HuggingFace Analyser

HuggingFace text generation and text classification models and related spaces analyser written in Python.

Get started

Local:

git clone git@github.com:AMK9978/HuggingAnalyser.git
pip install -r requirements.txt
python main.py

Docker:

git clone git@github.com:AMK9978/HuggingAnalyser.git
docker build hugging .
docker run --name hugging -itd hugging

Comparison and Results

Comparison.md

File Structure

• datasets: The obtained datasets from crawling used in the analysis
• media: The drawn figures by draw.py used in the analysis
• tests: The test file of the project containing four tests
• main.py: The main file of the project containing the crawler
• drawer.py: The helper module for drawing the figures

Authors

• Amir Karimi AMK9978

Built With

• Python 3.11
• requests
• BeautifulSoup4
• pandas
• matplotlib
• Black
• isort

Due to the above requirements, the application is blazing fast and lightweight. It does not use any browser instance. No captcha test was recorded during crawling.

License

This project is licensed under the MIT License.

Discussion on HuggingFace

There is an open thread on HuggingFace opened by Amir Karimi requesting the app size feature from HuggingFace Hub API:

Thread