Skip to content

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.

Improve some warning messages. (#15977)

comfyanonymous8a33128f2f8c
diff --git a/main.py b/main.py
index 50222cc5c..bf0ea7343 100644
--- a/main.py
+++ b/main.py
@@ -271,7 +271,7 @@ def dynamic_vram_supported():
if args.enable_dynamic_vram or (enables_dynamic_vram() and dynamic_vram_supported()):
if (not args.enable_dynamic_vram) and (comfy.model_management.torch_version_numeric < (2, 8)):
- logging.warning("Unsupported Pytorch detected. DynamicVRAM support requires Pytorch version 2.8 or later. Falling back to legacy ModelPatcher. VRAM estimates may be unreliable especially on Windows")
+ logging.warning("Unsupported Pytorch detected. DynamicVRAM support requires Pytorch version 2.8 or later (2.12+ is recommended). Falling back to legacy ModelPatcher. VRAM estimates may be unreliable especially on Windows")
else:
try:
aimdo_initialized = comfy_aimdo.control.init_devices((d.index, int(args.vram_headroom * 1024 ** 3)) for d in comfy.model_management.get_all_torch_devices())
@@ -602,7 +602,7 @@ if __name__ == "__main__":
"dynamic vram enabled please give us a detailed reports as this "
"argument will be removed soon. If you use gguf we recommend keeping "
"dynamic vram enabled and using native ComfyUI model formats instead. "
- "ComfyUI native formats like fp8 will be faster even if they are larger than your memory."
+ "ComfyUI native formats like fp8, int8 and w4a8 will be faster even if they are larger than your memory."
)
event_loop, _, start_all_func = start_comfyui()
try: