Update README.md, dependencies and scripts #41
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Update README.md to make SAM-6D more accessible. Many version upgrade in conda environment file. Minor changes to prepare.sh and demo.sh script to avoid possible errors.
|
I followed your repo and still the prepare.sh failed. Could you please suggest a fix? |
|
also regarding 2, you may find my answer here useful |
I am a bit confused by your error message, as neither version of CUDA should be installed or used. Can you please check if these are pointing to your sam6d environment: conda activate sam6d
which python
which pip
which nvccI suspect that there may be a misconfiguration that is causing the wrong packages to be used. If this is the case, please check Also, the following output would be helpful for further debugging. conda activate sam6d
conda info
conda listI have tested the updated I would appreciate your feedback. |
|
I also used to get this message when compiling PointNet: The reason why this occurs is that To fix this I had to manually obtain CUDA 12.1 from NVIDIA’s CUDA website and run the file script via export CUDA_HOME=/home/YOUR_USER/PATH_CUDA_YOU_INSTALLED/cuda-12.1/
export PATH=/home/YOUR_USER/PATH_CUDA_YOU_INSTALLED/cuda-12.1/bin:$PATH
export LD_LIBRARY_PATH=/home/YOUR_USER/PATH_CUDA_YOU_INSTALLED/cuda-12.1/lib64:$LD_LIBRARY_PATHIf you then run Once this is set, you should be able to compile PointNet without the pytorch version conflict you saw before |
|
Alternatively, if you don't want to manually install CUDA 12.1, something that also worked for me was to use the original repo from @JiehongLin which uses NOTE: I also changed the environment.yaml with cudatoolkit-dev=11.7:Once the conda environment is created and activated, conda takes care of appending to the env variables the path to the cudatoolkit-dev=11.7 and you shouldn't have issues when compiling PointNet. |
codinglist
added a commit
to codinglist/SAM-6D
that referenced
this pull request
Jan 7, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
First of all, I would like to take this opportunity to thank the authors for the great work and congratulate them on the acceptance of the paper for CVPR 2024. 🥳
This PR fixes several minor issues with the goal of making the SAM-6D code more accessible:
1. Changes in README.MD
$PWDwhen exporting: See Missing demo files #332. Updated environment.yml
3. Changes in prepare.sh and demo.sh
conda createandconda activatefrom script and instead included in README.MD: As this is on some systems, see Can't executeconda activatefrom bash script conda/conda#7980&&) in both scripts to ensure that code only continues of previous commands executed successfully: This resulted in confusing error traces and should avoid issues, e.g. ModuleNotFoundError: No module named 'pointnet2._ext' #19Disclaimer: I have tested the changes on several local machines and cloud servers. However only for running inference and not for training the PEM, as described in Pose Estimation Model (PEM) for SAM-6D.
Best regards!