Enable TaylorSeer caching by default for Flux and SD3.5#3709
Open
l-bat wants to merge 3 commits intoopenvinotoolkit:masterfrom
Open
Enable TaylorSeer caching by default for Flux and SD3.5#3709l-bat wants to merge 3 commits intoopenvinotoolkit:masterfrom
l-bat wants to merge 3 commits intoopenvinotoolkit:masterfrom
Conversation
l-bat
requested review from
likholat,
sbalandi,
sgonorov and
yatarkan
as code owners
github-actions
Bot
added
category: llm_bench
Contributor
There was a problem hiding this comment.
Pull request overview
Enables TaylorSeer caching by default for Flux and StableDiffusion3 Text2Image pipelines, and updates tooling/tests/docs so benchmarks and samples can still run baseline (no-caching) comparisons.
Changes:
- Turn on
taylorseer_configby default in Flux and StableDiffusion3 Text2Image pipeline default generation configs. - Update benchmarks/tests to explicitly disable caching where baseline behavior is required, and add a Python test asserting the new default.
- Refresh docs/samples to reflect the new default behavior and how to disable/customize caching.
Reviewed changes
Copilot reviewed 11 out of 11 changed files in this pull request and generated 3 comments.
Show a summary per file
| File | Description |
|---|---|
| tools/who_what_benchmark/tests/test_cli_image.py | Passes --taylorseer-config to disable caching for baseline-like CLI runs. |
| tools/llm_bench/README.md | Documents default-on caching for Flux/SD3 Text2Image and how to disable via CLI config. |
| tests/python_tests/test_image_generation.py | Adds SD3 tiny-random model constant + test asserting TaylorSeer default is enabled. |
| tests/python_tests/conftest.py | Adds tiny-random-sd3 model mapping for conversion/fixture use. |
| src/cpp/src/image_generation/stable_diffusion_3_pipeline.hpp | Enables default TaylorSeer config for SD3 Text2Image pipeline type. |
| src/cpp/src/image_generation/flux_pipeline.hpp | Enables default TaylorSeer config for Flux Text2Image pipeline type. |
| site/docs/concepts/optimization-techniques/diffusion-caching.md | Updates TaylorSeer docs to state default enablement and shows disable/customize patterns. |
| samples/python/image_generation/taylorseer_text2image.py | Explicitly disables caching for baseline image generation before enabling it for comparison. |
| samples/python/image_generation/README.md | Notes default enablement and updates baseline vs caching description. |
| samples/cpp/image_generation/taylorseer_text2image.cpp | Explicitly disables caching for baseline run; reuses config variable for cached run. |
| samples/cpp/image_generation/README.md | Notes default enablement and updates baseline vs caching description. |
l-bat
force-pushed
the
lt/enable_ts_imggen
branch
from
b69c6ca to
2aa4fa6
Compare
l-bat
force-pushed
the
lt/enable_ts_imggen
branch
from
2aa4fa6 to
51c4424
Compare
likholat
approved these changes
Apr 16, 2026
l-bat
force-pushed
the
lt/enable_ts_imggen
branch
from
51c4424 to
06eeb57
Compare
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.
Description
CVS-182423
Documentation: https://l-bat.github.io/openvino.genai/docs/concepts/optimization-techniques/diffusion-caching#image-generation-flux--stablediffusion3
GHA: https://github.com/l-bat/openvino.genai/actions/runs/24449838966
Checklist: