Runpod API documentation
Is the runpod API documented somewhere?
I've failed to find anything about it, and have had to resort to reverse engineering the webUI backend interaction (graphql mostly), and infer what the API might be by looking at how runpodctl makes API calls. Would be great to have the complete API documented, as it would allow creating much better tooling, saving users time and also increasing the overall value of the runpod platform (win-win)....
Is runpodctl abandonware?
I notice there's a lot of useful PRs for runpodctl, but zero comments nor activities on those from runpod's side. So I'm wondering if there is at least someone at runpod keeping an eye on the runpodctl github project, or if it's left to rot. Would be useful to know so the community knows whether it's time to fork it or create a replacement.
Docker argument issues - python sdk (docker_args)
If i copy paste the docker arguments that work fine on the website into the docker_args of runpod.create_pod method, i get error "Something went wrong. Please try again later or contact support".
The only difference is that the args are in triple quotes. Im not sure how I should be formatting them....
Modify existing FaceFusion template?
Is it possible to clone a template so that I can modify it? The template for FaceFusion Face Swapper and Enhancer is no longer being updated, and I’d like to modify it to use the updated code for version 3.0 from HuggingFace.
no cuda gpu detected
I dont know if this is a general problem or not
import torch torch.cuda.is_available()
gives error might not have cuda gpu
...
Solution:
When deploying a pod click filters then you will get will get these options. There is a pull down for CUDA Versions.
any way to control the restart policy of pods?
by default, it seems like runpod always restarts the pod after any termination. I am wondering whether there is a flag or other option to control the restart policy.
for instance, K8s have the following restart policy:
Container restart policy...
how can i deploy an instance with 4070, 4080 gpu?
when i deploy i only see 4090 and 3090 and other datacentre gpu and not RTX series gpus?
CAP_SYS_ADMIN privileges inside container
I am using a pytorch template and profiling some CUDA kernels. For the profiler to work inside the container, I need the container to be run with the
--cap-add=CAP_SYS_ADMIN
flag to docker run
, as far as I can tell the runpod platform does not offer control over the flags passed to docker run. Is there any way around this issue?
inside the container I see:...RunPod SD InvokeAI v3.3.0 Unable to import a model
I used this template a year ago and importing a model was easy enough. just copy the download url and paste it into the importer to add it. but today i tried it again and it always says undefined no matter what model i pick. i tried the URLs form both Civit ai and huggingface but still it says undefined.
Unable to restart pod
When restarting, the error log is as follows:
2024-10-01T09:17:54Z create container runpod/pytorch:2.1.0-py3.10-cuda11.8.0-devel-ubuntu22.04
2024-10-01T09:17:54Z 2.1.0-py3.10-cuda11.8.0-devel-ubuntu22.04 Pulling from runpod/pytorch...

connection refused SSH over exposed TCP
Everytime I am trying to connect via ssh over exposedd TCP I get a connection refused while I can connect normally usign the basic ssh terminal with no support for scp & sftp. WHich seems strange to me. I appreciate your help 🙂
Support for terminating pods via SkyPilot
Hi, I want to let my training runs go overnight and to terminate the pod once they are finished training. To do this, I am currently using SkyPilot. Whenever I try and stop a pod via SkyPilot, I get an error similar to
Stopping is currently not supported for RunPod
. Can RunPod please support this feature?rsync does not work
Hi, I am running a docker container in the cloud. Everything is working fine so far, I can connect via ssh with my public keys, everything is great. Except that I can't tranfer files using rsync. Everytime I am trying to transfer files via rsync I am getting asked for a password which I have never set. Does anybody has a solution for me?
Examples to connect I already tried (The paths are changed):
- rsync -avz -e "ssh" ~/documents/example.txt [email protected]:/root/example.txt
- rsync -avz -e "ssh -i /path/to/key" ~/documents/example.txt [email protected]:/root/example.txt...
Solution:
I just made it. I used a custom docker container. After readinghttps://blog.runpod.io/how-to-achieve-true-ssh-on-runpod/ I was able to solve it myself.
Build a docker compose yml file
I just made a runpod environment with gpu rtx 4090 and i have a github repository with yml file. I would like to run "docker-compose build" but I can't install docker properly in runpod environment any suggestions or helps please
Unable to Type into Terminal
Total newbie here. Am I losing it, or is there a reason I can't type at all into terminal? It just flashes where I would type. Thank you

Unable to Open or Delete a Folder
Hello. I am attempting to open a folder and nothing is happening when I attempt to open it. If I attempt to delete it, the attached error message is generated. This is for this workflow: https://civitai.com/models/790080/inpainting-simple-workflow-flux-or-upscale-or-lora-or-gguf. The workflow is also generating errors as if I don't have the flux 1-dev-q8_0.gguf file, but I'm not sure if I do, since I can't access the folder in question.
Any tips?
Thank you!...

Can create a Pod with an A1111 template
I've tried to create a pod with the "runpod/a1111:1.10.0.post7" template but it doesn't do anything. It stays on this screen:

We have detected a critical error on this machine which may affect some pods.
how can I get my pod data???
