O. Inha

For the singing, you might want to check out something like: - https://dreamtonics.com/en/synthesizerv - https://emvoiceapp.com/

This might be closest to what you hoped: https://www.suno.ai I don't mean to advertise other services in Github issues, but just because I remember this issue, as it's a reminder...

Thanks! Applied this fix to the Colab notebook.

Fixed, please refresh notebook and try again. I've been doing a lot of changes to the notebook in the past days, particularly concerning audio quality enhancements (it now produces relatively...

In the Colab notebook stereo is _simulated_ by separate generations of left and right channel and mashing them together into one stereo file. Left channel audio is generated by text...

Unable to reproduce. Possible causes for the error: - File path is not relative to My Drive. If you have "My Drive" or "MyDrive" in your path, it's very likely...

Not that I know of (apart from what formats torchaudio supports). I'm testing with about 9 sec long WAV (PCM S16 LE) stereo, 44100 hz, 16-bit, and experience no issues....

Duplicate of [issue#95](https://github.com/haoheliu/AudioLDM/issues/95) Solution: `pip install --upgrade transformers==4.29.0`

Not what you're asking, but fyi anyway, that I've added stereo _simulation_ and 44.1 kHz _conversion_ to the Colab notebook. If you set `stereo_width` > 0, it will generate a...

audioldm-s-full: https://zenodo.org/record/7600541/files/audioldm-s-full?download=1 audioldm-l-full: https://zenodo.org/record/7698295/files/audioldm-full-l.ckpt?download=1 audioldm-s-full-v2: https://zenodo.org/record/7698295/files/audioldm-full-s-v2.ckpt?download=1 audioldm-m-text-ft: https://zenodo.org/record/7813012/files/audioldm-m-text-ft.ckpt?download=1 audioldm-s-text-ft: https://zenodo.org/record/7813012/files/audioldm-s-text-ft.ckpt?download=1 audioldm-m-full: https://zenodo.org/record/7813012/files/audioldm-m-full.ckpt?download=1 Copy checkpoint of choice to directory `~/.cache/audioldm/`. Afaik `audioldm-m-full` is recommended. If I'm not mistaken, the largest one,...