[ET-VK][quantization] Implement layout-flexible quantize/dequantize operators#17261
Merged
SS-JIA merged 1 commit intogh/SS-JIA/405/origfrom Feb 5, 2026
Merged
[ET-VK][quantization] Implement layout-flexible quantize/dequantize operators#17261SS-JIA merged 1 commit intogh/SS-JIA/405/origfrom
SS-JIA merged 1 commit intogh/SS-JIA/405/origfrom
Conversation
…perators Pull Request resolved: #17106 Implemented quantize_per_tensor and dequantize_per_tensor GLSL shaders and C++ dispatch logic to support the new single-dimension packed INT8 layouts (kPackedInt8_4W, kPackedInt8_4C, kPackedInt8_4H). These operators enable conversion between floating-point tensors and packed int8 representations with per-tensor scale and zero-point parameters. The implementation includes: - GLSL shaders: quantize_per_tensor and dequantize_per_tensor with support for both texture->buffer and buffer->buffer data flows, including GL_EXT_debug_printf statements for debugging - QuantizeDequantize.cpp: Added dispatch functions for the new layouts and registered etvk.q_dq_8bit_per_tensor.default operator - Test infrastructure: Created q_dq_8bit_per_tensor test binary with DEBUG_MODE support and reference CPU implementation for validation The shaders implement the quantization formula ``` Q = clamp(round(x/scale) + zp, -128, 127) ``` and dequantization formula ``` x' = (Q - zp) \* scale ``` with proper int8 packing/unpacking using little-endian byte ordering and sign extension. ghstack-source-id: 338638544 @exported-using-ghexport Differential Revision: [D92061370](https://our.internmc.facebook.com/intern/diff/D92061370/)
pytorchbot
requested review from
SS-JIA,
kirklandsign and
larryliu0820
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/17261
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New Failure, 76 Pending, 1 Unrelated FailureAs of commit 1781382 with merge base 1cffd23 ( NEW FAILURE - The following job has failed:
FLAKY - The following job failed but was likely due to flakiness present on trunk:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
SS-JIA
approved these changes
Feb 5, 2026
SS-JIA
pushed a commit
that referenced
this pull request
Feb 6, 2026
…perators (#17261) Implemented quantize_per_tensor and dequantize_per_tensor GLSL shaders and C++ dispatch logic to support the new single-dimension packed INT8 layouts (kPackedInt8_4W, kPackedInt8_4C, kPackedInt8_4H). These operators enable conversion between floating-point tensors and packed int8 representations with per-tensor scale and zero-point parameters. The implementation includes: - GLSL shaders: quantize_per_tensor and dequantize_per_tensor with support for both texture->buffer and buffer->buffer data flows, including GL_EXT_debug_printf statements for debugging - QuantizeDequantize.cpp: Added dispatch functions for the new layouts and registered etvk.q_dq_8bit_per_tensor.default operator - Test infrastructure: Created q_dq_8bit_per_tensor test binary with DEBUG_MODE support and reference CPU implementation for validation The shaders implement the quantization formula Q = clamp(round(x/scale) + zp, -128, 127) and dequantization formula x' = (Q - zp) * scale, with proper int8 packing/unpacking using little-endian byte ordering and sign extension. Differential Revision: [D92061370](https://our.internmc.facebook.com/intern/diff/D92061370/) [ghstack-poisoned]
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
This PR was created by the merge bot to help merge the original PR into the main branch.
ghstack PR number: #17106 by @SS-JIA
^ Please use this as the source of truth for the PR details, comments, and reviews
ghstack PR base: https://github.com/pytorch/executorch/tree/gh/SS-JIA/399/base
ghstack PR head: https://github.com/pytorch/executorch/tree/gh/SS-JIA/399/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/gh/SS-JIA/405/orig
Merge bot PR head: https://github.com/pytorch/executorch/tree/gh/SS-JIA/399/orig
Differential Revision: D92061370
@diff-train-skip-merge