Fix device in rope module when using dynamic updates #35608
Merged
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Outdated suggestions cannot be applied.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.
What does this PR do?
If a model is initialized on cpu, then moved with
.cuda()
or.to(device)
, the parameters and buffers are automatically moved, but not the other internal tensors. This can lead to wrong device when going back toself.original_inv_freq
in the rope module as the model was originally on cpu, thusself.original_inv_freq
is still on cpu.This fixes it.