I’d like to run rasa on a raspberry pi (ideally in Docker) - just wondering if it’s possible?
My attempt to build failed when adding tensorflow. Has anyone got around this?
Also, is there any way of running rasa without tensorflow (e.g. if the responses for the bot were deterministic)?
Hey there,
it’s quite a pain to run TensorFlow on a raspberry. I assume you have seen this guide, Build from source for the Raspberry Pi | TensorFlow? Unfortunately TensorFlow is currently not an optional dependency.
I couldn’t get that to work, but managed to install tensorflow using piwheels in the end.
I have rasa-core and rasa-sdk containers running on pi now.
Haven’t yet managed to get duckling working with a haskell container
Hi, any luck with running rasa with duckling server on pi4?
Hi… I don’t generally comprehend being ready to run Rasa on RPi. Does it imply that I can run a model on RPi ? Would i be able to run the Rasa server in RPi too ? For instance on the off chance that I have a voice acknowledgment module interface with the RPi, I can separate the content, send that message to a Rasa server running in RPi and anticipate purposes, and so forth.