Arm backend: Enable test_llama_tosa_BI and related fixes #10681
Merged
Conversation
This file contains hidden or 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
First problem solved by adding quantize of scalar_tensor:
The where.self operator got a scalar_tensor input which
was not quantized. This happened since the where.self quantization
annotator uses the parent specs, which in this case where
non-existing. Adding the quantization of scalar_tensor sorts this out.
Secondly when quantizing scalar_tensor the following assert triggers:
expecting kwargs for aten op IR to be empty
Hence setting scalar_tensor kwargs to {}.
Finally trying to quantize -inf fails for scalar_tensor nodes fails
Fix it by adding the pass from qnn backend to replace -inf/inf.
Hence adding new pass ReplaceInfValues.
Co-authored-by: Per Åstrand <[email protected]>
Co-authored-by: Måns Nilsson <[email protected]>
Change-Id: I83b10c13de14fefcc7fd70108927f6b79f326ed8
mansnils
added
partner: arm
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/10681
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ⏳ No Failures, 2 PendingAs of commit 0cdb41c with merge base cd3b53d ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
martinlsm
approved these changes
May 5, 2025
jhelsby
pushed a commit
to jhelsby/executorch
that referenced
this pull request
May 9, 2025
First problem solved by adding quantize of scalar_tensor: The where.self
operator got a scalar_tensor input which was not quantized. This
happened since the where.self quantization annotator uses the parent
specs, which in this case where non-existing. Adding the quantization of
scalar_tensor sorts this out.
Secondly when quantizing scalar_tensor the following assert triggers:
expecting kwargs for aten op IR to be empty
Hence setting scalar_tensor kwargs to {}.
Finally trying to quantize -inf fails for scalar_tensor nodes fails Fix
it by adding the pass from qnn backend to replace -inf/inf. Hence adding
new pass ReplaceInfValues.
Co-authored-by: Per Åstrand <[email protected]>
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.
First problem solved by adding quantize of scalar_tensor: The where.self operator got a scalar_tensor input which was not quantized. This happened since the where.self quantization annotator uses the parent specs, which in this case where non-existing. Adding the quantization of scalar_tensor sorts this out.
Secondly when quantizing scalar_tensor the following assert triggers:
expecting kwargs for aten op IR to be empty
Hence setting scalar_tensor kwargs to {}.
Finally trying to quantize -inf fails for scalar_tensor nodes fails Fix it by adding the pass from qnn backend to replace -inf/inf. Hence adding new pass ReplaceInfValues.
Co-authored-by: Per Åstrand [email protected]
Co-authored-by: Måns Nilsson [email protected]
cc @digantdesai @freddan80 @per @zingo @oscarandersson8218