11 Replies
You need a larger GPU type
no. I change 48GB bug it still 16GB
@ashleyk
@jax editing your endpoint doesn't apply the changes, you have to scale your workers down to zero and back up again for the changes to take effect when change the GPU tier.
OK
Looks like its still on 16GB tier
You can probably try 24GB tier before going to 48GB.
OK，I've been using 16GB before, and I didn't redeploy it tonight, and suddenly it's OOM!
It happens sometimes, I have an A1111 endpoint that usually works fine with 24GB VRAM but occasionally I get OOM for some requests so I also had to upgrade to 48GB because of it.
I tried it. - OK, thanks！
Your customer service response time is great!
Thanks, I don't even work for RunPod 😆
Sorry, I remember the last time you helped me with my problem too haha, you're really enthusiastic, hope to become friends