Skip to content

Comments

Update train.py#25

Open
Miralan wants to merge 2 commits intojik876:masterfrom
Miralan:master
Open

Update train.py#25
Miralan wants to merge 2 commits intojik876:masterfrom
Miralan:master

Conversation

@Miralan
Copy link

@Miralan Miralan commented Dec 7, 2020

Solve imbalanced gpu memory at multi-gpu distributed training.

Solve imbalanced gpu memory at multi-gpu distributed training.
@v-nhandt21
Copy link

@Miralan Did you remember running time you use for distributed, second/iteration, Can you share the script for run distibuted

shivahr added a commit to bhashini-ai/hifi-gan that referenced this pull request Jan 14, 2022
@Aloento
Copy link

Aloento commented Mar 14, 2023

Why and how did you make this change?

@PussyCat0700
Copy link

Amazing! The line torch.cuda.set_device(rank) gives significant boost especially when training just started. Anyone knows why?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants