Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload folder using huggingface_hub

Upload final_model_deployed.pt with huggingface_hub

Upload models/cnn/saved/final_model.pt with huggingface_hub

Upload models/cnn/saved/final_model.pt with huggingface_hub

Upload final_model_deployed.pt with huggingface_hub

Upload final_model_deployed.pt with huggingface_hub

Upload final_model_deployed.pt with huggingface_hub

Upload folder using huggingface_hub

Requirements added

add HF space config

r

Server changes for deployment

added src/data/dataset/py to commits

OOP complete and functional rewrite

Changes to config.py

Download process now uses OOP. New Configuration logic uses OOP

Added Transformer training

Fixes and added transformer architecture (AST)

Corrected a major bug in labels which created a huge artifical class imbalance, model should now perform better

0.0.0.0 host for containerization and added Dcokerfile for easier local build

Removed Essentia dependency and used librosa instead because it didn't compile on arm64 architecture for docker export

dockerfile version of server.py

Hid docker files

Added dependencies list

Added support for other audio formats

Added web interface api for top k prediction

Minor changes to main.py

Added augment dataset support to CLI.

CLI access to all functions and project, from downloading data, preprocessing it, training a model, and testing on test data.

More cleaning of config.

Minor README.md changes

Refined prediction logic, the prediction folder for the test file must contain only the test file.

Config file logic and prediction logic.

Added config file. Added Prediction logic for a single audio file in a dedicated folder

removed src/data/__pycache__

Added READ.ME and src/data files

Initialize git repo.