RunPod

R

RunPod

We're a community of enthusiasts, engineers, and enterprises, all sharing insights on AI, Machine Learning and GPUs!

Join

⚡|serverless

⛅|pods

issue with websocket (wss) port on runpods

hi there! i’m working on a real-time sdxl example and have tried several times, but for some reason, the websocket (wss) port never works with runpods. it works fine with http, but not with wss. any help would be appreciated!...

I want to make kohya ss available through the http service [port 7860] button.

I installed kohya_ss on my pod. I created a starting script for it I added some variables in the runpod ui. ...
No description

Public IP security?

Hello everyone, I wonder how public IP security is provided when we expose the port in runpod?...

Automatically shut down pod after being idle for X mins

Hi guys, a general question: I'd like to automatically shut down my pod after being idle for 30 mins (so I can complete several trainings overnight). I've found this command bash -c "nohup sleep 2h; runpodctl stop pod $RUNPOD_POD_ID" & but the thing is that I have to pre-define after which period the pod gets shut down. Is there any functionality that automatically monitors the pod's status and shuts it down after being idle for X mins?

Updating Forge in Runpod

Hi everyone. I'm trying to set up Forge in Runpod. I can get it up and going using the basic SD Forge template (zlm1kalme2d6rf), but it appears to be an outdated version. Specifically, there's a UI indicator in the upper left of the current version labeled UI that allows you to select SD, XL, Flux or All, and it alters other input boxes based on which one of these fields you select. The Forge in this specific template doesn't seem to have this option. I've tried to update Forge by doing 'git pull https://github.com/lllyasviel/stable-diffusion-webui-forge.git' from the /workspace/stable-diffusion-webui-forge directory. It does something (I am pretty noobish at Git and terminal navigation), but it's either not updating properly, or it's not adding this upper left UI option. It's possible that this is a settings issue, however every YouTube tutorial that shows people installing Forge on their local machine loads this by default which suggests to me this is an old version I'm using. ...
No description

api having issues randomly

File ".../anaconda3/lib/python3.11/site-packages/runpod/api/graphql.py", line 37, in run_graphql_query raise error.QueryError( runpod.error.QueryError: Something went wrong. Please try again later or contact support. I get this randomly. i'm using the api to rent a single 4090 on secure servers with a template...

0 GPUs notif

I am getting this error when I attempt to start the pod: Start your pod without GPUs. This is useful for debugging non gpu-related problems or transferring data. If you have a volume configured, it will be retrieved and mounted. The price for this instance is $0.945/hour + disk costs. I have $41.91 loaded, so I'm unsure of why....

Running a Dockerized Pytorch-based computer vision app

I would like to run a pytorch-based computer vision program using runpod. I have a Dockerized application that relies on Cuda 12.1 and Pytorch 2.3.1. How can I do this with runpod?

Minimize the startup time of ComfyUI on serverless/pod GPU

Hello, Hope everybody is good. Thanks for this amazing community. I am currently facing an issue which is as follow: -> I am running ComfyUI on my local machine and on my local machine with my current workflow, models, and Loras loaded in around 15-30 seconds, and after that processing of the image starts which takes around 30 seconds. That is the desired or better result I am looking into on my paid Runpod serverless and pod GPU's...

Can I access Network Volume via VS Code Explorer?

I deployed a pod using a Network Volume. I am able to Remote-SSH into the pod using VS Code. When I open up the terminal in VS Code, I have access to the network volume (/workspace). I'd like to access the network volume via the VS Code Explorer, but as you can in the attached image, the EXPLORER: ROOT seems to be a different volume (I'm guessing it's the container volume). ...
No description

Error 804: forward compatibility was attempted on non supported HW

Getting a driver mismatch using the runpod provided container runpod/pytorch:2.4.0-py3.11-cuda12.4.1-devel-ubuntu22.04 on an 1 x RTX 4090 ID: 7wqshyg5hf93r3...

move network storage between data centers

Is there a way to move network volumes between different data centers on an ad-hoc basis? Currently, we select data centers primarily for cost reasons. However, in the near future, we will need to start doing tests on more powerful machines, and eventually, we may migrate most of our workloads to these higher-performance systems. The issue is that the current system doesn’t seem to support moving a network volume to a different data center. Ideally, we would like to have mirrors of a network volume across multiple data centers, which would significantly accelerate our production....

Forge webUI template is outdated, git pull is not updating it

Trying to run Forge with Flux the latest version. The template is outdated, and even if I do a git pull it's not updating it to the current version. I'm spending money on trying to get around this but looks like Forge will not work with the latest version on Runpod?
Solution:
Do a git clone instead of a git pull and it is solved

How to enable lsof command?

I tried checking some ports on my pod and I read that lsof can be used for that but when I try using it in my terminal I get a "bash: sudo: command not found" response Is there a way to enable lsof?...
Solution:
apt update & apt install lsof

How can I add jupyter notebook to my custom template?

How can I add jupyter notebook to my custom template?

any way to do this via api? runpodctl exec {podId} -- {command}

i know i can do subprocess run and wrap it in fast api, but trying to save 1 extra step if anyone knows, AI said its not possible.

I cannot send more than one env in pod creation using GraphQL

Greetings I send a graphql request using PHP, if the env field is a single env, it creates a pod, but when I try to send more than one env, I get an error. Error -> Something went wrong. Please try again later or contact support.
No description

Slow download speed

Complete RunPod newbie here. I'm using Flux.1-Dev-ComfyUI by Camenduru template. However, its download speed is very slow, even the pod shows connection speed is 14836 Mbps. Are there something I missing? Is this normal? It takes nearly 10 mins to complete downloading. (I know Flux.1 Dev is 32GB and it's being hosted on HuggingFace.)
Solution:
Okay, here's what I found. I guess Docker RUN command runs slower(~8x) than just directly using aria2c. So I created network volume, deployed empty(without model) ComfyUI image based on ghcr.io/ai-dock/comfyui:latest, mount network volume to /workspace/storage/stable_diffusion/models, and manually downloaded models there. After first download to network volume, I don't have to download 32GB model again....

ComfyUI in aitrepreneur/comfyui:2.3.5 template cannot be updated anymore. Alternatives?

aitrepreneur/comfyui:2.3.5, the replacement for ashleykza/comfyui:2.4.0, doesn't work if we manually force the update of ComfyUI. The server runs but there is a mismatch in the expected IP and the one in the syste, AFAIK. Probably due to the big set of changes in the recent updates of ComfyUI. Any alternative template with an updated comfyUI and minimal overhead?

I'm having trouble ihaving docker inside a pod i want to push my image using my pod

docker: Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?. .......