jamesf/Perplexica
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Adds Google VertexAI as model provider

Browse source
This commit is contained in:
Hristo 2024-05-14 15:05:17 -04:00
parent cef75279c5
commit ebbe18ab45
12 changed files with 101 additions and 77 deletions
Download patch file Download diff file Expand all files Collapse all files

21
README.md
View file

@ -112,16 +112,17 @@ You need to edit the ports accordingly.
1a: Copy the `sample.env` file to `.env`
1b: Copy the `deploy/gcp/sample.env` file to `deploy/gcp/.env`
2a: Fillout desired LLM provider access keys etc. in `.env`
- Note: you will have to comeback and edit this file again once you have the address of the K8s backend deploy
2b: Fillout the GCP info in `deploy/gcp/.env`
3: Edit `GCP_REPO` to the correct docker image repo path if you are using something other than Container registry
4: Edit the `PREFIX` if you would like images and GKE entities to be prefixed with something else
5: In `deploy/gcp` run `make init` to initialize terraform
6: Follow the normal Preplexica configuration steps outlined in the project readme
7: Auth docker with the appropriate credential for repo Ex. for `gcr.io` -> `gcloud auth configure-docker`
8: In `deploy/gcp` run `make build-deplpy` to build and push the project images to the repo, create a GKE cluster and deploy the app
9: Once deployed successfully edit the `.env` file in the root project folder and update the `REMOTE_BACKEND_ADDRESS` with the remote k8s deployment address and port
10: In root project folder run `make rebuild-run-app-only`
- Note: you will have to comeback and edit this file again once you have the address of the K8s backend deploy
2b: Fillout the GCP info in `deploy/gcp/.env`
3: Edit `GCP_REPO` to the correct docker image repo path if you are using something other than Container registry
4: Edit the `PREFIX` if you would like images and GKE entities to be prefixed with something else
5: In `deploy/gcp` run `make init` to initialize terraform
6: Follow the normal Preplexica configuration steps outlined in the project readme
7: Auth docker with the appropriate credential for repo Ex. for `gcr.io` -> `gcloud auth configure-docker`
8: In `deploy/gcp` run `make build-deplpy` to build and push the project images to the repo, create a GKE cluster and deploy the app
9: Once deployed successfully edit the `.env` file in the root project folder and update the `REMOTE_BACKEND_ADDRESS` with the remote k8s deployment address and port
10: In root project folder run `make rebuild-run-app-only`
If you configured everything correctly frontend app will run locally and provide you with a local url to open it.
Now you can run queries against the remotely deployed backend from your local machine. :celebrate: