Spaces:
Runtime error
Runtime error
- β Ensure app.py audio input sounds like training data | |
- β Use a huggingface transformer with the dataset | |
- Verify that the training spectrogram matches the predict spectrogram | |
- Count number of example misses in dataset loading | |
- Verify windowing and jitter params in Song Dataset | |
- Create an attention-based network | |
- β Increase parameter count in network | |
- Verify that labels really match what is on the music4dance site | |
- β Read the Medium series about audio DL | |
- double check \_rectify_duration | |
- β Filter out songs that have only one vote | |
- β Download songs from [Best Ballroom](https://www.youtube.com/channel/UC0bYSnzAFMwPiEjmVsrvmRg) | |
- β fix nan values | |
- Try higher mels (224) and more ffts (2048) | |
- Verify random sample of dataset outputs by hand. | |
- Train with non music data and add a non music category | |
- Add back class weights | |
- Add back multi label classification | |