Sage Attention installed correctly, But doesn't work in comfy #331
Description
Hello, Can someone help me with setting up sage attention? yes I have gone through every step and made sure I am using the right compatible version
Computer Specs :
Ryzen 5 7600
12gb Vram (3060)
32gb ram
ComfyUI startup code
Total VRAM 12288 MB, total RAM 31832 MB
pytorch version: 2.9.1+cu128
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce RTX 3060 : cudaMallocAsync
Using async weight offloading with 2 streams
Enabled pinned memory 14324.0
working around nvidia conv3d memory bug.
Using sage attention
Python version: 3.12.10 (tags/v3.12.10:0cc8128, Apr 8 2025, 12:21:36) [MSC v.1943 64 bit (AMD64)]
ComfyUI version: 0.4.0
ComfyUI frontend version: 1.33.13
My comfy starts normally, it does say using "sage attention". it loads all the model normally. But when the comfy workflow reaches the Ksampler and starts the rendering
This error(attached below) is being spammed at me continously . and as you can see the Sage attention is not being used instead pytorch is used. I have also made sure I have the right Triton version. And this is not in a particular workflow, rather present in all. Also I do get an image output and there is nothing wrong with the image as well, it's just I am not able to use the Sage attention
I have attached in here the screen shot of the error
Kindly help me with what is going on??