UX: configure ZeroGPU to use 'xlarge' hardware tier (96GB VRAM) 2fbae11 akhaliq HF Staff commited on Jun 1
Fix: return single JSON string from @app .api for reliable Gradio Server serialization 1eba2d5 akhaliq HF Staff commited on Jun 1
Fix output display: return plain strings from API + add /api/file endpoint for serving generated files fd5957e akhaliq HF Staff commited on Jun 1
Speed up generation by defaulting to 360p and 3s duration 9cf5f96 akhaliq HF Staff commited on May 31
Handle ZeroGPU startup PyTorch CUDA set_device restrictions in preloading 3302cc2 akhaliq HF Staff commited on May 31
Optimize model preloading on ZeroGPU by initializing pipeline pools at module scope d31983f akhaliq HF Staff commited on May 31
Convert app to gradio.Server with custom obsidian glassmorphic frontend d06a8f3 akhaliq HF Staff commited on May 31
