File size: 1,515 Bytes
9ba99e9 5830dee 9ba99e9 5830dee |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 |
---
license: mit
language:
- en
tags:
- medi
- medical
pretty_name: python
size_categories:
- 10K<n<100K
---
# What is the Dataset About?🤷🏼♂️
---
The dataset is useful for training a Generative Language Model for the Medical application and instruction purposes, the dataset consists of various thoughs proposed by the people [**mentioned as the Human** ] and there responses including Medical Terminologies not limited to but including names of the drugs, prescriptions, yogic exercise suggessions, breathing exercise suggessions and few natural home made prescriptions.
# How the Dataset was made?😅
---
I have used all the available opensource datasets and combined them into a single datsource for training, which is completely opensourced and somewhat reliable.
* There is another refined and updated version of this datset here 👉🏼 [Link](https://huggingface.co/datasets/Mohammed-Altaf/medical-instruction-120k)
## Example Training Scripts:
* Qlora Fine Tuning -
## Tips:
This is my first dataset to upload on HuggingFace, so below are the thing I wish I could have known
* always save your final dataset before uploading to hub as a json with lines.
* The json should have the records orientation, which will be helpful while loading the dataset properly without any error.
```{python}
# use below if you are using pandas for data manipulation
train.to_json("dataset_name.json", orient='records', lines=True)
test.to_json("dataset_name.json", orient='records', lines=True)
``` |