I have a python function which uses OpenAI embeddings to do similarity search. It also uses LangChain, Google VertexAI and Qdrant. When deploying function this error comes up while installing dependencies. Output is in file attached. As I understand, the environment is missing some libraries. Is there any way to modify base environment? Or how do I fix that?
I edited "commands" in appwrite.json to this: apk --no-cache add musl-dev linux-headers g++ gcc openblas openblas-dev && pip install -r requirements.txt
And now it doesn't output any errors, but gets stuck on "installing build dependencies"
Does it fail? Or does it say processing?
To be honest, I just rewrote everything in another language. I managed to get it deployed successfully, but deploy times are huge + it didn't execute, problem with libraries once again.
[SOLVED] Problems with building python function
Python was slow or the new runtime?
Btw, the build container is different than the runtime container and it you have some runtime OS dependencies, you would need to make sure they're installed in the runtime container.
I've done this before by downloading the APK dependencies and ensuring they get installed in the function code
Recommended threads
- How to Avoid Double Requests in function...
I'm currently using Appwrite's `functions.createExecution` in my project. I want to avoid double requests when multiple actions (like searching or pagination) a...
- Send Email Verification With REST
I am using REST to create a user on the server side after receiving form data from the client. After the account is successfully created i wanted to send the v...
- Use different email hosts for different ...
Hello, I have 2 projects and i want to be able to set up email templates in the projects. Both projects will have different email host configurations. I see ...