README.md

---
title: Chatbot
emoji: 🫑
colorFrom: green 
colorTo: indigo
sdk: gradio
sdk_version: 3.23.0
app_file: chatbot/app.py
pinned: false
license: unknown
---

# Vegan Recipe Chatbot

![Screenshot of Chatbot initial interface](docs/assets/chatbot_init.png)

## Quickstart

`git clone` the repo  

```sh
cd lv-recipe-chatbot
```  

Make sure to use the version of python specified in `py_version.txt`  
Create a virtual environment.

```sh
python3 -m venv env
```

Activate the env and install dependencies.

```sh
source env/bin/activate
pip install -r requirements.txt
pip install -r requirements/dev.txt
```

Put API secrets in .env

```sh
cp .env.example .env
# edit .env with your secret key(s). Only OPEN_AI_KEY is required.
```

One option is to enter the poetry environment.  

```sh
poetry shell
```

Then start the Gradio demo.

```sh
python app.py
```

## Dependencies

If a new dependency for devlepment is helpful for developers, add it to `dev.txt`.  
If it is a dependency for the app that is imported in source code, add it to `core.txt`.  
Then run:

```sh
pipreqs --force
```

This will update our `requirements.txt` to include the depenency as it should be pinned in the environment.  

## Useful links

* [Task Matrix (Formerly Visual ChatGPT)](https://github.com/microsoft/TaskMatrix)
* [LangChain](https://python.langchain.com/en/latest/index.html)
* [LLM Prompt Engineering](https://www.promptingguide.ai)
* [OpenAI best practices for prompts](https://help.openai.com/en/articles/6654000-best-practices-for-prompt-engineering-with-openai-api)