pytorch/vision

Classification references does not work without distributed setup

Open

#6.529 aberto em 1 de set. de 2022

Ver no GitHub
 (2 comments) (0 reactions) (0 assignees)Python (15.050 stars) (6.858 forks)batch import
bughelp wantedmodule: reference scripts

Description

If you don't set the respective env vars

https://github.com/pytorch/vision/blob/d5bd8b728f14c33b339fc45c90ca39be339bce3f/references/classification/utils.py#L255-L258

training will not be distributed and in turn the backend will not be initialized. However, during evaluation we check

https://github.com/pytorch/vision/blob/d5bd8b728f14c33b339fc45c90ca39be339bce3f/references/classification/train.py#L88

unguarded, which then fails with

RuntimeError: Default process group has not been initialized, please make sure to call init_process_group.

cc @datumbox

Guia do colaborador