Added optional feature maps

fschmid56 / EfficientAT

This repository aims at providing efficient CNNs for Audio Tagging. We provide AudioSet pre-trained models ready for downstream training and extraction of audio embeddings.

MIT License

233 stars 44 forks source link

Added optional feature maps #13

Closed turian closed 1 year ago

turian commented 1 year ago

Addresses #11 and unblocks https://github.com/lucidrains/audiolm-pytorch/issues/177 without changing functionality for existing users.