---
title: DISC-Med-SFT
canonical_url: "https://www.modelscope.cn/datasets/AI-ModelScope/DISC-Med-SFT"
md_url: "https://www.modelscope.cn/datasets/AI-ModelScope/DISC-Med-SFT.md"
repository: AI-ModelScope/DISC-Med-SFT
last_updated: 2024-09-02
license: "Apache License 2.0"
storage_size: "765 MB"
domain:
  - configs
  - size_categories
  - task_categories
  - tasks
  - language
tasks:
  - "{config_name=default, data_files=[{path=DISC-Med-SFT_released.csv, split=train}]}"
  - "100K<n<1M"
  - conversational
  - nlp
  - zh
downloads: 2410
stars: 6
---

# DISC-Med-SFT

> DISC-Med-SFT - AI-ModelScope 在 ModelScope 开源的数据集。configs: configname: default datafiles: path: DISC-Med-SFTreleased.csv split: train language: zh license: Apache License 2.0 sizecategories: 100K<n<1M tags: medical sft chat taskcategories: question-answering…

AI-ModelScope/DISC-Med-SFT 是 ModelScope 魔搭社区上的{config_name=default, data_files=[{path=DISC-Med-SFT_released.csv, split=train}]}、100K<n<1M、conversational数据集，涉及 configs、size_categories、task_categories 领域，存储大小 765 MB，采用 Apache License 2.0 许可。

- **Repository**: AI-ModelScope/DISC-Med-SFT
- **License**: Apache License 2.0
- **Tasks**: {config_name=default, data_files=[{path=DISC-Med-SFT_released.csv, split=train}]}, 100K<n<1M, conversational, nlp, zh
- **Domain**: configs, size_categories, task_categories, tasks, language
- **Storage size**: 765 MB
- **Downloads**: 2410
- **Stars**: 6
- **Last updated**: 2024-09-02

Source: https://www.modelscope.cn/datasets/AI-ModelScope/DISC-Med-SFT

---

configs:
- config_name: default
  data_files:
  - path: DISC-Med-SFT_released.csv
    split: train
language:
- zh
license: Apache License 2.0
size_categories:
- 100K<n<1M
tags:
- medical
- sft
- chat
task_categories:
- question-answering
- conversational
tasks:
- chat
- nlp
---


本仓库拷贝自: https://huggingface.co/datasets/Flmc/DISC-Med-SFT

可以通过modelscope SDK下载
```python
from modelscope import MsDataset
train_dataset = MsDataset.load('AI-ModelScope/DISC-Med-SFT', 
                               subset_name='default', split='train').to_hf_dataset()

print(train_dataset)
"""Out[0]
Dataset({
    features: ['_id', 'source', 'conversation'],
    num_rows: 464898
})
"""
```

This is a repository containing a subset of the DISC-Med-SFT Dataset.

Check [DISC-MedLLM](https://github.com/FudanDISC/DISC-MedLLM) for more information.
