Cannot install moshi-mlx

[boji123]

Backend impacted

The PyTorch implementation

Operating system

Linux

Hardware

CPU

Description

`INFO: pip is looking at multiple versions of moshi-mlx to determine which version is compatible with other requirements. This could take a while. Collecting moshi_mlx Using cached http://mirrors.tencentyun.com/pypi/packages/95/bc/f1952075f7eb11155cd818ac6a42a5dfd31f3fe7ae94856eb405bae18aa1/moshi_mlx-0.0.0-py3-none-any.whl (1.4 kB) ERROR: Cannot install moshi-mlx==0.0.0 and moshi-mlx==0.1.0 because these package versions have conflicting dependencies.

The conflict is caused by: moshi-mlx 0.1.0 depends on mlx<0.18 and >=0.17.2 moshi-mlx 0.0.0 depends on mlx<0.18 and >=0.17.2

To fix this you could try to:

1. loosen the range of package versions you've specified
2. remove package versions to allow pip to attempt to solve the dependency conflict

ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts `

Extra information

`conda create -n moshi python=3.10 conda activate moshi

export RUSTUP_UPDATE_ROOT=https://mirrors.ustc.edu.cn/rust-static/rustup export RUSTUP_DIST_SERVER=https://mirrors.ustc.edu.cn/rust-static curl https://sh.rustup.rs -sSf | sh # install Cargo, the Rust package manager . "$HOME/.cargo/env" # source Cargo

pip install moshi # moshi PyTorch, from PyPI pip install moshi_mlx # moshi MLX, from PyPI`

blocked at pip install moshi_mlx

Environment

Fill in the following information on your system.

• Operating system version:

If the backend impacted is PyTorch:

• Python version:
• PyTorch version:
• CUDA version (run python -c 'import torch; print(torch.version.cuda)'):
• GPU model and memory:

If the backend is MLX:

• Mac model:

[do-me]

You cannot install the mlx version (designed for Apple Silicon) on Linux. Try with the CUDA version if you have a GPU. CPU-only doesn't work yet afaik.

[adefossez]

closing as @do-me summarized the situation :)

[EmaSuriano]

@do-me What about for Windows? I have a GPU and I would love to see how it works there :)

[do-me]

Sure that should work with the normal cuda version. Just make sure to install the right backend (not -mlx).