[Model] Add LongCat-Image-Edit support (#392)

Signed-off-by: huangxulong <huangxulong@oschina.cn>
Signed-off-by: elijah <f1renze.142857@gmail.com>
Co-authored-by: huangxulong <huangxulong@oschina.cn>
This commit is contained in:
elijahandhuangxulong authored and GitHub committed 2025-12-22 21:31:50 +08:00
1 parent f363108425
commit ef827022b8
7 files changed
+843 -261

No files matched your search

+73
View File
@@ -186,6 +186,77 @@ def enable_cache_for_wan22(pipeline: Any, cache_config: Any) -> Callable[[int],
return refresh_cache_context
def enable_cache_for_longcat_image(pipeline: Any, cache_config: Any) -> Callable[[int], None]:
"""Enable cache-dit for LongCatImage pipeline.
Args:
pipeline: The LongCatImage pipeline instance.
cache_config: DiffusionCacheConfig instance with cache configuration.
"""
# Build DBCacheConfig for transformer
db_cache_config = _build_db_cache_config(cache_config)
calibrator = None
if cache_config.enable_taylorseer:
taylorseer_order = cache_config.taylorseer_order
calibrator = TaylorSeerCalibratorConfig(taylorseer_order=taylorseer_order)
logger.info(f"TaylorSeer enabled with order={taylorseer_order}")
# Build ParamsModifier for transformer
modifier = ParamsModifier(
cache_config=db_cache_config,
calibrator_config=calibrator,
)
logger.info(
f"Enabling cache-dit on LongCatImage transformer with BlockAdapter: "
f"Fn={db_cache_config.Fn_compute_blocks}, "
f"Bn={db_cache_config.Bn_compute_blocks}, "
f"W={db_cache_config.max_warmup_steps}, "
)
# Enable cache-dit using BlockAdapter for transformer
cache_dit.enable_cache(
(
BlockAdapter(
transformer=pipeline.transformer,
blocks=[
pipeline.transformer.transformer_blocks,
pipeline.transformer.single_transformer_blocks,
],
forward_pattern=[ForwardPattern.Pattern_1, ForwardPattern.Pattern_1],
params_modifiers=[modifier],
)
),
cache_config=db_cache_config,
)
def refresh_cache_context(pipeline: Any, num_inference_steps: int, verbose: bool = True) -> None:
"""Refresh cache context for the transformer with new num_inference_steps.
Args:
pipeline: The LongCatImage pipeline instance.
num_inference_steps: New number of inference steps.
"""
if cache_config.scm_steps_mask_policy is None:
cache_dit.refresh_context(pipeline.transformer, num_inference_steps=num_inference_steps, verbose=verbose)
else:
cache_dit.refresh_context(
pipeline.transformer,
cache_config=DBCacheConfig().reset(
num_inference_steps=num_inference_steps,
steps_computation_mask=cache_dit.steps_mask(
mask_policy=cache_config.scm_steps_mask_policy,
total_steps=num_inference_steps,
),
steps_computation_policy=cache_config.scm_steps_policy,
),
verbose=verbose,
)
return refresh_cache_context
def enable_cache_for_flux(pipeline: Any, cache_config: Any) -> Callable[[int], None]:
"""Enable cache-dit for Flux dual-transformer architecture.
@@ -267,6 +338,8 @@ CUSTOM_DIT_ENABLERS.update(
{
"WanPipeline": enable_cache_for_wan22,
"FluxPipeline": enable_cache_for_flux,
"LongcatImagePipeline": enable_cache_for_longcat_image,
"LongcatImageEditPipeline": enable_cache_for_longcat_image,
}
)