Repository navigation
Faster Nonlocal denoise: GEMM ONNX graph, halo 40, remove Maximum - #19
Merged
Merged
Conversation
…ality - Derive the ONNX model with gemm-1x1-v2 (drop zero-bias nodes, run 1x1 convolutions as matrix multiplications): 370 -> 153 ms per tile on CUDA, same frozen fixture gates. - Tile halo 64 -> 40 (150 tiles instead of 247 for 32.5 MP); a single accepted model hash with 320/40/240 tiling. - Remove the Maximum / four-rotation option and the quality parameter from the native pipeline, MCP schema, Python tooling, docs and skills. - Pin the republished Hugging Face revision (nonlocal-v2) and add an ONNX install test against it. - Includes pending MCP session-storage and render/session edits. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
gemm-1x1-v2): 370 -> 153 ms per 320x320 tile on an RTX 5060 Ti, same frozen fixture gates on CUDA and CPU.nonlocal-v2) and pinned. Older installs are rejected until reinstalled withinstall_model kind=nonlocal.qualityeverywhere (breaking, no compatibility shim).Testing
cargo fmt --check,cargo clippy --all-targets --all-features -- -D warnings,cargo test --lib(224 passed)🤖 Generated with Claude Code