593363408154e-07,
|
| 2071 |
+
"loss": 0.0004,
|
| 2072 |
+
"reward": 1.0,
|
| 2073 |
+
"reward_std": 0.0,
|
| 2074 |
+
"rewards/equation_reward_func": 0.0,
|
| 2075 |
+
"rewards/format_reward_func": 1.0,
|
| 2076 |
+
"step": 318
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"completion_length": 26.062500655651093,
|
| 2080 |
+
"epoch": 0.17066666666666666,
|
| 2081 |
+
"grad_norm": 0.003874147172074953,
|
| 2082 |
+
"kl": 0.3551025390625,
|
| 2083 |
+
"learning_rate": 2.037805247011482e-07,
|
| 2084 |
+
"loss": 0.0004,
|
| 2085 |
+
"reward": 1.0,
|
| 2086 |
+
"reward_std": 0.0,
|
| 2087 |
+
"rewards/equation_reward_func": 0.0,
|
| 2088 |
+
"rewards/format_reward_func": 1.0,
|
| 2089 |
+
"step": 320
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"completion_length": 22.450521290302277,
|
| 2093 |
+
"epoch": 0.17173333333333332,
|
| 2094 |
+
"grad_norm": 0.0036545323789303995,
|
| 2095 |
+
"kl": 0.3665771484375,
|
| 2096 |
+
"learning_rate": 1.9800663244276127e-07,
|
| 2097 |
+
"loss": 0.0004,
|
| 2098 |
+
"reward": 1.0,
|
| 2099 |
+
"reward_std": 0.0,
|
| 2100 |
+
"rewards/equation_reward_func": 0.0,
|
| 2101 |
+
"rewards/format_reward_func": 1.0,
|
| 2102 |
+
"step": 322
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"completion_length": 33.1458340883255,
|
| 2106 |
+
"epoch": 0.1728,
|
| 2107 |
+
"grad_norm": 0.006124961025482053,
|
| 2108 |
+
"kl": 0.358642578125,
|
| 2109 |
+
"learning_rate": 1.9229545593825363e-07,
|
| 2110 |
+
"loss": 0.0004,
|
| 2111 |
+
"reward": 0.9973958358168602,
|
| 2112 |
+
"reward_std": 0.007365695666521788,
|
| 2113 |
+
"rewards/equation_reward_func": 0.0,
|
| 2114 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 2115 |
+
"step": 324
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"completion_length": 25.63541728258133,
|
| 2119 |
+
"epoch": 0.17386666666666667,
|
| 2120 |
+
"grad_norm": 0.004330409449251065,
|
| 2121 |
+
"kl": 0.38134765625,
|
| 2122 |
+
"learning_rate": 1.8664818124261373e-07,
|
| 2123 |
+
"loss": 0.0004,
|
| 2124 |
+
"reward": 1.0,
|
| 2125 |
+
"reward_std": 0.0,
|
| 2126 |
+
"rewards/equation_reward_func": 0.0,
|
| 2127 |
+
"rewards/format_reward_func": 1.0,
|
| 2128 |
+
"step": 326
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"completion_length": 30.52343863248825,
|
| 2132 |
+
"epoch": 0.17493333333333333,
|
| 2133 |
+
"grad_norm": 0.004429155190479566,
|
| 2134 |
+
"kl": 0.3779296875,
|
| 2135 |
+
"learning_rate": 1.8106598114017397e-07,
|
| 2136 |
+
"loss": 0.0004,
|
| 2137 |
+
"reward": 1.0,
|
| 2138 |
+
"reward_std": 0.0,
|
| 2139 |
+
"rewards/equation_reward_func": 0.0,
|
| 2140 |
+
"rewards/format_reward_func": 1.0,
|
| 2141 |
+
"step": 328
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"completion_length": 29.343750715255737,
|
| 2145 |
+
"epoch": 0.176,
|
| 2146 |
+
"grad_norm": 0.004794753409205765,
|
| 2147 |
+
"kl": 0.361083984375,
|
| 2148 |
+
"learning_rate": 1.7555001490105486e-07,
|
| 2149 |
+
"loss": 0.0004,
|
| 2150 |
+
"reward": 1.0,
|
| 2151 |
+
"reward_std": 0.0,
|
| 2152 |
+
"rewards/equation_reward_func": 0.0,
|
| 2153 |
+
"rewards/format_reward_func": 1.0,
|
| 2154 |
+
"step": 330
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"completion_length": 28.692708909511566,
|
| 2158 |
+
"epoch": 0.17706666666666668,
|
| 2159 |
+
"grad_norm": 0.005714723884457156,
|
| 2160 |
+
"kl": 0.3719482421875,
|
| 2161 |
+
"learning_rate": 1.7010142804041783e-07,
|
| 2162 |
+
"loss": 0.0004,
|
| 2163 |
+
"reward": 1.0,
|
| 2164 |
+
"reward_std": 0.0,
|
| 2165 |
+
"rewards/equation_reward_func": 0.0,
|
| 2166 |
+
"rewards/format_reward_func": 1.0,
|
| 2167 |
+
"step": 332
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"completion_length": 24.77864646911621,
|
| 2171 |
+
"epoch": 0.17813333333333334,
|
| 2172 |
+
"grad_norm": 0.004745602217539158,
|
| 2173 |
+
"kl": 0.35888671875,
|
| 2174 |
+
"learning_rate": 1.6472135208057125e-07,
|
| 2175 |
+
"loss": 0.0004,
|
| 2176 |
+
"reward": 1.0,
|
| 2177 |
+
"reward_std": 0.0,
|
| 2178 |
+
"rewards/equation_reward_func": 0.0,
|
| 2179 |
+
"rewards/format_reward_func": 1.0,
|
| 2180 |
+
"step": 334
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"completion_length": 30.609375655651093,
|
| 2184 |
+
"epoch": 0.1792,
|
| 2185 |
+
"grad_norm": 0.004920986126883945,
|
| 2186 |
+
"kl": 0.380126953125,
|
| 2187 |
+
"learning_rate": 1.5941090431598653e-07,
|
| 2188 |
+
"loss": 0.0004,
|
| 2189 |
+
"reward": 1.0,
|
| 2190 |
+
"reward_std": 0.0,
|
| 2191 |
+
"rewards/equation_reward_func": 0.0,
|
| 2192 |
+
"rewards/format_reward_func": 1.0,
|
| 2193 |
+
"step": 336
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"completion_length": 32.19010490179062,
|
| 2197 |
+
"epoch": 0.18026666666666666,
|
| 2198 |
+
"grad_norm": 0.004022967836188304,
|
| 2199 |
+
"kl": 0.3558349609375,
|
| 2200 |
+
"learning_rate": 1.5417118758126408e-07,
|
| 2201 |
+
"loss": 0.0004,
|
| 2202 |
+
"reward": 1.0,
|
| 2203 |
+
"reward_std": 0.0,
|
| 2204 |
+
"rewards/equation_reward_func": 0.0,
|
| 2205 |
+
"rewards/format_reward_func": 1.0,
|
| 2206 |
+
"step": 338
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"completion_length": 28.093750953674316,
|
| 2210 |
+
"epoch": 0.18133333333333335,
|
| 2211 |
+
"grad_norm": 0.003832900467363359,
|
| 2212 |
+
"kl": 0.3629150390625,
|
| 2213 |
+
"learning_rate": 1.490032900221068e-07,
|
| 2214 |
+
"loss": 0.0004,
|
| 2215 |
+
"reward": 1.0,
|
| 2216 |
+
"reward_std": 0.0,
|
| 2217 |
+
"rewards/equation_reward_func": 0.0,
|
| 2218 |
+
"rewards/format_reward_func": 1.0,
|
| 2219 |
+
"step": 340
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"completion_length": 28.463542699813843,
|
| 2223 |
+
"epoch": 0.1824,
|
| 2224 |
+
"grad_norm": 0.005487525138346315,
|
| 2225 |
+
"kl": 0.3526611328125,
|
| 2226 |
+
"learning_rate": 1.4390828486934058e-07,
|
| 2227 |
+
"loss": 0.0004,
|
| 2228 |
+
"reward": 1.0,
|
| 2229 |
+
"reward_std": 0.0,
|
| 2230 |
+
"rewards/equation_reward_func": 0.0,
|
| 2231 |
+
"rewards/format_reward_func": 1.0,
|
| 2232 |
+
"step": 342
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"completion_length": 24.825521528720856,
|
| 2236 |
+
"epoch": 0.18346666666666667,
|
| 2237 |
+
"grad_norm": 0.0031850458946458297,
|
| 2238 |
+
"kl": 0.36962890625,
|
| 2239 |
+
"learning_rate": 1.3888723021603526e-07,
|
| 2240 |
+
"loss": 0.0004,
|
| 2241 |
+
"reward": 1.0,
|
| 2242 |
+
"reward_std": 0.0,
|
| 2243 |
+
"rewards/equation_reward_func": 0.0,
|
| 2244 |
+
"rewards/format_reward_func": 1.0,
|
| 2245 |
+
"step": 344
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"completion_length": 42.6744801402092,
|
| 2249 |
+
"epoch": 0.18453333333333333,
|
| 2250 |
+
"grad_norm": 0.0035741821948086553,
|
| 2251 |
+
"kl": 0.34521484375,
|
| 2252 |
+
"learning_rate": 1.3394116879776567e-07,
|
| 2253 |
+
"loss": 0.0003,
|
| 2254 |
+
"reward": 1.0,
|
| 2255 |
+
"reward_std": 0.0,
|
| 2256 |
+
"rewards/equation_reward_func": 0.0,
|
| 2257 |
+
"rewards/format_reward_func": 1.0,
|
| 2258 |
+
"step": 346
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"completion_length": 32.55729299783707,
|
| 2262 |
+
"epoch": 0.1856,
|
| 2263 |
+
"grad_norm": 0.0047430778934828615,
|
| 2264 |
+
"kl": 0.3699951171875,
|
| 2265 |
+
"learning_rate": 1.2907112777606576e-07,
|
| 2266 |
+
"loss": 0.0004,
|
| 2267 |
+
"reward": 1.0,
|
| 2268 |
+
"reward_std": 0.0,
|
| 2269 |
+
"rewards/equation_reward_func": 0.0,
|
| 2270 |
+
"rewards/format_reward_func": 1.0,
|
| 2271 |
+
"step": 348
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"completion_length": 32.25260508060455,
|
| 2275 |
+
"epoch": 0.18666666666666668,
|
| 2276 |
+
"grad_norm": 0.003936651676485424,
|
| 2277 |
+
"kl": 0.3486328125,
|
| 2278 |
+
"learning_rate": 1.2427811852511395e-07,
|
| 2279 |
+
"loss": 0.0003,
|
| 2280 |
+
"reward": 1.0,
|
| 2281 |
+
"reward_std": 0.0,
|
| 2282 |
+
"rewards/equation_reward_func": 0.0,
|
| 2283 |
+
"rewards/format_reward_func": 1.0,
|
| 2284 |
+
"step": 350
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"completion_length": 30.427084147930145,
|
| 2288 |
+
"epoch": 0.18773333333333334,
|
| 2289 |
+
"grad_norm": 0.0023807576550031397,
|
| 2290 |
+
"kl": 0.3360595703125,
|
| 2291 |
+
"learning_rate": 1.1956313642169973e-07,
|
| 2292 |
+
"loss": 0.0003,
|
| 2293 |
+
"reward": 1.0,
|
| 2294 |
+
"reward_std": 0.0,
|
| 2295 |
+
"rewards/equation_reward_func": 0.0,
|
| 2296 |
+
"rewards/format_reward_func": 1.0,
|
| 2297 |
+
"step": 352
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"completion_length": 34.77864670753479,
|
| 2301 |
+
"epoch": 0.1888,
|
| 2302 |
+
"grad_norm": 0.15089837821656296,
|
| 2303 |
+
"kl": 0.34228515625,
|
| 2304 |
+
"learning_rate": 1.1492716063850971e-07,
|
| 2305 |
+
"loss": 0.0003,
|
| 2306 |
+
"reward": 0.9973958358168602,
|
| 2307 |
+
"reward_std": 0.007365695666521788,
|
| 2308 |
+
"rewards/equation_reward_func": 0.0,
|
| 2309 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 2310 |
+
"step": 354
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"completion_length": 25.59895896911621,
|
| 2314 |
+
"epoch": 0.18986666666666666,
|
| 2315 |
+
"grad_norm": 0.004901087575209092,
|
| 2316 |
+
"kl": 0.3863525390625,
|
| 2317 |
+
"learning_rate": 1.1037115394078162e-07,
|
| 2318 |
+
"loss": 0.0004,
|
| 2319 |
+
"reward": 1.0,
|
| 2320 |
+
"reward_std": 0.0,
|
| 2321 |
+
"rewards/equation_reward_func": 0.0,
|
| 2322 |
+
"rewards/format_reward_func": 1.0,
|
| 2323 |
+
"step": 356
|
| 2324 |
+
},
|
| 2325 |
+
{
|
| 2326 |
+
"completion_length": 27.411459028720856,
|
| 2327 |
+
"epoch": 0.19093333333333334,
|
| 2328 |
+
"grad_norm": 0.2515997335919245,
|
| 2329 |
+
"kl": 0.415283203125,
|
| 2330 |
+
"learning_rate": 1.058960624863629e-07,
|
| 2331 |
+
"loss": 0.0004,
|
| 2332 |
+
"reward": 0.9947916716337204,
|
| 2333 |
+
"reward_std": 0.014731391333043575,
|
| 2334 |
+
"rewards/equation_reward_func": 0.0,
|
| 2335 |
+
"rewards/format_reward_func": 0.9947916716337204,
|
| 2336 |
+
"step": 358
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"completion_length": 24.864584028720856,
|
| 2340 |
+
"epoch": 0.192,
|
| 2341 |
+
"grad_norm": 0.0046814573951274975,
|
| 2342 |
+
"kl": 0.3677978515625,
|
| 2343 |
+
"learning_rate": 1.015028156292212e-07,
|
| 2344 |
+
"loss": 0.0004,
|
| 2345 |
+
"reward": 1.0,
|
| 2346 |
+
"reward_std": 0.0,
|
| 2347 |
+
"rewards/equation_reward_func": 0.0,
|
| 2348 |
+
"rewards/format_reward_func": 1.0,
|
| 2349 |
+
"step": 360
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"completion_length": 26.3255215883255,
|
| 2353 |
+
"epoch": 0.19306666666666666,
|
| 2354 |
+
"grad_norm": 0.003523171914321313,
|
| 2355 |
+
"kl": 0.3526611328125,
|
| 2356 |
+
"learning_rate": 9.719232572644187e-08,
|
| 2357 |
+
"loss": 0.0004,
|
| 2358 |
+
"reward": 1.0,
|
| 2359 |
+
"reward_std": 0.0,
|
| 2360 |
+
"rewards/equation_reward_func": 0.0,
|
| 2361 |
+
"rewards/format_reward_func": 1.0,
|
| 2362 |
+
"step": 362
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"completion_length": 17.119792222976685,
|
| 2366 |
+
"epoch": 0.19413333333333332,
|
| 2367 |
+
"grad_norm": 0.0033418329191399116,
|
| 2368 |
+
"kl": 0.3404541015625,
|
| 2369 |
+
"learning_rate": 9.296548794875658e-08,
|
| 2370 |
+
"loss": 0.0003,
|
| 2371 |
+
"reward": 1.0,
|
| 2372 |
+
"reward_std": 0.0,
|
| 2373 |
+
"rewards/equation_reward_func": 0.0,
|
| 2374 |
+
"rewards/format_reward_func": 1.0,
|
| 2375 |
+
"step": 364
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"completion_length": 23.22135478258133,
|
| 2379 |
+
"epoch": 0.1952,
|
| 2380 |
+
"grad_norm": 0.003827547965209838,
|
| 2381 |
+
"kl": 0.3699951171875,
|
| 2382 |
+
"learning_rate": 8.882318009464123e-08,
|
| 2383 |
+
"loss": 0.0004,
|
| 2384 |
+
"reward": 1.0,
|
| 2385 |
+
"reward_std": 0.0,
|
| 2386 |
+
"rewards/equation_reward_func": 0.0,
|
| 2387 |
+
"rewards/format_reward_func": 1.0,
|
| 2388 |
+
"step": 366
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"completion_length": 25.372396528720856,
|
| 2392 |
+
"epoch": 0.19626666666666667,
|
| 2393 |
+
"grad_norm": 0.005771301327474302,
|
| 2394 |
+
"kl": 0.390869140625,
|
| 2395 |
+
"learning_rate": 8.476626240802099e-08,
|
| 2396 |
+
"loss": 0.0004,
|
| 2397 |
+
"reward": 1.0,
|
| 2398 |
+
"reward_std": 0.0,
|
| 2399 |
+
"rewards/equation_reward_func": 0.0,
|
| 2400 |
+
"rewards/format_reward_func": 1.0,
|
| 2401 |
+
"step": 368
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"completion_length": 28.466146528720856,
|
| 2405 |
+
"epoch": 0.19733333333333333,
|
| 2406 |
+
"grad_norm": 0.005304014054373385,
|
| 2407 |
+
"kl": 0.3919677734375,
|
| 2408 |
+
"learning_rate": 8.079557739962128e-08,
|
| 2409 |
+
"loss": 0.0004,
|
| 2410 |
+
"reward": 0.9973958358168602,
|
| 2411 |
+
"reward_std": 0.007365695666521788,
|
| 2412 |
+
"rewards/equation_reward_func": 0.0,
|
| 2413 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 2414 |
+
"step": 370
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"completion_length": 28.109375596046448,
|
| 2418 |
+
"epoch": 0.1984,
|
| 2419 |
+
"grad_norm": 0.008746822168976117,
|
| 2420 |
+
"kl": 0.381591796875,
|
| 2421 |
+
"learning_rate": 7.691194967200098e-08,
|
| 2422 |
+
"loss": 0.0004,
|
| 2423 |
+
"reward": 1.0,
|
| 2424 |
+
"reward_std": 0.0,
|
| 2425 |
+
"rewards/equation_reward_func": 0.0,
|
| 2426 |
+
"rewards/format_reward_func": 1.0,
|
| 2427 |
+
"step": 372
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"completion_length": 29.783854722976685,
|
| 2431 |
+
"epoch": 0.19946666666666665,
|
| 2432 |
+
"grad_norm": 0.003852702379103747,
|
| 2433 |
+
"kl": 0.3544921875,
|
| 2434 |
+
"learning_rate": 7.311618574830569e-08,
|
| 2435 |
+
"loss": 0.0004,
|
| 2436 |
+
"reward": 1.0,
|
| 2437 |
+
"reward_std": 0.0,
|
| 2438 |
+
"rewards/equation_reward_func": 0.0,
|
| 2439 |
+
"rewards/format_reward_func": 1.0,
|
| 2440 |
+
"step": 374
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"completion_length": 37.625001072883606,
|
| 2444 |
+
"epoch": 0.20053333333333334,
|
| 2445 |
+
"grad_norm": 0.006486759121452462,
|
| 2446 |
+
"kl": 0.382568359375,
|
| 2447 |
+
"learning_rate": 6.940907390477457e-08,
|
| 2448 |
+
"loss": 0.0004,
|
| 2449 |
+
"reward": 1.0,
|
| 2450 |
+
"reward_std": 0.0,
|
| 2451 |
+
"rewards/equation_reward_func": 0.0,
|
| 2452 |
+
"rewards/format_reward_func": 1.0,
|
| 2453 |
+
"step": 376
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"completion_length": 22.640625834465027,
|
| 2457 |
+
"epoch": 0.2016,
|
| 2458 |
+
"grad_norm": 0.005855790645544301,
|
| 2459 |
+
"kl": 0.3614501953125,
|
| 2460 |
+
"learning_rate": 6.579138400703715e-08,
|
| 2461 |
+
"loss": 0.0004,
|
| 2462 |
+
"reward": 1.0,
|
| 2463 |
+
"reward_std": 0.0,
|
| 2464 |
+
"rewards/equation_reward_func": 0.0,
|
| 2465 |
+
"rewards/format_reward_func": 1.0,
|
| 2466 |
+
"step": 378
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"completion_length": 33.35937559604645,
|
| 2470 |
+
"epoch": 0.20266666666666666,
|
| 2471 |
+
"grad_norm": 0.0045352649165850195,
|
| 2472 |
+
"kl": 0.3572998046875,
|
| 2473 |
+
"learning_rate": 6.22638673502327e-08,
|
| 2474 |
+
"loss": 0.0004,
|
| 2475 |
+
"reward": 1.0,
|
| 2476 |
+
"reward_std": 0.0,
|
| 2477 |
+
"rewards/equation_reward_func": 0.0,
|
| 2478 |
+
"rewards/format_reward_func": 1.0,
|
| 2479 |
+
"step": 380
|
| 2480 |
+
},
|
| 2481 |
+
{
|
| 2482 |
+
"completion_length": 37.419271528720856,
|
| 2483 |
+
"epoch": 0.20373333333333332,
|
| 2484 |
+
"grad_norm": 0.0055192004314618985,
|
| 2485 |
+
"kl": 0.3720703125,
|
| 2486 |
+
"learning_rate": 5.882725650298787e-08,
|
| 2487 |
+
"loss": 0.0004,
|
| 2488 |
+
"reward": 1.0,
|
| 2489 |
+
"reward_std": 0.0,
|
| 2490 |
+
"rewards/equation_reward_func": 0.0,
|
| 2491 |
+
"rewards/format_reward_func": 1.0,
|
| 2492 |
+
"step": 382
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"completion_length": 35.075521647930145,
|
| 2496 |
+
"epoch": 0.2048,
|
| 2497 |
+
"grad_norm": 0.00906841179693692,
|
| 2498 |
+
"kl": 0.388671875,
|
| 2499 |
+
"learning_rate": 5.548226515528132e-08,
|
| 2500 |
+
"loss": 0.0004,
|
| 2501 |
+
"reward": 1.0,
|
| 2502 |
+
"reward_std": 0.0,
|
| 2503 |
+
"rewards/equation_reward_func": 0.0,
|
| 2504 |
+
"rewards/format_reward_func": 1.0,
|
| 2505 |
+
"step": 384
|
| 2506 |
+
},
|
| 2507 |
+
{
|
| 2508 |
+
"completion_length": 44.68489694595337,
|
| 2509 |
+
"epoch": 0.20586666666666667,
|
| 2510 |
+
"grad_norm": 0.005997242826384812,
|
| 2511 |
+
"kl": 0.363037109375,
|
| 2512 |
+
"learning_rate": 5.222958797023036e-08,
|
| 2513 |
+
"loss": 0.0004,
|
| 2514 |
+
"reward": 1.0,
|
| 2515 |
+
"reward_std": 0.0,
|
| 2516 |
+
"rewards/equation_reward_func": 0.0,
|
| 2517 |
+
"rewards/format_reward_func": 1.0,
|
| 2518 |
+
"step": 386
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"completion_length": 36.622397005558014,
|
| 2522 |
+
"epoch": 0.20693333333333333,
|
| 2523 |
+
"grad_norm": 0.10183351032451872,
|
| 2524 |
+
"kl": 0.375244140625,
|
| 2525 |
+
"learning_rate": 4.9069900439828115e-08,
|
| 2526 |
+
"loss": 0.0004,
|
| 2527 |
+
"reward": 0.9973958358168602,
|
| 2528 |
+
"reward_std": 0.007365695666521788,
|
| 2529 |
+
"rewards/equation_reward_func": 0.0,
|
| 2530 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 2531 |
+
"step": 388
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"completion_length": 27.940105140209198,
|
| 2535 |
+
"epoch": 0.208,
|
| 2536 |
+
"grad_norm": 0.006355436428529637,
|
| 2537 |
+
"kl": 0.3638916015625,
|
| 2538 |
+
"learning_rate": 4.600385874466256e-08,
|
| 2539 |
+
"loss": 0.0004,
|
| 2540 |
+
"reward": 1.0,
|
| 2541 |
+
"reward_std": 0.0,
|
| 2542 |
+
"rewards/equation_reward_func": 0.0,
|
| 2543 |
+
"rewards/format_reward_func": 1.0,
|
| 2544 |
+
"step": 390
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"completion_length": 26.825521230697632,
|
| 2548 |
+
"epoch": 0.20906666666666668,
|
| 2549 |
+
"grad_norm": 0.005474495697015851,
|
| 2550 |
+
"kl": 0.3861083984375,
|
| 2551 |
+
"learning_rate": 4.303209961764587e-08,
|
| 2552 |
+
"loss": 0.0004,
|
| 2553 |
+
"reward": 1.0,
|
| 2554 |
+
"reward_std": 0.0,
|
| 2555 |
+
"rewards/equation_reward_func": 0.0,
|
| 2556 |
+
"rewards/format_reward_func": 1.0,
|
| 2557 |
+
"step": 392
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"completion_length": 43.97395968437195,
|
| 2561 |
+
"epoch": 0.21013333333333334,
|
| 2562 |
+
"grad_norm": 0.0034248967357224185,
|
| 2563 |
+
"kl": 0.3330078125,
|
| 2564 |
+
"learning_rate": 4.015524021178196e-08,
|
| 2565 |
+
"loss": 0.0003,
|
| 2566 |
+
"reward": 1.0,
|
| 2567 |
+
"reward_std": 0.0,
|
| 2568 |
+
"rewards/equation_reward_func": 0.0,
|
| 2569 |
+
"rewards/format_reward_func": 1.0,
|
| 2570 |
+
"step": 394
|
| 2571 |
+
},
|
| 2572 |
+
{
|
| 2573 |
+
"completion_length": 33.33333444595337,
|
| 2574 |
+
"epoch": 0.2112,
|
| 2575 |
+
"grad_norm": 0.006434693223212768,
|
| 2576 |
+
"kl": 0.3935546875,
|
| 2577 |
+
"learning_rate": 3.7373877972001255e-08,
|
| 2578 |
+
"loss": 0.0004,
|
| 2579 |
+
"reward": 1.0,
|
| 2580 |
+
"reward_std": 0.0,
|
| 2581 |
+
"rewards/equation_reward_func": 0.0,
|
| 2582 |
+
"rewards/format_reward_func": 1.0,
|
| 2583 |
+
"step": 396
|
| 2584 |
+
},
|
| 2585 |
+
{
|
| 2586 |
+
"completion_length": 32.841146647930145,
|
| 2587 |
+
"epoch": 0.21226666666666666,
|
| 2588 |
+
"grad_norm": 0.004929606396670415,
|
| 2589 |
+
"kl": 0.3687744140625,
|
| 2590 |
+
"learning_rate": 3.46885905110873e-08,
|
| 2591 |
+
"loss": 0.0004,
|
| 2592 |
+
"reward": 1.0,
|
| 2593 |
+
"reward_std": 0.0,
|
| 2594 |
+
"rewards/equation_reward_func": 0.0,
|
| 2595 |
+
"rewards/format_reward_func": 1.0,
|
| 2596 |
+
"step": 398
|
| 2597 |
+
},
|
| 2598 |
+
{
|
| 2599 |
+
"completion_length": 25.093750596046448,
|
| 2600 |
+
"epoch": 0.21333333333333335,
|
| 2601 |
+
"grad_norm": 0.005791877214539501,
|
| 2602 |
+
"kl": 0.3695068359375,
|
| 2603 |
+
"learning_rate": 3.20999354897229e-08,
|
| 2604 |
+
"loss": 0.0004,
|
| 2605 |
+
"reward": 1.0,
|
| 2606 |
+
"reward_std": 0.0,
|
| 2607 |
+
"rewards/equation_reward_func": 0.0,
|
| 2608 |
+
"rewards/format_reward_func": 1.0,
|
| 2609 |
+
"step": 400
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"completion_length": 32.02604269981384,
|
| 2613 |
+
"epoch": 0.2144,
|
| 2614 |
+
"grad_norm": 0.005202107306989119,
|
| 2615 |
+
"kl": 0.3798828125,
|
| 2616 |
+
"learning_rate": 2.9608450500678562e-08,
|
| 2617 |
+
"loss": 0.0004,
|
| 2618 |
+
"reward": 1.0,
|
| 2619 |
+
"reward_std": 0.0,
|
| 2620 |
+
"rewards/equation_reward_func": 0.0,
|
| 2621 |
+
"rewards/format_reward_func": 1.0,
|
| 2622 |
+
"step": 402
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"completion_length": 27.648438274860382,
|
| 2626 |
+
"epoch": 0.21546666666666667,
|
| 2627 |
+
"grad_norm": 0.005371398662773084,
|
| 2628 |
+
"kl": 0.3856201171875,
|
| 2629 |
+
"learning_rate": 2.721465295716996e-08,
|
| 2630 |
+
"loss": 0.0004,
|
| 2631 |
+
"reward": 1.0,
|
| 2632 |
+
"reward_std": 0.0,
|
| 2633 |
+
"rewards/equation_reward_func": 0.0,
|
| 2634 |
+
"rewards/format_reward_func": 1.0,
|
| 2635 |
+
"step": 404
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"completion_length": 27.742188096046448,
|
| 2639 |
+
"epoch": 0.21653333333333333,
|
| 2640 |
+
"grad_norm": 0.006378137301419084,
|
| 2641 |
+
"kl": 0.3594970703125,
|
| 2642 |
+
"learning_rate": 2.4919039985404622e-08,
|
| 2643 |
+
"loss": 0.0004,
|
| 2644 |
+
"reward": 1.0,
|
| 2645 |
+
"reward_std": 0.0,
|
| 2646 |
+
"rewards/equation_reward_func": 0.0,
|
| 2647 |
+
"rewards/format_reward_func": 1.0,
|
| 2648 |
+
"step": 406
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"completion_length": 24.539063334465027,
|
| 2652 |
+
"epoch": 0.2176,
|
| 2653 |
+
"grad_norm": 0.006280450050690689,
|
| 2654 |
+
"kl": 0.3812255859375,
|
| 2655 |
+
"learning_rate": 2.2722088321343258e-08,
|
| 2656 |
+
"loss": 0.0004,
|
| 2657 |
+
"reward": 1.0,
|
| 2658 |
+
"reward_std": 0.0,
|
| 2659 |
+
"rewards/equation_reward_func": 0.0,
|
| 2660 |
+
"rewards/format_reward_func": 1.0,
|
| 2661 |
+
"step": 408
|
| 2662 |
+
},
|
| 2663 |
+
{
|
| 2664 |
+
"completion_length": 29.023438096046448,
|
| 2665 |
+
"epoch": 0.21866666666666668,
|
| 2666 |
+
"grad_norm": 0.007309421154996187,
|
| 2667 |
+
"kl": 0.390380859375,
|
| 2668 |
+
"learning_rate": 2.0624254211693894e-08,
|
| 2669 |
+
"loss": 0.0004,
|
| 2670 |
+
"reward": 1.0,
|
| 2671 |
+
"reward_std": 0.0,
|
| 2672 |
+
"rewards/equation_reward_func": 0.0,
|
| 2673 |
+
"rewards/format_reward_func": 1.0,
|
| 2674 |
+
"step": 410
|
| 2675 |
+
},
|
| 2676 |
+
{
|
| 2677 |
+
"completion_length": 27.057292222976685,
|
| 2678 |
+
"epoch": 0.21973333333333334,
|
| 2679 |
+
"grad_norm": 0.004269729933642793,
|
| 2680 |
+
"kl": 0.3533935546875,
|
| 2681 |
+
"learning_rate": 1.8625973319162602e-08,
|
| 2682 |
+
"loss": 0.0004,
|
| 2683 |
+
"reward": 1.0,
|
| 2684 |
+
"reward_std": 0.0,
|
| 2685 |
+
"rewards/equation_reward_func": 0.0,
|
| 2686 |
+
"rewards/format_reward_func": 1.0,
|
| 2687 |
+
"step": 412
|
| 2688 |
+
},
|
| 2689 |
+
{
|
| 2690 |
+
"completion_length": 29.026042342185974,
|
| 2691 |
+
"epoch": 0.2208,
|
| 2692 |
+
"grad_norm": 0.0076460713410529055,
|
| 2693 |
+
"kl": 0.4102783203125,
|
| 2694 |
+
"learning_rate": 1.672766063197789e-08,
|
| 2695 |
+
"loss": 0.0004,
|
| 2696 |
+
"reward": 0.9973958358168602,
|
| 2697 |
+
"reward_std": 0.007365695666521788,
|
| 2698 |
+
"rewards/equation_reward_func": 0.0,
|
| 2699 |
+
"rewards/format_reward_func": 0.9973958358168602,
|
| 2700 |
+
"step": 414
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"completion_length": 39.75000101327896,
|
| 2704 |
+
"epoch": 0.22186666666666666,
|
| 2705 |
+
"grad_norm": 0.00506495764972648,
|
| 2706 |
+
"kl": 0.365234375,
|
| 2707 |
+
"learning_rate": 1.492971037770924e-08,
|
| 2708 |
+
"loss": 0.0004,
|
| 2709 |
+
"reward": 1.0,
|
| 2710 |
+
"reward_std": 0.0,
|
| 2711 |
+
"rewards/equation_reward_func": 0.0,
|
| 2712 |
+
"rewards/format_reward_func": 1.0,
|
| 2713 |
+
"step": 416
|
| 2714 |
+
},
|
| 2715 |
+
{
|
| 2716 |
+
"completion_length": 24.554688572883606,
|
| 2717 |
+
"epoch": 0.22293333333333334,
|
| 2718 |
+
"grad_norm": 0.003635383901655059,
|
| 2719 |
+
"kl": 0.356689453125,
|
| 2720 |
+
"learning_rate": 1.3232495941396637e-08,
|
| 2721 |
+
"loss": 0.0004,
|
| 2722 |
+
"reward": 1.0,
|
| 2723 |
+
"reward_std": 0.0,
|
| 2724 |
+
"rewards/equation_reward_func": 0.0,
|
| 2725 |
+
"rewards/format_reward_func": 1.0,
|
| 2726 |
+
"step": 418
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"completion_length": 27.872396171092987,
|
| 2730 |
+
"epoch": 0.224,
|
| 2731 |
+
"grad_norm": 0.005264483567820246,
|
| 2732 |
+
"kl": 0.374755859375,
|
| 2733 |
+
"learning_rate": 1.1636369788008971e-08,
|
| 2734 |
+
"loss": 0.0004,
|
| 2735 |
+
"reward": 1.0,
|
| 2736 |
+
"reward_std": 0.0,
|
| 2737 |
+
"rewards/equation_reward_func": 0.0,
|
| 2738 |
+
"rewards/format_reward_func": 1.0,
|
| 2739 |
+
"step": 420
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"completion_length": 26.59114670753479,
|
| 2743 |
+
"epoch": 0.22506666666666666,
|
| 2744 |
+
"grad_norm": 0.0037470230331634537,
|
| 2745 |
+
"kl": 0.3553466796875,
|
| 2746 |
+
"learning_rate": 1.014166338924627e-08,
|
| 2747 |
+
"loss": 0.0004,
|
| 2748 |
+
"reward": 1.0,
|
| 2749 |
+
"reward_std": 0.0,
|
| 2750 |
+
"rewards/equation_reward_func": 0.0,
|
| 2751 |
+
"rewards/format_reward_func": 1.0,
|
| 2752 |
+
"step": 422
|
| 2753 |
+
},
|
| 2754 |
+
{
|
| 2755 |
+
"completion_length": 26.52343863248825,
|
| 2756 |
+
"epoch": 0.22613333333333333,
|
| 2757 |
+
"grad_norm": 0.0050309925811448716,
|
| 2758 |
+
"kl": 0.371826171875,
|
| 2759 |
+
"learning_rate": 8.748687154702672e-09,
|
| 2760 |
+
"loss": 0.0004,
|
| 2761 |
+
"reward": 1.0,
|
| 2762 |
+
"reward_std": 0.0,
|
| 2763 |
+
"rewards/equation_reward_func": 0.0,
|
| 2764 |
+
"rewards/format_reward_func": 1.0,
|
| 2765 |
+
"step": 424
|
| 2766 |
+
},
|
| 2767 |
+
{
|
| 2768 |
+
"completion_length": 42.177085161209106,
|
| 2769 |
+
"epoch": 0.2272,
|
| 2770 |
+
"grad_norm": 0.004690685442083978,
|
| 2771 |
+
"kl": 0.3511962890625,
|
| 2772 |
+
"learning_rate": 7.457730367402549e-09,
|
| 2773 |
+
"loss": 0.0004,
|
| 2774 |
+
"reward": 1.0,
|
| 2775 |
+
"reward_std": 0.0,
|
| 2776 |
+
"rewards/equation_reward_func": 0.0,
|
| 2777 |
+
"rewards/format_reward_func": 1.0,
|
| 2778 |
+
"step": 426
|
| 2779 |
+
},
|
| 2780 |
+
{
|
| 2781 |
+
"completion_length": 33.70052194595337,
|
| 2782 |
+
"epoch": 0.22826666666666667,
|
| 2783 |
+
"grad_norm": 0.0037686388761955696,
|
| 2784 |
+
"kl": 0.35205078125,
|
| 2785 |
+
"learning_rate": 6.269061123724162e-09,
|
| 2786 |
+
"loss": 0.0004,
|
| 2787 |
+
"reward": 1.0,
|
| 2788 |
+
"reward_std": 0.0,
|
| 2789 |
+
"rewards/equation_reward_func": 0.0,
|
| 2790 |
+
"rewards/format_reward_func": 1.0,
|
| 2791 |
+
"step": 428
|
| 2792 |
+
},
|
| 2793 |
+
{
|
| 2794 |
+
"completion_length": 27.914063453674316,
|
| 2795 |
+
"epoch": 0.22933333333333333,
|
| 2796 |
+
"grad_norm": 0.003330651432955464,
|
| 2797 |
+
"kl": 0.354736328125,
|
| 2798 |
+
"learning_rate": 5.182926277723821e-09,
|
| 2799 |
+
"loss": 0.0004,
|
| 2800 |
+
"reward": 1.0,
|
| 2801 |
+
"reward_std": 0.0,
|
| 2802 |
+
"rewards/equation_reward_func": 0.0,
|
| 2803 |
+
"rewards/format_reward_func": 1.0,
|
| 2804 |
+
"step": 430
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"completion_length": 34.72656321525574,
|
| 2808 |
+
"epoch": 0.2304,
|
| 2809 |
+
"grad_norm": 0.004691669166337136,
|
| 2810 |
+
"kl": 0.3543701171875,
|
| 2811 |
+
"learning_rate": 4.199551389870659e-09,
|
| 2812 |
+
"loss": 0.0004,
|
| 2813 |
+
"reward": 1.0,
|
| 2814 |
+
"reward_std": 0.0,
|
| 2815 |
+
"rewards/equation_reward_func": 0.0,
|
| 2816 |
+
"rewards/format_reward_func": 1.0,
|
| 2817 |
+
"step": 432
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"completion_length": 21.466146230697632,
|
| 2821 |
+
"epoch": 0.23146666666666665,
|
| 2822 |
+
"grad_norm": 0.006847054731283313,
|
| 2823 |
+
"kl": 0.3829345703125,
|
| 2824 |
+
"learning_rate": 3.3191406802041688e-09,
|
| 2825 |
+
"loss": 0.0004,
|
| 2826 |
+
"reward": 1.0,
|
| 2827 |
+
"reward_std": 0.0,
|
| 2828 |
+
"rewards/equation_reward_func": 0.0,
|
| 2829 |
+
"rewards/format_reward_func": 1.0,
|
| 2830 |
+
"step": 434
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"completion_length": 24.791667580604553,
|
| 2834 |
+
"epoch": 0.23253333333333334,
|
| 2835 |
+
"grad_norm": 0.004002274972123922,
|
| 2836 |
+
"kl": 0.4224853515625,
|
| 2837 |
+
"learning_rate": 2.541876985923119e-09,
|
| 2838 |
+
"loss": 0.0004,
|
| 2839 |
+
"reward": 1.0,
|
| 2840 |
+
"reward_std": 0.0,
|
| 2841 |
+
"rewards/equation_reward_func": 0.0,
|
| 2842 |
+
"rewards/format_reward_func": 1.0,
|
| 2843 |
+
"step": 436
|
| 2844 |
+
},
|
| 2845 |
+
{
|
| 2846 |
+
"completion_length": 26.74479252099991,
|
| 2847 |
+
"epoch": 0.2336,
|
| 2848 |
+
"grad_norm": 0.004841844168761783,
|
| 2849 |
+
"kl": 0.3985595703125,
|
| 2850 |
+
"learning_rate": 1.867921723415433e-09,
|
| 2851 |
+
"loss": 0.0004,
|
| 2852 |
+
"reward": 1.0,
|
| 2853 |
+
"reward_std": 0.0,
|
| 2854 |
+
"rewards/equation_reward_func": 0.0,
|
| 2855 |
+
"rewards/format_reward_func": 1.0,
|
| 2856 |
+
"step": 438
|
| 2857 |
+
},
|
| 2858 |
+
{
|
| 2859 |
+
"completion_length": 30.26562613248825,
|
| 2860 |
+
"epoch": 0.23466666666666666,
|
| 2861 |
+
"grad_norm": 0.0037147860959137894,
|
| 2862 |
+
"kl": 0.3653564453125,
|
| 2863 |
+
"learning_rate": 1.2974148547362228e-09,
|
| 2864 |
+
"loss": 0.0004,
|
| 2865 |
+
"reward": 1.0,
|
| 2866 |
+
"reward_std": 0.0,
|
| 2867 |
+
"rewards/equation_reward_func": 0.0,
|
| 2868 |
+
"rewards/format_reward_func": 1.0,
|
| 2869 |
+
"step": 440
|
| 2870 |
+
},
|
| 2871 |
+
{
|
| 2872 |
+
"completion_length": 39.00781399011612,
|
| 2873 |
+
"epoch": 0.23573333333333332,
|
| 2874 |
+
"grad_norm": 0.0036142159951702123,
|
| 2875 |
+
"kl": 0.3477783203125,
|
| 2876 |
+
"learning_rate": 8.304748585417076e-10,
|
| 2877 |
+
"loss": 0.0003,
|
| 2878 |
+
"reward": 1.0,
|
| 2879 |
+
"reward_std": 0.0,
|
| 2880 |
+
"rewards/equation_reward_func": 0.0,
|
| 2881 |
+
"rewards/format_reward_func": 1.0,
|
| 2882 |
+
"step": 442
|
| 2883 |
+
},
|
| 2884 |
+
{
|
| 2885 |
+
"completion_length": 22.03385466337204,
|
| 2886 |
+
"epoch": 0.2368,
|
| 2887 |
+
"grad_norm": 0.004359450323671437,
|
| 2888 |
+
"kl": 0.3602294921875,
|
| 2889 |
+
"learning_rate": 4.671987054842841e-10,
|
| 2890 |
+
"loss": 0.0004,
|
| 2891 |
+
"reward": 1.0,
|
| 2892 |
+
"reward_std": 0.0,
|
| 2893 |
+
"rewards/equation_reward_func": 0.0,
|
| 2894 |
+
"rewards/format_reward_func": 1.0,
|
| 2895 |
+
"step": 444
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"completion_length": 25.75520932674408,
|
| 2899 |
+
"epoch": 0.23786666666666667,
|
| 2900 |
+
"grad_norm": 0.0055438604826414635,
|
| 2901 |
+
"kl": 0.366455078125,
|
| 2902 |
+
"learning_rate": 2.076618380744133e-10,
|
| 2903 |
+
"loss": 0.0004,
|
| 2904 |
+
"reward": 1.0,
|
| 2905 |
+
"reward_std": 0.0,
|
| 2906 |
+
"rewards/equation_reward_func": 0.0,
|
| 2907 |
+
"rewards/format_reward_func": 1.0,
|
| 2908 |
+
"step": 446
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"completion_length": 29.468751192092896,
|
| 2912 |
+
"epoch": 0.23893333333333333,
|
| 2913 |
+
"grad_norm": 0.004686764994007478,
|
| 2914 |
+
"kl": 0.3499755859375,
|
| 2915 |
+
"learning_rate": 5.191815501343066e-11,
|
| 2916 |
+
"loss": 0.0003,
|
| 2917 |
+
"reward": 1.0,
|
| 2918 |
+
"reward_std": 0.0,
|
| 2919 |
+
"rewards/equation_reward_func": 0.0,
|
| 2920 |
+
"rewards/format_reward_func": 1.0,
|
| 2921 |
+
"step": 448
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"completion_length": 23.330729842185974,
|
| 2925 |
+
"epoch": 0.24,
|
| 2926 |
+
"grad_norm": 0.002691527304301135,
|
| 2927 |
+
"kl": 0.359375,
|
| 2928 |
+
"learning_rate": 0.0,
|
| 2929 |
+
"loss": 0.0004,
|
| 2930 |
+
"reward": 1.0,
|
| 2931 |
+
"reward_std": 0.0,
|
| 2932 |
+
"rewards/equation_reward_func": 0.0,
|
| 2933 |
+
"rewards/format_reward_func": 1.0,
|
| 2934 |
+
"step": 450
|
| 2935 |
+
},
|
| 2936 |
+
{
|
| 2937 |
+
"epoch": 0.24,
|
| 2938 |
+
"step": 450,
|
| 2939 |
+
"total_flos": 0.0,
|
| 2940 |
+
"train_loss": 0.00035568679777363087,
|
| 2941 |
+
"train_runtime": 8958.7654,
|
| 2942 |
+
"train_samples_per_second": 1.206,
|
| 2943 |
+
"train_steps_per_second": 0.05
|
| 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 |
+
}
|