StepVideoT2VConfig(model_path: str = '', pipeline_config_path: str | None = None, embedded_cfg_scale: float = 6.0, flow_shift: int = 13, disable_autocast: bool = False, dit_config: DiTConfig = StepVideoConfig(), dit_precision: str = 'bf16', vae_config: VAEConfig = StepVideoVAEConfig(), vae_precision: str = 'bf16', vae_tiling: bool = False, vae_sp: bool = False, image_encoder_config: EncoderConfig = EncoderConfig(), image_encoder_precision: str = 'fp32', text_encoder_configs: tuple[EncoderConfig, ...] = (lambda: (EncoderConfig(),))(), text_encoder_precisions: tuple[str, ...] = (lambda: ('fp32',))(), preprocess_text_funcs: tuple[Callable[[str], str], ...] = (lambda: (preprocess_text,))(), postprocess_text_funcs: tuple[Callable[[BaseEncoderOutput], tensor], ...] = (lambda: (postprocess_text,))(), pos_magic: str = '超高清、HDR 视频、环境光、杜比全景声、画面稳定、流畅动作、逼真的细节、专业级构图、超现实主义、自然、生动、超细节、清晰。', neg_magic: str = '画面暗、低分辨率、不良手、文本、缺少手指、多余的手指、裁剪、低质量、颗粒状、签名、水印、用户名、模糊。', timesteps_scale: bool = False, mask_strategy_file_path: str | None = None, STA_mode: STA_Mode = STA_INFERENCE, skip_time_steps: int = 15, dmd_denoising_steps: list[int] | None = None, ti2v_task: bool = False, boundary_ratio: float | None = None, precision: str = 'bf16')
Bases: PipelineConfig
Base configuration for StepVideo pipeline architecture.