11 Replies
Also, why am I getting billed while I'm unable to use comfyUI?
The pod is having issues after crashing, and also the terminal UI has long latency
Could I get a refund for this? Has been stuck like this for 2 hr+s
wasted hours worth of credits just past week. Something must have broken on their side but wont admit it
Good morning everyone! Having the same problem
runpod comfyui container Error response from daemon: No such image: runpod/stable-diffusion:comfy-ui-5.0.0
Any understanding on what happened? Hoping for a speedy recovery
@Arc @zz_zac Hey guys, the issue seems to be solved for me, I just managed to successfully startup the pod with SD ComfyUI templateOne issue they have told me is the mismatch between the cuda version of the Pod Template vs the Host. With 5090 we need to select 12.8 but the host runs 12.9 so that causes a mismatch...if you set CUDA version to 12.8 it should work.
However there is a speed issue somewhere. I'm sure when they see huge decline in revenue they will look deeper.
I will give this a try. Looks promising because on random occasions I got CUDA mismatch error in my logs.. Sometimes, nothing shows up including the logs
CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with
TORCH_USE_CUDA_DSA to enable device-side assertions.
Getting this CUDA error@zz_zac For now, just use a 4090, I should get the job done. There's a problem right now about pytorch, I don't quite get it yet, but using a previous gen GPU did work for me
I'm using 5090 with Cuda 12.8 version. All running fine now
Nice