split_external_data - helper to break ONNX model external data into mutliple files w/ max size #354
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
rahul-tuli
previously approved these changes
Aug 4, 2023
dbogunowicz
previously approved these changes
Aug 7, 2023
bfineran
dismissed stale reviews from dbogunowicz and rahul-tuli
The base branch was changed.
added 2 commits
August 29, 2023 14:26
…utliple files w/ max size
bfineran
force-pushed
the
split-external-data
branch
from
e1261d3 to
ad52e32
Compare
|
approved pre rebase |
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Outdated suggestions cannot be applied.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.
utility function that breaks ONNX model with a single external data file into multiple external data files with a given max size
internal see doc here: https://github.com/neuralmagic/internal-docs/pull/21
max file size defaults to 16GB
test_plan:
given a plain resnet50 file saved with external data (using
sparseml.utils.onnx.save_onnx) saved tomodel.onnx model.data.onnx)running
split_external_data("model.onnx", 30000000)yieldedmodel was able to compile and running through deepsparse evaluation gave the expected accuracy