|
--- |
|
dataset_info: |
|
features: |
|
- name: conversations |
|
list: |
|
- name: from |
|
dtype: string |
|
- name: value |
|
dtype: string |
|
splits: |
|
- name: train |
|
num_bytes: 79860701 |
|
num_examples: 30000 |
|
download_size: 38685452 |
|
dataset_size: 79860701 |
|
configs: |
|
- config_name: default |
|
data_files: |
|
- split: train |
|
path: data/train-* |
|
task_categories: |
|
- question-answering |
|
- text-generation |
|
language: |
|
- zh |
|
--- |
|
|
|
# Kyara: Knowledge Yielding Adaptive Retrieval Augmentation for LLM Fine-tuning |
|
|
|
<p align="left"> |
|
🤗 <a href="https://huggingface.co./zake7749/gemma-2-2b-it-chinese-kyara-dpo">Hugging Face</a>  ｜ 🚀<a href="https://github.com/zake7749/kyara">Github</a>  ｜  📑 <a href="#">Paper</a>  ｜  📖 <a href="#">English</a>  ｜  📖 <a href="https://github.com/zake7749/kyara">Chinese</a> |
|
</p> |
|
<div style="text-align: center;"> |
|
<img src="https://i.imgur.com/QiWlcYJ.jpeg" alt="kyara"/> |
|
</div> |
|
|
|
This is a dataset for training [Kyara](https://huggingface.co./zake7749/gemma-2-2b-it-chinese-kyara-dpo). |
|
|
|
We applied [PersonaHub](https://huggingface.co./datasets/proj-persona/PersonaHub) to generate 50K math problems and used `Gemini-1.5-Flash` to filter out data with obvious errors in calculation and reasoning. However, this dataset still contains some computational and logical errors; please use it with caution. |
