jy0205/Pyramid-Flow
[ICLR 2025] Pyramidal Flow Matching for Efficient Video Generative Modeling
Leverages autoregressive flow matching to interpolate between multi-resolution latents during generation, enabling simultaneous decompression and video synthesis with a single DiT backbone. Supports flexible inference modes including image-to-video, multi-GPU acceleration, and CPU offloading for sub-8GB memory constraints. Available in miniFLUX and SD3 variants on Hugging Face, trained exclusively on open-source datasets to generate up to 10-second 768p videos at 24 FPS.
3,171 stars. No commits in the last 6 months.
Stars
3,171
Forks
300
Language
Python
License
MIT
Category
Last pushed
Dec 21, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/jy0205/Pyramid-Flow"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
milad1378yz/MOTFM
Flow Matching for Medical Image Synthesis: Bridging the Gap Between Speed and Quality
X-GenGroup/Flow-Factory
A unified framework for easy reinforcement learning in Flow-Matching models
OpenImagingLab/FlashVSR
[CVPR 2026] Towards Real-Time Diffusion-Based Streaming Video Super-Resolution — An efficient...
fallenshock/FlowEdit
Official implementation of the paper: "FlowEdit: Inversion-Free Text-Based Editing Using...
haidog-yaqub/MeanFlow
Pytorch Implementation (unofficial) of the paper "Mean Flows for One-step Generative Modeling"...