Reinforcing Few-step Generators via Reward-Tilted Distribution Matching
🔄 In a Training Loop
Yushi Huang
Harahan
AI & ML interests
Efficient AIGC, especially for video/image generation and MLLMs
Recent Activity
authored a paper 23 days ago
Flash-VAED: Plug-and-Play VAE Decoders for Efficient Video Generation authored a paper 23 days ago
Real-Time Human Frontal View Synthesis from a Single Image authored a paper 23 days ago
Salt: Self-Consistent Distribution Matching with Cache-Aware Training for Fast Video GenerationOrganizations
None yet
models 6
Harahan/FLUX1-12B-RTDMD
Text-to-Image • Updated • 2
Harahan/FLUX2-9B-RTDMD
Text-to-Image • Updated • 1
Harahan/FLUX2-4B-RTDMD
Text-to-Image • Updated • 3
Harahan/SD35M-RTDMD
Text-to-Image • Updated • 1
Harahan/QVGen-CogVideoX-2B-W4A4
Text-to-Video • Updated • 17 • 2
Harahan/QVGen-Wan-1_3B-W4A4
Text-to-Video • Updated • 3 • 2
datasets 0
None public yet