LongLive-Plug: Once-for-All Distillation for Video Generation

Hugging Face Daily Papers Papers
video-generation distillation lora once-for-all few-step-sampling classifier-free-guidance autoregressive-generation

Summary

LongLive-Plug is a once-for-all distillation framework that learns reusable LoRA adapters for video generation models, enabling training-free, plug-and-play deployment across various tasks without per-target retraining.

Video diffusion models are increasingly developed into specialized models for diverse downstream tasks, and this development often includes a distillation stage, for example to accelerate sampling or to improve long-video generation. This stage is typically repeated for every specialized model. We introduce LongLive-Plug, a once-for-all distillation framework that learns reusable capabilities as LoRAs on a base model for training-free, plug-and-play deployment to compatible downstream models. These capabilities include single-pass classifier-free guidance, few-step sampling, and long-context error correction for autoregressive generation. The adapters remain reusable even when downstream models add conditioning branches, expand output channels. Despite training at a fixed guidance scale, our dedicated CFG LoRA provides text guidance control through its inference weight. Combining it with a few-step LoRA simultaneously preserves few-step generation and CFG controllability on downstream tasks. We verify training-free deployment on 54 downstream models across three backbone families and eight task categories, including world modeling, robotics, editing, and multimodal generation. The approach may support additional compatible models. Each capability can thus be distilled once per backbone family and reused without per-target retraining.
Original Article
View Cached Full Text

Cached at: 09/30/26, 04:18 AM

Paper page - LongLive-Plug: Once-for-All Distillation for Video Generation

Source: https://huggingface.co/papers/2609.38154 Published on Sep 29

·

Submitted byhttps://huggingface.co/Andyson

Syangon Sep 30

Authors:

,

,

,

,

,

,

,

,

,

,

Abstract

Videodiffusionmodelsareincreasinglydevelopedintospecializedmodelsfordiversedownstreamtasks,andthisdevelopmentoftenincludesadistillationstage,forexampletoacceleratesamplingortoimprovelong-videogeneration.Thisstageistypicallyrepeatedforeveryspecializedmodel.WeintroduceLongLive-Plug,aonce-for-alldistillationframeworkthatlearnsreusablecapabilitiesasLoRAsonabasemodelfortraining-free,plug-and-playdeploymenttocompatibledownstreammodels.Thesecapabilitiesincludesingle-passclassifier-freeguidance,few-stepsampling,andlong-contexterrorcorrectionforautoregressivegeneration.Theadaptersremainreusableevenwhendownstreammodelsaddconditioningbranches,expandoutputchannels.Despitetrainingatafixedguidancescale,ourdedicatedCFGLoRAprovidestextguidancecontrolthroughitsinferenceweight.Combiningitwithafew-stepLoRAsimultaneouslypreservesfew-stepgenerationandCFGcontrollabilityondownstreamtasks.Weverifytraining-freedeploymenton54downstreammodelsacrossthreebackbonefamiliesandeighttaskcategories,includingworldmodeling,robotics,editing,andmultimodalgeneration.Theapproachmaysupportadditionalcompatiblemodels.Eachcapabilitycanthusbedistilledonceperbackbonefamilyandreusedwithoutper-targetretraining.

View arXiv pageView PDFProject pageGitHub2.64kAdd to collection

Get this paper in your agent:

hf papers read 2609\.38154

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2609.38154 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.38154 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.38154 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Real-Time Long Video Generation (GitHub Repo)

TLDR AI

NVlabs releases LongLive 2.0, a parallel infrastructure for real-time long video generation using NVFP4 quantization, supporting both training and inference. It achieves 45.7 FPS and is accepted at ICLR 2026.

Video2LoRA: Parametric Video Internalization for Vision-Language Models

Hugging Face Daily Papers

This paper introduces Video2LoRA, a method that predicts Low-Rank Adaptation (LoRA) weights directly from video representations, enabling efficient video processing in frozen vision-language models. It reduces visual token load by up to 1500x and query TTFT by 6-80x while maintaining performance on video summarization and captioning benchmarks.