env error:Could not build wheels for mamba_ssm, which is required to install pyproject.toml-based projects**

[edwardhuanghz]

docker search anaconda3 docker pull continuumio/anaconda3 docker run -it -d --name tsldev -p 6888:8888 -p 6822:22 -p 6006:6006 -e NVIDIA_DRIVER_CAPABILITIES=compute,utility -e NVIDIA_VISIBLE_DEVICES=all --gpus all -v D:\aidev:/aidev continuumio/anaconda3 /bin/bash docker exec -it tsldev bash

conda create -n tsldev python=3.8 conda activate tsldev

conda deactivate

conda env remove --n tsldev

pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple

pip install --upgrade -i https://pypi.tuna.tsinghua.edu.cn/simple einops

when install mamba_ssm pip install --upgrade -i https://pypi.tuna.tsinghua.edu.cn/simple einops mamba_ssm occur a error,see below ERROR: Could not build wheels for mamba_ssm, which is required to install pyproject.toml-based projects

[listview998]

It seems that using mamba_ssm currently only supports Linux systems and cannot be used temporarily under Windows. First, you need to check if the system environment is Linux

[edwardhuanghz]

I run it on the docker of debian os in window 11

[Lijuming33]

全局检索mamba，把mamba : mamba模型注释，把导入语句中的mamba删除即可

[Harryx2019]

pip3 install mamba_ssm 可以装上

[ztxtech]

You'd better create a new conda environment without docker and configure a correct mirror of PyPI. Maybe the official mirror of PyPI works. Then you will know what the problem is.

# delete the tuna mirror
pip install -r requirements.txt

[wuhaixu2016]

Hi, many thanks for your discussion. mamba_ssm is not platform-friendly. Thus, we have removed the default usage of Mamba in the latest commit: https://github.com/thuml/Time-Series-Library/commit/2d734a17045419cde6bfe7fc0e7fe99d8ce8d940 If you have some problems about installing mamba_ssm, I think their official repo provides sufficient instructions: https://github.com/mamba-org/mamba