Some of my libraries are not imported/detected

[BenDes21]

Hi there, some of my libraries that I included into my snapshot are not detected when the worker init, Im using the same version of comfyUI with same version of lib on my laptop and all working perfectly.

Wondering why it's not working with the workers :(

[BenDes21]

snapshot file is not loaded anymore since the update @TimPietrusky @TimPietruskyRunPod

[AIExplorer25]

Have you found a solution for it?

[TimPietrusky]

@BenDes21 @AIExplorer25 "restore snapshot" will be replaced by a super minimal dockerfile, see #116 (comment)

[TimPietrusky]

🎉 This issue has been resolved in version 5.0.0 🎉

The release is available on GitHub release

Your semantic-release bot 📦🚀

[TimPietrusky]

@AIExplorer25 you can use https://github.com/runpod-workers/worker-comfyui/blob/main/docs/customization.md#method-1-custom-dockerfile to create a custom dockerfile based on your workflow.

if you don't want to do this manually, then ping me on our Discord, send me your workflow and then i convert it into a serverless endpoint using the comfy-agent i'm working on

[AIExplorer25]

@AIExplorer25 you can use https://github.com/runpod-workers/worker-comfyui/blob/main/docs/customization.md#method-1-custom-dockerfile to create a custom dockerfile based on your workflow.

if you don't want to do this manually, then ping me on our Discord, send me your workflow and then i convert it into a serverless endpoint using the comfy-agent i'm working on

@TimPietrusky I managed to install the custom nodes. If you could help me, my workflow is generating video and I don't know how to return the video. The rp_handler.py only has code to download images, I want to download the video which got generated, I can see the video is getting generated, but the response I'm getting is not having the video, I tried to update the 'rp_handler.py' code, still not able to download the video. Do you have any sample code to download video, that would be really helpful. Thanks.

[TimPietrusky]

@AIExplorer25 like i said before: please ping me on our Discord, you can find me there as "Tim" or "NERDDISCO", because i would need to see your workflow in order to help out and also test your specific problem.

if you don't want that, then please open a new issue with your workflow and the use case of receiving a video, that is also ok!