o ٷi|@s^dZddlmZmZddlmZddlmZddlm Z GdddeZ Gdd d ej Z dS) a Base cache backend interface for diffusion models. This module defines the abstract base class that all cache backends must implement. Cache backends provide a unified interface for applying different caching strategies to transformer models. Main cache backend implementations: 1. CacheDiTBackend: Implements cache-dit acceleration (DBCache, SCM, TaylorSeer) using the cache-dit library. Inherits from CacheBackend. Used via cache_backend="cache_dit". 2. TeaCacheBackend: Hook-based backend for TeaCache acceleration. Inherits from CacheBackend. Used via cache_backend="tea_cache". All backends implement the same interface: - enable(pipeline): Enable cache on the pipeline - refresh(pipeline, num_inference_steps, verbose): Refresh cache state - is_enabled(): Check if cache is enabled )ABCabstractmethod)AnyN)DiffusionCacheConfigc @speZdZdZdefddZededdfdd Zedded e d e ddfd dZ de fddZ de fddZdS) CacheBackenda Abstract base class for cache backends. All cache backend implementations (CacheDiTBackend, TeaCacheBackend, etc.) inherit from this base class and implement the enable() and refresh() methods to manage cache lifecycle. Cache backends apply caching strategies to transformer models to accelerate inference. Different backends use different underlying mechanisms (e.g., cache-dit library for CacheDiTBackend, hooks for TeaCacheBackend), but all share the same unified interface. Attributes: config: DiffusionCacheConfig instance containing cache-specific configuration parameters enabled: Boolean flag indicating whether cache is enabled (set to True after enable() is called) configcCs||_d|_dS)z Initialize cache backend with configuration. Args: config: DiffusionCacheConfig instance with cache-specific parameters FN)renabled)selfrr R/home/ubuntu/.local/lib/python3.10/site-packages/vllm_omni/diffusion/cache/base.py__init__1s zCacheBackend.__init__pipelinereturnNcCtd)a Enable cache on the pipeline. This method applies the caching strategy to the transformer(s) in the pipeline. The specific implementation depends on the backend (e.g., hooks for TeaCacheBackend, cache-dit library for CacheDiTBackend). Called once during pipeline initialization. Args: pipeline: Diffusion pipeline instance. The backend can extract: - transformer: via pipeline.transformer - model_type: via pipeline.__class__.__name__ z"Subclasses must implement enable()NotImplementedError)r r r r r enable;szCacheBackend.enableTnum_inference_stepsverbosecCr)aQ Refresh cache state for new generation. This method should clear any cached values and reset counters/accumulators. Called at the start of each generation to ensure clean state. Args: pipeline: Diffusion pipeline instance. The backend can extract: - transformer: via pipeline.transformer num_inference_steps: Number of inference steps for the current generation. May be used for cache context updates. verbose: Whether to log refresh operations (default: True) z#Subclasses must implement refresh()r)r r rrr r r refreshKszCacheBackend.refreshcCs|jS)z Check if cache is enabled on this backend. Returns: True if cache is enabled, False otherwise. )rr r r r is_enabled\szCacheBackend.is_enabledcCs|jjd|jdS)Nz(config=)) __class____name__rrr r r __repr__eszCacheBackend.__repr__T)r __module__ __qualname____doc__rr rrrintboolrrstrrr r r r rs  rcs0eZdZfddZddeffdd ZZS)CachedTransformerc std|_dS)NF)superr do_true_cfg)r kwargsrr r r js  zCachedTransformer.__init__Tenable_separate_cfgc s||_tjdi|dS)Nr )r(r$__init_subclass__)clsr(r&r'r r r)nsz#CachedTransformer.__init_subclass__r)rrrr r!r) __classcell__r r r'r r#is r#) rabcrrtypingrtorch.nnnnvllm_omni.diffusion.datarrModuler#r r r r s   J