Update README.md
Browse files
README.md
CHANGED
|
@@ -6,24 +6,27 @@ language:
|
|
| 6 |
- zh
|
| 7 |
tags:
|
| 8 |
- medical
|
|
|
|
|
|
|
| 9 |
---
|
| 10 |
-
# MedArk-KI-
|
| 11 |
MedArk-KI is an integrated large-scale, high-quality Chinese SFT dataset, designed for medical knowledge injection into LLMs.
|
| 12 |
|
| 13 |
MedArk-KI involving in two types of medical knowledge: Traditional Chinese Medicine (TCM) and Western Medicine(WM). It consists of 4 subsets as shown in the tabel:
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
| 17 |
-
|
|
| 18 |
-
|
|
| 19 |
-
|
|
| 20 |
-
|
|
|
|
|
| 21 |
|
| 22 |
# 1.DX-xxK
|
| 23 |
|
| 24 |
A total of 5,3554 high-quality Q&A pairs were collected from the DingXiang website, 丁香医生 (https://dxy.com/diseases), and underwent rigorous data cleaning.
|
| 25 |
Each answer was authored by a doctor and reviewed by another doctor, ensuring that all information is accurate and reliable.
|
| 26 |
-
These Q&A pairs encompass 8 key areas of medical knowledge across
|
| 27 |
|
| 28 |
|
| 29 |
# 2.HT-xxk
|
|
@@ -44,5 +47,4 @@ https://www.dayi.org.cn/
|
|
| 44 |
|
| 45 |
# 4.MB-xxk
|
| 46 |
|
| 47 |
-
https://github.com/scienceasdf/medical-books
|
| 48 |
-
|
|
|
|
| 6 |
- zh
|
| 7 |
tags:
|
| 8 |
- medical
|
| 9 |
+
size_categories:
|
| 10 |
+
- 1M<n<10M
|
| 11 |
---
|
| 12 |
+
# MedArk-KI-146M
|
| 13 |
MedArk-KI is an integrated large-scale, high-quality Chinese SFT dataset, designed for medical knowledge injection into LLMs.
|
| 14 |
|
| 15 |
MedArk-KI involving in two types of medical knowledge: Traditional Chinese Medicine (TCM) and Western Medicine(WM). It consists of 4 subsets as shown in the tabel:
|
| 16 |
+
|
| 17 |
+
| Name | Volume | Author | Reviewer | Type | Source | Format |
|
| 18 |
+
|------|---------|-----------|-------------|------|-------------------------|-----------------|
|
| 19 |
+
| DX | 5,3554 | Doctor | Doctor | WM | Web | HTML |
|
| 20 |
+
| HT | 135,8093| ChatGPT | ERNIE-Speed | WM | Books, Encyclopedia, Web| PDF, TXT, HTML |
|
| 21 |
+
| TCM | 3,8294 | ERNIE 4.0 | ERNIE-Speed | TCM | Books, KG | PDF, Latex, Json|
|
| 22 |
+
| MB | 1,4543 |ERNIE-Speed|ERNIE-Speed | WM | Books | Latex |
|
| 23 |
+
|Totol | 146,4484| | | | | |
|
| 24 |
|
| 25 |
# 1.DX-xxK
|
| 26 |
|
| 27 |
A total of 5,3554 high-quality Q&A pairs were collected from the DingXiang website, 丁香医生 (https://dxy.com/diseases), and underwent rigorous data cleaning.
|
| 28 |
Each answer was authored by a doctor and reviewed by another doctor, ensuring that all information is accurate and reliable.
|
| 29 |
+
These Q&A pairs encompass 8 key areas of medical knowledge across 3412 diseases in 31 departments: disease introduction, symptoms, causes, diagnosis, treatment, lifestyle, prevention, and consultation guidance.
|
| 30 |
|
| 31 |
|
| 32 |
# 2.HT-xxk
|
|
|
|
| 47 |
|
| 48 |
# 4.MB-xxk
|
| 49 |
|
| 50 |
+
https://github.com/scienceasdf/medical-books
|
|
|