Update multi-gpu notebook to set cupy device #675
Merged
Conversation
This file contains 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
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
Documentation previewhttps://nvidia-merlin.github.io/Transformers4Rec/review/pr-675 |
marcromeyn
approved these changes
Apr 18, 2023
# for free
to join this conversation on GitHub.
Already have an account?
# to comment
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Outdated suggestions cannot be applied.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.
Depends on NVIDIA-Merlin/dataloader#135.
Goals ⚽
Update the multi-gpu training notebook to work with the new version of the Merlin dataloader.
Implementation Details 🚧
The new dataloader keeps the data in cupy arrays and converts each batch to torch tensors at the last minute, while the old dataloader converted everything to torch tensors as soon as the dataset was loaded from dask-cudf. Since we are using cupy in the dataloader now, we have to ensure that cupy uses the correct device by using
cupy.cuda.Device. This should ideally be set inmerlin.io.Datasetso users don't have to set this, but we add this to the notebook as a short-term workaround.Testing Details 🔍