Anime Search App Using Weaviate and Gradio

This project demonstrates how to build a simple anime search application using Weaviate for vector search and Gradio for creating an interactive UI. The app allows users to search for anime using embeddings stored in a Weaviate database.

---

Prerequisites

Before getting started, ensure you have the following installed on your system:

• Python (3.10.6)
• Docker and Docker Compose

---

Installation

Follow these steps to set up and run the application:

1. Install Required Python Libraries

Run the following command to install the necessary Python dependencies:

pip install weaviate-client pandas

2. Start Weaviate Using Docker

Use Docker Compose to spin up a Weaviate instance:

docker-compose up -d

This will start a local Weaviate server that will store and manage the anime data.

3. Preprocess the Dataset (Optional)

This dataset has been preprocess so you don't need to run this. But if you want to preprocess a custom dataset, modify the clean-dataset.py file according to your needs. Once modified, run the script:

python clean-dataset.py

The original dataset from: https://www.kaggle.com/datasets/andreuvallhernndez/myanimelist-jikan

4. Create a Collection in Weaviate

Run the following script to create the anime collection and insert the dataset into the Weaviate instance:

python create-collection.py

Wait for the data to be inserted into Weaviate. Once the process is complete, you should see output similar to this:

This process taking pretty much time depend on your dataset.

5. Launch the Anime Search App

Run the search application:

python search-anime.py

After the script starts, you’ll see a local URL (e.g., http://127.0.0.1:7860). Open this URL in your browser to interact with the app.

enjoy