Wrap up the distributed benchmark and get them running on conda_mast
Open
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 1k
- Forks
- 346
- PR merge metrics
- No merged PRs in 30d
Description
We would like to introduce basic distributed benchmarking support on synthetic data.
The idea is to wrap up single-GPU model on DDP/FSDP, then get them running on conda_mast.
The initial OSS distributed userbenchmark could be used as the starting point:
https://github.com/pytorch/benchmark/tree/main/userbenchmark/distributed
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the existing userbenchmark/distributed entry point linked in the issue and review how its synthetic benchmarks are structured. Determine how the single-GPU model should run under DDP/FSDP and on conda_mast. Done means basic distributed benchmarks execute successfully in that environment.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- distributed-systems, machine-learning, performance
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100