JavierCCC

You can use windows_exporter in order to gather metrics from windows processes: https://github.com/prometheus-community/windows_exporter

same problem here. Trying to run a rag flow using ollama.

+1

> Still not possible in 2024 ? I think it is still as needed as possible.

i'm in the same situation using llama3.1 instruct through nvdia NMI container...

Same situation here. If i have to decide to leave a product, that product will not be grafana.

That example is not using vLLM, but the original code is using it. Same issue even loading the model with vLLM.

I figure out I was using the non async vllm engine. I switch to it but now I'm looking for a way to integrate async vllm engine with langchain. But...

It doesnt work. The only option that seems to work is using NSSM... can't believe this. I've opened a new question about this topic. If promtail binary can't run by...

This is still an issue