https://github.com/timelesshc/anime-recommender-system
An anime recommender system using LLM and RAG
https://github.com/timelesshc/anime-recommender-system
anime chromadb fullstack-app llm rag rag-chatbot
Last synced: 4 months ago
JSON representation
An anime recommender system using LLM and RAG
- Host: GitHub
- URL: https://github.com/timelesshc/anime-recommender-system
- Owner: timelesshc
- Created: 2025-08-30T22:34:32.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2025-09-01T04:08:03.000Z (5 months ago)
- Last Synced: 2025-09-01T06:45:47.680Z (5 months ago)
- Topics: anime, chromadb, fullstack-app, llm, rag, rag-chatbot
- Language: Python
- Homepage:
- Size: 2.18 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
### 1. Project Introduction
An anime recommendation system using LLM and RAG pipelines.
- **Tech stack**
- Groq --> LLM
- HuggingFace --> Embedding model
- Langchain --> AI framework to interact with LLM
- GCP VM --> Virtual machine on the cloud
- Minikube --> For making a Kubternetes Cluster where you can deploy your application
- Streamlit --> To make UI or frontend of the app
- Docker --> For containerization of the app during deployment
- Grafana Cloud --> Monitoring your Kubernetes Clusters
- Chroma DB --> Local vector store for storing embeddings
- GitHub --> Work as a Source Code Management (SCM) for your project
- **Workflow**
### 2. Configure VM Instance
- **Clone your GitHub repo**
```bash
git clone https://github.com/timelesshc/anime-recommender-system.git
ls
cd anime-recommender-system
ls # You should see the contents of your project
```
- **Install Docker**
- Search: "Install Docker on Ubuntu"
- Open the first official Docker website (docs.docker.com)
- Scroll down and copy the **first big command block** and paste into your VM terminal
- Then copy and paste the **second command block**
- Then run the **third command** to test Docker:
```bash
docker run hello-world
```
- **Run Docker without sudo**
- On the same page, scroll to: **"Post-installation steps for Linux"**
- Paste all 4 commands one by one to allow Docker without `sudo`
- Last command is for testing
- **Enable Docker to start on boot**
- On the same page, scroll down to: **"Configure Docker to start on boot"**
- Copy and paste the command block (2 commands):
```bash
sudo systemctl enable docker.service
sudo systemctl enable containerd.service
```
- **Verify Docker Setup**
```bash
systemctl status docker # You should see "active (running)"
docker ps # No container should be running
docker ps -a # Should show "hello-world" exited container
```
### 3. Configure Minikube inside VM
- **Install Minikube**
- Open browser and search: `Install Minikube`
- Open the first official site (minikube.sigs.k8s.io) with `minikube start` on it
- Choose:
- **OS:** Linux
- **Architecture:** *x86*
- Select **Binary download**
- Reminder: You have already done this on Windows, so you're familiar with how Minikube works
- **Install Minikube Binary on VM**
- Copy and paste the installation commands from the website into your VM terminal
- **Start Minikube Cluster**
```bash
minikube start
```
- This uses Docker internally, which is why Docker was installed first
- **Install kubectl**
- Search: `Install kubectl`
- Run the first command with `curl` from the official Kubernetes docs
- Run the second command to validate the download
- Instead of installing manually, go to the **Snap section** (below on the same page)
```bash
sudo snap install kubectl --classic
```
- Verify installation:
```bash
kubectl version --client
```
- **Check Minikube Status**
```bash
minikube status # Should show all components running
kubectl get nodes # Should show minikube node
kubectl cluster-info # Cluster info
docker ps # Minikube container should be running
```
### 4. Interlink your Github on VSCode and on VM
```bash
git config --global user.email "xxx@xxx.com"
git config --global user.name "xxx"
git add .
git commit -m "commit"
git push origin main
```
- When prompted:
- **Username**: `xxx`
- **Password**: GitHub token (paste, it's invisible)
---
### 5. Build and Deploy your APP on VM
```bash
## Point Docker to Minikube
eval $(minikube docker-env)
docker build -t llmops-app:latest .
kubectl create secret generic llmops-secrets \
--from-literal=GROQ_API_KEY="" \
--from-literal=HUGGINGFACEHUB_API_TOKEN=""
kubectl apply -f llmops-k8s.yaml
kubectl get pods
### U will see pods runiing
# Do minikube tunnel on one terminal
minikube tunnel
# Open another terminal
kubectl port-forward svc/llmops-service 8501:80 --address 0.0.0.0
## Now copy external ip and :8501 and see ur app there....
```
### 6. GRAFANA CLOUD MONITORING
```bash
## Open another VM terminal for Grfana cloud
kubectl create ns monitoring
kubectl get ns
## Make account on Grfaana cloud
### Install HELM - Search on Google
-- Copy commands from script section..
-- U will get 3 commands
## Come to grafana cloud --> Left pane observability --> Kubernetes--> start sending data
## In backend installation --> Hit install
## Give your clustername and namespace there : minikube and monitoring in our case
## Select kubernetes
## Keep other things on as default
## Here only create new access token give name lets give minikube-token & Create it and save it somewhere..
## Select helm and deploy helm charts is already generated...
## Come to terminal --> Create a file
vi values.yaml
## Paste all from there to your file now remove last EOF part & and also initial part save that initial part we need it..
Example :
helm repo add grafana https://grafana.github.io/helm-charts &&
helm repo update &&
helm upgrade --install --atomic --timeout 300s grafana-k8s-monitoring grafana/k8s-monitoring \
--namespace "monitoring" --create-namespace --values - <<'EOF'
### Remove this above intial part and save it somewhere
Then Esc+wq! amd save the file
## Now use the copied command just make some modification:
Remove that EOF part and instead write
--values values.yaml
Example:
helm repo add grafana https://grafana.github.io/helm-charts &&
helm repo update &&
helm upgrade --install --atomic --timeout 300s grafana-k8s-monitoring grafana/k8s-monitoring \
--namespace "monitoring" --create-namespace --values values.yaml
## Paste this command on VM u will get status deployed revision 1
## It means it was a SUCESS
To check:
kubectl get pods -n monitoring
# These are all should be running.....
Go to grafana cloud again..
And below u will get go to homepage click it..
Just refresh the page and boom..
Now u can see metrics related to your kubernetes cluster..
---Explore it for yourself now
---Make sure to do cleanup
```