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

[s2s] 1 GPU test for run_distributed_eval #8297

Closed
sshleifer opened this issue Nov 4, 2020 · 5 comments · Fixed by #8315
Closed

[s2s] 1 GPU test for run_distributed_eval #8297

sshleifer opened this issue Nov 4, 2020 · 5 comments · Fixed by #8315
Assignees
Labels
Help wanted Extra attention is needed, help appreciated

Comments

@sshleifer
Copy link
Contributor

Add test coverage for run_distributed_eval.py that can run on 1 GPU.
The command:

python -m torch.distributed.launch --nproc_per_node=1 run_distributed_eval.py --model_name Helsinki-NLP/opus-mt-en-ro --save_dir opus_wmt_en_ro_gens --data_dir wmt_en_ro

works on 1 GPU.

After adding test coverage, we could try to improve API consistency between run_distributed_eval.py and run_eval.py .

@sshleifer sshleifer added the Help wanted Extra attention is needed, help appreciated label Nov 4, 2020
@sshleifer
Copy link
Contributor Author

wdyt @stas00

@stas00
Copy link
Contributor

stas00 commented Nov 4, 2020

I will work on that, thank you.

@stas00
Copy link
Contributor

stas00 commented Nov 4, 2020

A minor correction to the command (corrected data_dir):

python -m torch.distributed.launch --nproc_per_node=2 run_distributed_eval.py --model_name Helsinki-NLP/opus-mt-en-ro --save_dir test_data/opus_wmt_en_ro_gens --data_dir test_data/wmt_en_ro

Question: why only 1 gpu? we currently don't have it tested at all.

@sshleifer
Copy link
Contributor Author

I thought 1 GPU test coverage would be runnable in current CI/by more users.
But if much easier to test 2 gpu/easy to add test for 2 GPU that is great!

@stas00
Copy link
Contributor

stas00 commented Nov 4, 2020

Bottom line - run with as many GPUs as available.

Thank you for clarifying.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Help wanted Extra attention is needed, help appreciated
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants