Revert "Revert "[train] TransformersPredictor: Add support for custom pipeline class"" #36705
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
Signed-off-by: Kai Fricke <kai@anyscale.com>
Yard1
approved these changes
Jun 23, 2023
8 tasks
arvind-chandra
pushed a commit
to lmco/ray
that referenced
this pull request
Aug 31, 2023
… pipeline class"" (ray-project#36705) Reverts ray-project#36701 This re-activates the changes in ray-project#36494 which were generally working. The problem was that an import of `TFPreTrainedModel` on a GPU instance seems to initialize the GPU and make it unusable by Ray workers, so that CUDA memory allocations fail. Thus, imports of TF modules should be guarded behind the TYPE_CHECKING variable: ``` if TYPE_CHECKING: # ... from transformers.modeling_utils import PreTrainedModel from transformers.modeling_tf_utils import TFPreTrainedModel ``` Signed-off-by: Kai Fricke <kai@anyscale.com> Signed-off-by: e428265 <arvind.chandramouli@lmco.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in 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Outdated suggestions cannot be applied.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.
Reverts #36701
This re-activates the changes in #36494 which were generally working. The problem was that an import of
TFPreTrainedModelon a GPU instance seems to initialize the GPU and make it unusable by Ray workers, so that CUDA memory allocations fail.Thus, imports of TF modules should be guarded behind the TYPE_CHECKING variable: