Will there be FP4 quantization?
#10 opened about 20 hours ago
		by
		
				
							
						hlufeng
	
Unet Missing ..
								5
#9 opened about 24 hours ago
		by
		
				
							
						hanswurstwerner
	
1030 FP8 E4M3 version is bad result in comfyui
π
							
						3
				
									3
	#7 opened 3 days ago
		by
		
				
							
						t8star
	
Is there any chance that you will do something similar for T2V models?
#5 opened 5 days ago
		by
		
				
							
						Ruzarh
	
Should Wan2.2-Distill-Models used with Distill Loras?
π
							
						1
				#4 opened 10 days ago
		by
		
				
							
						ChathKay
	
add pipeline tag for better discoverability :)
π₯
							
						1
				#3 opened 15 days ago
		by
		
				
							
						linoyts
	
ComfyUI will still cast all layers to fp8
									3
	#2 opened 18 days ago
		by
		
				
							
						silveroxides
	
Error with quantized models using the "Load Diffusion Model" node in ComfyUI
									3
	#1 opened 18 days ago
		by
		
				
							
						xuanwoa