Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
### Introduction
|
| 3 |
+
|
| 4 |
+
This is a modified dataset from a part of [JMedBench](https://huggingface.co/datasets/Coldog2333/JMedBench).
|
| 5 |
+
It consists of 674,954 lines of Japanese and English medical text. The column is named "text".
|
| 6 |
+
|
| 7 |
+
Only the train subset is extracted from the data source and then concatenated, where the question-answering samples are made into one sentence by
|
| 8 |
+
```
|
| 9 |
+
Questions:{Question}Answer:{Answer}
|
| 10 |
+
```
|
| 11 |
+
for English samples and
|
| 12 |
+
```
|
| 13 |
+
質問:{Question}回答:{Answer}
|
| 14 |
+
```
|
| 15 |
+
for Japanese samples.
|
| 16 |
+
|
| 17 |
+
No other modification was applied.
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
### Limitations
|
| 21 |
+
|
| 22 |
+
Please read the original data sources.
|
| 23 |
+
|
| 24 |
+
### Licenses
|
| 25 |
+
|
| 26 |
+
Please follow the original data sources.
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
### Intention
|
| 30 |
+
|
| 31 |
+
This dataset is meant to be used for the continual pretraining of medical LLMs.
|
| 32 |
+
|
| 33 |
+
No other usage is recommended.
|
| 34 |
+
|
| 35 |
+
|