Popular repositories Loading
-
Repeated-runs-become-slower-due-to-VRAM-offload-regression-with-DynamicVRAM-leak-spill-
Repeated-runs-become-slower-due-to-VRAM-offload-regression-with-DynamicVRAM-leak-spill- PublicObserved during normal image generation. In our case, ComfyUI-N-Nodes / RIFE appears to keep GPU memory outside ComfyUI's model manager. Under heavier VRAM pressure, later runs can slow dramaticall…
-
SDXL-16GB-AMD-GPU-VAE-VRAM-Sampler-Performance
SDXL-16GB-AMD-GPU-VAE-VRAM-Sampler-Performance PublicPractical SDXL testing on RX 9060 XT 16 GB with ComfyUI/ROCm: batch scaling, VRAM headroom, VAE behavior, multiple resolutions, sampler/step behavior, throughput and long-run performance.
Something went wrong, please refresh the page to try again.
If the problem persists, check the GitHub status page or contact support.
If the problem persists, check the GitHub status page or contact support.