Emotion Music Model V2
This repository hosts the artifacts for the Model V2 recommender:
kmeans_model.pkl: trained KMeans modelmetadata.csv: song metadata aligned with features
How to use
Download files from this repo, place them in:
modelV2/data/processed/
Then run:
python -m modelV2.src.recommender
Training
These artifacts were generated using the scripts in this project:
python modelV2/src/preprocess.py
python modelV2/train.py
Notes
The recommender maps emotions to clusters and samples songs from the target cluster.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support