Repository navigation
[quantization] Quantize Gemma4 PLE arithmetic for Circle export - #1076
Merged
mhs4670go merged 3 commits intoOct 22, 2026
Merged
Conversation
Observe the projection-plus-token ADD output and quantize the two PLE multiplication scale constants in both prefill and decode. Reuse the frozen model observers at export boundaries and preserve the existing final per-layer-input quantization contract. Reject legacy PLE checkpoints without the required calibration state. Add focused Circle dtype/qparam regression checks and update the Gemma4 export documentation. TICO-DCO-1.0-Signed-off-by: seongwoo chae <mhs4670go@naver.com>
Apply the UFMT blank-line changes reported by Check format CI. TICO-DCO-1.0-Signed-off-by: seongwoo chae <mhs4670go@naver.com>
Allow the quantized Circle regression test to pass its activation quantization configuration to the tiny Gemma4 text model. TICO-DCO-1.0-Signed-off-by: seongwoo chae <mhs4670go@naver.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Observe the projection-plus-token ADD output and quantize the two PLE multiplication scale constants in both prefill and decode. Reuse the frozen model observers at export boundaries and preserve the existing final per-layer-input quantization contract.
Reject legacy PLE checkpoints without the required calibration state. Add focused Circle dtype/qparam regression checks and update the Gemma4 export documentation.
TICO-DCO-1.0-Signed-off-by: seongwoo chae mhs4670go@naver.com