Hi everyone, I am starting a new project at work that will be based on BentoML and I have a few questions to help me plan better: 1. Is there any support for integrating ComfyUI Nodes with every Service with the aim of visualising the output at every step of the processing? 2. I saw that BentoML has (?) a Ray integration but the documentation is empty, can somebody clarify it? Also I have asked a more detailed question about the structure of the project [here](https://github.com/orgs/bentoml/discussions/5204), would appreciate some guidance on that too ❤️ Thanks for all the help!

Is there any support for integrating ComfyUI Nodes with every Service with the aim of visualising the output at every step of the processing?

Could you please elaborate a bit more on this question?

> Could you please elaborate a bit more on this question? Basically I wanna wrap every Service as a ComfyUI node so I can use ComfyUI to debug the whole project. This would be enabled by having the access to the input and output of every service in a visual way that also showcases the full workflow. > We had a Ray integration but it did not keep up with the latest of the Ray development due to the lack community interests. What is your use case for running BentoML on Ray? I am interested in local deployment on our research cluster that would also work with the fractional GPUs in order to allow horizontal scaling of models automatically through Ray.

With a custom implementation of nodes, it’s possible to embed bento services in ComfyUi nodes. However, I think it would be best for us to learn more about your use cases. Would you like to get on a quick call?

Booked, cheers!