Kevin Hu
Kevin Hu
Closed due to no further response for days.
It's not supported yet. You could utilize Ollama/XInference... to server embedding model.
You could depoy a `slim` version of docker image which has no build-in embedding and rerank models.
Fetching more than 10K chunks has not been supported yet.
It's not for controlling the number of docs recalled. It's already deprecated for the current version.
I guess it's about memory shortage. But it will restart itself and redo faild tasks.
It's comming soon.
Appreciations! Could you please give some screen shot GIF or something to demostrate the scenario?
Cool! There're conflicts, please resolve them.
CI failure.