When on 4000 ADA, it's RANDOMLY NOT DETECTING GPU!
When on 4000 ADA, it's RANDOMLY NOT DETECTING GPU! Yesterday I set it up and it's okay. Today I set it up and not detecting GPU. NVIDIA-SMI says had it.
Why is that???
It doesn't comfortable if I need to always install torch cuda everytime, waste of time and money.
6 Replies
You can try using this tool, and also provide pod id so RunPod can investigate.
https://discord.com/channels/912829806415085598/1213495584539811860
The question is not about to investigate since we can install all of it by ourself. The question is, why sometimes it's okay, and sometimes not when we setup new.
I don't want to check everytime I start it. And I don't even need to investigate, because if it can't run then it means I must install all the way back to the torch cuda support and it takes time ( and of course money because the GPU rent ).
I'm sorry if I'm mad. I'm using a cheap GPU, but it's still matter to me because I'm not that rich. Don't need to waste my money and time to something that isn't stable.
The third time I try, it's working well again. But it really pisses me off if it's happen when I use the pricey one. Not even use the pricey one and I already pissed off like this.
Again I'm sorry. I know you just meant to help, but it's not useful for me. Or even, I don't really care about that solution, I want some kind of stability, or even cashback when it has a problem like this because wasting time so much ( But I know it's not possible at all ).
You can keep getting the bad pod if you don't help RunPod to diagnose it and find the pod with an issue.
Again, just for an argument... Since I paid for this, I don't feel like to has additional work task for me even though it's kind of simple.
I don't really want to argue because I know it's hard to achieve perfection in any kind. But right now, I just feel pissed. Maybe I will help later when I have more free time.
And again, if you can give me some kind of special account or discount, I really gladly help. Since I will be using your service so often ( That's why I'm pissed ), and will create pod again and again and again almost everyday maybe =="...
@flash-singh / @JM can assist. For sure, if you have technical issues with pods, feel free to just log the pod ID in the future, but the staff like JM / Flash, if there are technical issues runpod ive seen does offer refund / give credits for issues like this
@akiratoya13 how did you verify this? using nvidia-smi? if you run into it again please pm pod id, once we dig into it bit more, we can provide credits
for any type of credit request, some type of resource id is a must, helps us close the loop, you might be able to use the audit log to find it also