Inspiration: Working effectively on some classification models which can be trained in 10-15 minutes

What it does: Split the images and then train KNN model on it. At the end test the model and tell the accuracy of the trained model.

How we built it: Using the KNN on the Animal classification

Challenges we ran into: The images are separated in their respective folders. But they are not separated in train test.

Accomplishments that we're proud of: Using the trained model, created an application using streamlit. then deploy it on hugging face. step by step process

What we learned: Create a new space on Hugging Face and select the Streamlit template. Upload app.py, knn_model.pkl, and class_names.pkl files to the space.Hugging Face will automatically deploy the app. Kaggle Constraints: Kaggle does not support running interactive Streamlit apps within the notebook itself, but you can use it for developing your model and app. Hugging Face Deployment: It's better to deploy the app on Hugging Face Spaces, as they provide an easy-to-use platform for deploying Streamlit apps.

What's next for Animal Classification Model/ Bot: scikit-learn: Used for loading and working with the KNN model (joblib is used to load it).

Built With

  • huggingface
  • kaggle
  • python
  • streamlit
Share this project:

Updates