diffusers-internal-dev/abot-world-interactive
Pin diffusers to abot-world-modular@206121e08 (tiny VAE decode in the streaming preset)
Pin diffusers to abot-world-modular@206121e08 (tiny VAE decode in the streaming preset)
Pin diffusers to abot-world-modular@5a29766dd (WanDecodeCache streaming decode)
Raise per-session ZeroGPU budget to 150 s
Disable SSR: the Node proxy does not forward the /ws WebSocket upgrade
Live front-end: gradio.Server + WebSocket paced frame stream (shell from acvlab/abot-world-interactive), engine = pipe.stream(action_source=...)
Give each block 6 s of ZeroGPU duration budget (measured ~5 s/block)
Pin diffusers to abot-world-modular@ffee5d2d9 (bust the cached pip layer)
Run on ZeroGPU: spaces.GPU session generator + command queue for controls
Barebone interactive ABot-World on Modular Diffusers: one pipe.stream(action_source=...) loop
initial commit
