[fix]: bugfix for RAY_EXPERIMENTAL_NOSET_ASCEND/CUDA_RT_VISIBLE_DEVICES in RL#151
Open
xiazhahe wants to merge 1 commit into
Open
[fix]: bugfix for RAY_EXPERIMENTAL_NOSET_ASCEND/CUDA_RT_VISIBLE_DEVICES in RL#151xiazhahe wants to merge 1 commit into
xiazhahe wants to merge 1 commit into
Conversation
xiazhahe
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
If using a reinforcement learning framework, such as Verl, and setting the global variables
RAY_EXPERIMENTAL_NOSET_ASCEND_RT_VISIBLE_DEVICESorRAY_EXPERIMENTAL_NOSET_CUDA_RT_VISIBLE_DEVICES, the behavior changes from each card only being able to see its own card (i.e., each card is rank 0) to all cards in the device being visible. However,device = get_device_name()implies that tensors are loaded on rank 0 by default, which can easily lead to all cards' tensors being loaded on rank 0, causing an OOM (Out of Memory) error in weight = f.get_tensor(name).Therefore, in this PR, the tensor is loaded from the current device rank obtained by torch instead of rank 0 by default.