Scaling Mixture-of-Experts Video Pretraining for Embodied Intelligence
-
Updated
Sep 17, 2026 - Python
Scaling Mixture-of-Experts Video Pretraining for Embodied Intelligence
Selective fused FP8 checkpoint + ComfyUI runtime for LingBot-Video Dense 1.3B. Measured 1.27x in a local one-step RTX 5080 sampler smoke test; T2V, TI2V, experimental FLF.
Community ComfyUI toolkit for LingBot-Video: structured prompt directors, low-VRAM GGUF inference, diagnostics, and experimental edit/conditioning nodes.
To associate your repository with the lingbot-video topic, visit your repo's landing page and select "manage topics."