[feat
] Integrate safetensors with Dense, etc. modules too.
#2722
+141
−29
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Outdated suggestions cannot be applied.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.
Hello!
Pull Request overview
Details
This adds the
safe_serialization: bool = True
option to all remaining modules that previously saved and loaded exclusively with"pytorch_model.bin"
. A notable edge case is the LSTM model, which has to be saved oncpu
due to some internal magic thattorch
does when moving a LSTM model to CUDA.cc @philschmid @beurkinger This should allow re-adding
model.safetensors
for the Dense layer in LaBSE once you've moved your dependencies up to the (yet unreleased) 3.0.1. Semi-related: I suspect that moving from 2.* to 3.* will not result in notable issues.