search

ItzCrazyKns / Perplexica

Perplexica is an AI-powered search engine. It is an Open source alternative to Perplexity AI
MIT License
13.8k stars 1.31k forks source link

@ItzCrazyKns Can you help having the correct configurations that must be adapted everywhere universally? #10

Closed arunkumarakvr closed 4 months ago

arunkumarakvr commented 5 months ago

@ItzCrazyKns Is this Perplexica repo already production ready? If not, when will it become production ready?

Screenshot_20240418_155627_Chrome.png

This is not working on GitHub CodeSpace. Why? Anyways to restart the containers or to check if env var setup working properly?

Originally posted by @arunkumarakvr in https://github.com/ItzCrazyKns/Perplexica/issues/8#issuecomment-2063486742

ItzCrazyKns commented 5 months ago

This is not working on GitHub CodeSpace. Why? Anyways to..

Github codespaces aren't meant to run things on production so we'll not talk it about here but all I can say about it is that the frontend is not being able to connect to the websocket as the environment variable passed by Docker is meant for local uses. To deploy it you can make use of Docker but you would have to change the environment variables in the docker compose file, specially the backend API URL and the websocket address (you need to change the IP to the IP of the server you are using to deploy it). I believe if you are deploying it somewhere to use it with multiple devices you must have even a little knowledge about Docker then you'll be able to do it with ease.

Is this Perplexica repo already production ready? If not, when will it become production ready?

I first need to understand what do you mean by production before I can provide you any information.

arunkumarakvr commented 5 months ago

@ItzCrazyKns Sorry for the confusion, as you already mentioned it's under development, it wouldn't be possible to use it in a production environment I think. And also what I meant by env var is OpenAI API Key but not the ones inside the docker files because as you may see in the above screenshot, there is no search result or text generation happening. That's all I meant, while I tried running it on GitHub CodeSpace particularly.

ItzCrazyKns commented 5 months ago

It might be because the websocket connection couldn't be established due to configuration reasons. Please update those in the Docker compose file before rebuilding the images. And yes, Perplexica is under heavy development so I won't recommend it for production yet but lets see in the future

ItzCrazyKns commented 5 months ago

Hello, could you please respond to this issue within the next 24 hours? If I don't hear back from you by then, I'll assume that everything is resolved and will mark it as completed and closed.

arunkumarakvr commented 5 months ago

Hello, could you please respond to this issue within the next 24 hours? If I don't hear back from you by then, I'll assume that everything is resolved and will mark it as completed and closed.

@ItzCrazyKns I don't know how to respond to this, as this issue still persists even today. But this issue is never resolved.

arunkumarakvr commented 5 months ago

This is not working on GitHub CodeSpace. Why? Anyways to..

Github codespaces aren't meant to run things on production so we'll not talk it about here but all I can say about it is that the frontend is not being able to connect to the websocket as the environment variable passed by Docker is meant for local uses. To deploy it you can make use of Docker but you would have to change the environment variables in the docker compose file, specially the backend API URL and the websocket address (you need to change the IP to the IP of the server you are using to deploy it). I believe if you are deploying it somewhere to use it with multiple devices you must have even a little knowledge about Docker then you'll be able to do it with ease.

Please help with the right IP address that must be used in the docker-compose file. Is it 0.0.0.0:3001 or 'external ip of server like vm':3001 or any docker internal domain cited in the ollama feat issues that must be changed or should I change such an IP wherever localhost or 127.0.0.1 across this entire Perplexica repo?

arunkumarakvr commented 5 months ago

It might be because the websocket connection couldn't be established due to configuration reasons. Please update those in the Docker compose file before rebuilding the images. And yes, Perplexica is under heavy development so I won't recommend it for production yet but lets see in the future

Please help with the correct or right configuration that must be adapted. In both the .env and config.toml files, the configuration issues continue to persist everywhere from local to cloud.

ItzCrazyKns commented 5 months ago

I've set up a Discord server to improve communication and provide support more effectively. It will be easier to share files, message, and address your issues. You can join the server here: https://discord.gg/HUDjfQvM.

ItzCrazyKns commented 5 months ago

This is not working on GitHub CodeSpace. Why? Anyways to..

Github codespaces aren't meant to run things on production so we'll not talk it about here but all I can say about it is that the frontend is not being able to connect to the websocket as the environment variable passed by Docker is meant for local uses. To deploy it you can make use of Docker but you would have to change the environment variables in the docker compose file, specially the backend API URL and the websocket address (you need to change the IP to the IP of the server you are using to deploy it). I believe if you are deploying it somewhere to use it with multiple devices you must have even a little knowledge about Docker then you'll be able to do it with ease.

Please help with the right IP address that must be used in the docker-compose file. Is it 0.0.0.0:3001 or 'external ip of server like vm':3001 or any docker internal domain cited in the ollama feat issues that must be changed or should I change such an IP wherever localhost or 127.0.0.1 across this entire Perplexica repo?

If you are using the old feat/ollama branch kindly update to the latest version of the master branch as it contains the latest code supporting local models. Before we make it run on the network we need to make sure it runs locally on the computer. As I said in the other issue, please update the Ollama URL to http://host.docker.internal:11434. This should fix the issue for no output from the model. Afterwards you can change the IP to the private IP of the computer where it is hosted and it should work completely fine.

gscfwid commented 5 months ago

I've set up a Discord server to improve communication and provide support more effectively. It will be easier to share files, message, and address your issues. You can join the server here: https://discord.gg/HUDjfQvM.

The discord link is not available now, @ItzCrazyKns could you please update it?

ItzCrazyKns commented 5 months ago

I've set up a Discord server to improve communication and provide support more effectively. It will be easier to share files, message, and address your issues. You can join the server here: https://discord.gg/HUDjfQvM.

The discord link is not available now, @ItzCrazyKns could you please update it?

https://discord.gg/jFc3nnF3

ItzCrazyKns commented 4 months ago

Documentation for this has been added https://github.com/ItzCrazyKns/Perplexica/blob/master/docs/installation/NETWORKING.md

arunkumarakvr commented 4 months ago

@ItzCrazyKns these APIs for ollama LLMs or Azure OpenAI LLMs or embeddings models don't work at all despite these changes.