FastVideo Releases FastH3 V1: 4-Step Sparse Distilled LoRA for Minimax H3
Solitary_Thinker · reddit · 2026-08-29
FastVideo team released FastH3 V1, a 4-step sparse distilled checkpoint and LoRA for the Minimax H3 model. Key features include:
- Performance: Trained on 1000+ B200 hours, supporting variable resolution, aspect ratio, and duration.
- Open Source: Full weights and pre-extracted LoRA, with dense attention and synthetic data ablations available. Requires VSA kernel.
- Roadmap: Optimizations for local hardware (RTX, Apple MLX) and integration of Nvidia's Parallel Decoding Distillation (PDD).
More from Multimodal
- Krea 3 beta launches: 2K edits in under 8s, claims Nano Banana Pro-level quality — Scobleizer · 2026-10-16
- FLUX 3 Image Klein lands on fal: box-drawn layouts, multi-reference placement, style transfer — bfl_ai · 2026-10-16
- SGLang ships Day-0 support for FLUX 3 Image Klein: 10 reference images, BF16/FP8/NVFP4 variants — ying11231 · 2026-10-16
- Higgsfield teases Kling 4.0 rollout with a monkey-driven demo video — azed_ai · 2026-10-16
- Halcyon early-fusion image models match late-fusion with 275x less pretraining compute, open-sourced — davidwromero · 2026-10-16
- Running Wan2.1 14B FP16 on a 12GB laptop GPU: VRAM swap stress test finishes in 4m11s — RYuu-1120 · 2026-10-16