When I set num_worker=8, the following error will occurs
Open
- Dominant language
- Python
- Stars
- 799
- Forks
- 111
- PR merge metrics
- No merged PRs in 30d
Description
When I try to run the pretrain.py file, and set the work=8, the following error will appear. The training is very slow when the worker=4, the gpu utilization is very low, How do I solve it? thanks
Contributor guide
No contributing guide indexed for this repository
Research direction
Start with pretrain.py and inspect how the worker setting is handled. Reproduce the reported behavior with worker values of 4 and 8, using the error shown in the attached image; done means the error is resolved and training throughput or GPU utilization improves.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- machine-learning, performance
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100