facebookresearch/fairseq

Add support for OpenMPI Rank Assignment

Open

#4,645 创建于 2022年8月11日

在 GitHub 查看
 (0 评论) (0 反应) (0 负责人)Python (6,224 fork)batch import
enhancementhelp wantedneeds triage

仓库指标

Star
 (29,107 star)
PR 合并指标
 (30 天内没有已合并 PR)

描述

🚀 Feature Request

Currently, there is no way to read the variables OMPI_COMM_WORLD_SIZE, OMPI_COMM_WORLD_RANK, and OMPI_COMM_WORLD_LOCAL_RANK to assign ranks to any individual processes launched with OpenMPI

Motivation

I'd like to launch jobs with OpenMPI without using multiple-data mode, which can cause issues with command length if I have to specify all of the ranks manually.

Pitch

Add _infer_ompi_init function in fairseq/distributed/utils.py which reads the environment variables:

MASTER_ADDR
MASTER_PORT
OMPI_COMM_WORLD_SIZE
OMPI_COMM_WORLD_RANK
OMPI_COMM_WORLD_LOCAL_RANK

and updates the distributed config accordingly.

Alternatives

N/A

Additional context

N/A

贡献者指南