Share the environment that worked.
I have been successfully run. Environment follows: cuda 11.8 python 3.10.13 pytorch 2.1.1 causal_conv1d 1.1.1 mamba-ssm 1.2.0.post1
pip install torch==2.1.1 torchvision==0.16.1 torchaudio==2.1.1 --index-url https://download.pytorch.org/whl/cu118
pip install causal_conv1d==1.1.1
pip install mamba-ssm==1.2.0.post1
# copy mamba-ssm dir in vim to conda env site-package dir
cp -rf mamba-1p1p1/mamba_ssm /opt/miniconda3/envs/mamba/lib/python3.10/site-packages
it also works for me, thanks
Hello, my environment is the same as yours, but I still report an error: Mamba.init() got an unexpected keyword argument 'bimamba'. Do you know how to solve it?
Hello, my environment is the same as yours, but I still report an error: Mamba.init() got an unexpected keyword argument 'bimamba'. Do you know how to solve it?
This issue may be you didn't copy the mamba-ssm dir of Vim to the envs site-packages dir. Because Vim's author modified the mamba-ssm in their code. The mamba-ssm package that installed via pip is official package. So, you must copy mamba-ssm dir of Vim to virtual environment after pip install mamba-ssm.
cp -rf mamba-1p1p1/mamba_ssm /opt/miniconda3/envs/**mamba**/lib/python3.10/site-packages
Hello, my environment is the same as yours, but I still report an error: Mamba.init() got an unexpected keyword argument 'bimamba'. Do you know how to solve it?
This issue may be you didn't copy the mamba-ssm dir of Vim to the envs site-packages dir. Because Vim's author modified the mamba-ssm in their code. The mamba-ssm package that installed via pip is official package. So, you must copy mamba-ssm dir of Vim to virtual environment after pip install mamba-ssm.
cp -rf mamba-1p1p1/mamba_ssm /opt/miniconda3/envs/**mamba**/lib/python3.10/site-packages
Thanks a lot. I know what the problem is
This also worked for me, thank you!
Hello, in my environment “causal_conv1d_cuda“ can't be imported ,do you encount the same problem?
Hello, in my environment “causal_conv1d_cuda“ can't be imported ,do you encount the same problem?您好,在我的环境中“causal_conv1d_cuda”无法导入,您遇到同样的问题吗?
This problem might caused by the wrong version of causal_conv1d mamba-ssm. Follow this issue, import exact version of the packages as causal_conv1d==1.1.1 and mamba-ssm==1.2.0.post1!
I think the cuda version is important to keep 11.8.
Thank you for your sharing, but you did not find that this environment is only for vim model, and it is completely unavailable for another seg model, and the author did not upload complete code, may I ask if you have successfully reproduced the seg model?
Thank you for your sharing, but you did not find that this environment is only for vim model, and it is completely unavailable for another seg model, and the author did not upload complete code, may I ask if you have successfully reproduced the seg model?
I also want to ask.
Hello, I have encountered the same problem - TypeError: Mamba.init() got an unexpected keyword argument 'bimamba_type'. and I have also copied the mamba-ssm dir of Vim to the envs site-packages dir. It is useful and I solved this problem. But when I run the code, it brings a new problem and it looks like the same problem-Mamba.init() got an unexpected keyword argument 'if_devide_out'. How can I solve it?
I have been successfully run. Environment follows: cuda 11.8 python 3.10.13 pytorch 2.1.1 causal_conv1d 1.1.1 mamba-ssm 1.2.0.post1
pip install torch==2.1.1 torchvision==0.16.1 torchaudio==2.1.1 --index-url https://download.pytorch.org/whl/cu118 pip install causal_conv1d==1.1.1 pip install mamba-ssm==1.2.0.post1 # copy mamba-ssm dir in vim to conda env site-package dir cp -rf mamba-1p1p1/mamba_ssm /opt/miniconda3/envs/mamba/lib/python3.10/site-packages
Hello, I have encountered the same problem - TypeError: Mamba.init() got an unexpected keyword argument 'bimamba_type'. and I have also copied the mamba-ssm dir of Vim to the envs site-packages dir. It is useful and I solved this problem. But when I run the code, it brings a new problem and it looks like the same problem-Mamba.init() got an unexpected keyword argument 'if_devide_out'. How can I solve it?