Generating Question Answer Database from Unstructured Text
Project description
QA Genie
English | हिंदी
QA Genie is a Python package designed for generating questions and answers from unstructured data.
This package is built using the unofficial API of HuggingChat: hugchat. It leverages HuggingChat's capabilities for question and answer generation.
Note
This package is in its alpha release and more functionality will be added soon!
Update 1.0.0a3: This update enables the user to adjust iteration time. (Solves #1) Update 1.0.0a4: Enables user to get raw text output from chatbot for manual cleaning. Also fixes bugs in cleaning.
Installation
pip install qa_genie
or
pip3 install qa_genie
Usage
email = "your_email@example.com" # huggingface account email
password = "your_password" # huggingface account password
model = "meta" # use "meta" to use meta-llama/Llama-2-70b-chat-hf or "oasst" to use OpenAssistant/oasst-sft-6-llama-30b
# Initialize chatbot
chatbot = get_generator(email, password, model)
# Example usage with a single text
text = "Lorem ipsum dolor sit amet, consectetur adipiscing elit."
result_single = extract_qa(chatbot, text, num_qn=3) # returns pandas.DataFrame with num_qn questions and answers
# Example usage with multiple texts
texts = ["Text 1", "Text 2", "Text 3"]
result_multiple = extract_qas(chatbot, texts, num_qn_each=3) # return pandas.DataFrame with num_qn_each questions and answers generated for each text
Important Note
As mentioned by Soulter, Server resources are precious, it is not recommended to request this API in a high frequency.
Contributing
Feel free to contribute to QA Genie by creating issues, submitting pull requests, or suggesting improvements. Your contributions are highly appreciated :)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for qa_genie-1.0.0a4-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 11f2138fa1bf80605ae11a13b64d0a8d0385e95a7ea8d39e6cd56c8539507fc1 |
|
MD5 | f9226c573900f1e1f74ddf4e175efff4 |
|
BLAKE2b-256 | 485c9f120a65d0a93131a3e28c23d81ae793fdfe263873de60a31b9e7f83ed79 |