Skip to content

vllm_omni.diffusion.models.seedvr2.long_video

One asynchronous SeedVR2 request, bounded model windows, one MP4 writer.

FPS module-attribute 

FPS = 24

JOB_ID module-attribute 

JOB_ID = re.compile('[0-9a-f]{32}')

LOSSLESS module-attribute 

LOSSLESS = {'preset': 'ultrafast', 'qp': '0'}

MAX_FRAMES module-attribute 

MAX_FRAMES = 7200

OVERLAP module-attribute 

OVERLAP = 4

UPLOAD_CHUNK module-attribute 

UPLOAD_CHUNK = 1 << 20

active_task module-attribute 

active_task: Task[None] | None = None

job_lock module-attribute 

job_lock = asyncio.Lock()

logger module-attribute 

logger = logging.getLogger(__name__)

router module-attribute 

router = APIRouter()

JobCancelledError

Bases: JobError

The client asked the job to stop.

JobError

Bases: Exception

An error whose message is safe to return to the client.

cancel_long_video async 

cancel_long_video(job_id: str) -> dict[str, str]

Ask the running job to stop; it settles at the next window boundary.

create_long_video async 

create_long_video(
    raw_request: Request,
    input_references: UploadFile = File(...),
    num_frames: int = Form(...),
    size: str = Form(...),
    loop_input: bool = Form(False),
    prompt: str = Form(" "),
    seed: int = Form(7723),
    extra_params: str | None = Form(None),
) -> dict[str, str | int]

download_long_video async 

download_long_video(job_id: str) -> FileResponse

get_long_video async 

get_long_video(job_id: str) -> dict[str, str | int]

register_routes 

register_routes(app: FastAPI, port: int) -> None

Attach this model's optional long-video API to its serving process.