Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

activateEnvironment.sh is missing from the repository #2370

Closed
scvance opened this issue Jan 22, 2024 · 3 comments · Fixed by #2409
Closed

activateEnvironment.sh is missing from the repository #2370

scvance opened this issue Jan 22, 2024 · 3 comments · Fixed by #2409

Comments

@scvance
Copy link

scvance commented Jan 22, 2024

activateEnvironment.sh is referenced in the slurm example linked here https://github.com/huggingface/accelerate/blob/main/examples/slurm/submit_multinode.sh

I cannot find the file activateEnvironment.sh anywhere.

Any help would be greatly appreciated. Thanks!

@muellerzr
Copy link
Collaborator

cc @TJ-Solergibert :)

@TJ-Solergibert
Copy link
Contributor

TJ-Solergibert commented Jan 23, 2024

In the activateEnvironment.sh you should include the steps to activate the environment to run the experiments. This includes setting up the python interpreter, environment variables, etc.

The following configuration is what I need to do in my cluster in order to run the experiments. First I load the necessary modules to load anaconda and GPU dependencies. Then I activate the conda environment which I want to use to perform the experiments. Finally I set up some environment variables, mainly to run the experiments offline because we don’t have internet access.

module purge
module load anaconda3/2020.02 cuda/10.2 cudnn/8.0.5 nccl/2.9.9 arrow/7.0.0 openmpi
source activate /home/nct01/nct01328/pytorch_antoni_local

export HF_HOME=/gpfs/projects/nct01/nct01328/
export HF_LOCAL_HOME=/gpfs/projects/nct01/nct01328/HF_LOCAL
export HF_DATASETS_OFFLINE=1
export TRANSFORMERS_OFFLINE=1
export PYTHONPATH=/home/nct01/nct01328/transformers-in-supercomputers:$PYTHONPATH 
export GPUS_PER_NODE=4

You should end up having something very similar, good luck!

@muellerzr
Copy link
Collaborator

@TJ-Solergibert maybe we want to include this in the README under the SLURM part 👀

(Or add a readme to the SLURM folder?)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants