Skip to content

vllm_omni.diffusion.models.minimax_h3.lora

Loader for the LightX2V Turbo MiniMax-H3 LoRA family.

The published artifacts differ in task family, step count, training resolution and LoRA alpha, so :class:TurboSpec carries the contract of the file actually loaded and the pipeline validates each request against it. Only the Diffusers-PEFT exports are served. The native FlashGen contract lives in .npu.lora.

logger module-attribute 

logger = init_logger(__name__)

TurboSpec dataclass

The sampler contract one Turbo artifact was distilled for.

alpha instance-attribute 

alpha: float

audio_shift instance-attribute 

audio_shift: float

denoise_steps instance-attribute 

denoise_steps: int

filename instance-attribute 

filename: str

rank instance-attribute 

rank: int

sigma_points property 

sigma_points: int

Sigma points the API contract expects: one more than the forwards.

supported_tasks property 

supported_tasks: frozenset[str]

task_family instance-attribute 

task_family: str

fl2v (serves t2va and fl2va) or ref2v (serves ref2va).

video_shift instance-attribute 

video_shift: float

load_minimax_h3_turbo_lora 

load_minimax_h3_turbo_lora(
    *,
    partition: str,
    lora_request: LoRARequest,
    lora_path: str | Path,
    dtype: dtype,
    unsupported_offload_mode: str | None = None,
) -> tuple[LoRAModel, PEFTHelper, TurboSpec] | None

Load any published LightX2V Turbo artifact through the legacy manager.

parse_turbo_filename 

parse_turbo_filename(name: str) -> TurboSpec | None

Return the contract a Turbo filename encodes, or None.

alpha is the rank-matched default; the loader replaces it with the artifact's declared value.