Add Windows support

[YavorGIvanov]

• fairseq2 currently does not support Windows.
• Related Issue

We use fairseq2 just for the speech tokenizer (the model converting audio waveform to XLSR units), so should be straightforward to workaround this by either patching fairseq2 to work on Windows or similar approach.

[Maki9009]

soo does it work on windows now?

[YavorGIvanov]

Not yet. The issue has not been tackled, but I expect soon either some of us or the community to make progress.

[Maki9009]

what about docker or pinokio is it possible to get i to run on those?

[Maki9009]

Installing on Windows
fairseq2 does not have native support for Windows and there are no plans to support it in the foreseeable future. However, you can use fairseq2 via the Windows Subsystem for Linux (a.k.a. WSL) along with full CUDA support introduced in WSL 2. Please follow the instructions in the Installing on Linux section for a WSL-based installation.

[Maki9009]

would wsl2 not work? to get it to run on windows?

[YavorGIvanov]

It should work, yes @Maki9009

[Maki9009]

ill be honest.. idk how to do that. sooo i hope you guys implement it somehow would really appreciate it. if you get the time to do so

[coezbek]

Just commenting: Playdiffusion works well on WSL.