Installation & Deployment

TIP

If using Docker or Hugging Face Space, you can skip directly to step 5

1. Clone the repository:

git clone https://github.com/ADT109119/ChatPDF-LineBot-Docs.git

2. Install server dependencies:

cd ChatPDF-LineBot
pip install -r requirements.txt

Consider running in a virtual environment. Here are the commands for creating a virtual environment using Python's built-in venv:

cd ChatPDF-LineBot
python -m venv .\venv
.\venv\Scripts\activate
pip install -r requirements.txt

3. Install Node dependencies:

cd ChatPDF-LineBot/View
npm install

4. Build the frontend:

cd View
npm install
npm run build

5. Log in to the LINE platform

  • Create a Line Bot.
    • Add a new provider, e.g., "My Provider".
    • Under "My Provider", create a new channel of type "Messaging API", e.g., "My AI Assistant".
    • Go to the "My AI Assistant" channel page, select the "Messaging API" tab, and generate a channel access token.
  • Create a LIFF page.
    • Under the same Provider (e.g., "My Provider"), create a new channel of type "Line Login", e.g., "My LIFF Page".
    • Go to the "My LIFF Page" channel page, select the "LIFF" tab, and create a new LIFF page.
    • For the Endpoint URL, enter https://<your-domain-or-IP-address>/liff (you can change this later).
    • For Scopes, select "chat_message.write" and "openid".
    • Click Add to create.

6. Log in to OpenAI or Groq platform (or other platforms)

  • Generate an OpenAI API key.
  • Optionally generate other platform API keys (e.g., Groq), but ensure the API is compatible with OpenAI and remember to update BASE_URL in the next step.

7. Set environment variables

You can set environment variables directly on the server or Docker.

If running on a local machine, copy the .env.example file, rename it to .env, and set the API keys, models, and other parameters.

See Environment Variables List for details

8. Run the server

Using Docker

Docker Image Size (tag)

docker run -d \
--name chatpdf-linebot \
-p 8000:8000 \
-v /local/file/store/path:/app/db \
-e LINE_CHANNEL_ACCESS_TOKEN=YOUR_LINE_CHANNEL_ACCESS_TOKEN \
-e LINE_CHANNEL_SECRET=YOUR_LINE_CHANNEL_SECRET \
-e LINE_LIFF_ID=YOUR_LINE_LIFF_ID \
-e LINE_LOGIN_CHANNEL_ID=YOUR_LINE_LOGIN_CHANNEL_ID \
-e MODEL_NAME=YOUR_MODELS \
-e BASE_URL=BASE_URLS \
-e API_KEY=API_KEYS \
adt109119/chatpdf-linebot

Running locally

If steps 1-4 were completed without issues, simply run the following command to start the server:

The default PORT is 8000

python .\main.py

Deploy with Hugging Face Space

This project can be quickly deployed on Hugging Face Space.

Supports custom URLs with CloudFlare Tunnel.

WARNING

If you haven't purchased permanent storage from Hugging Face,

files will be lost every time you update or change environment variable.

9. Finalize LINE settings

  • Line Bot settings

    • Go to the "My AI Assistant" channel page, select the "Messaging API" tab, set the "Webhook URL" to your app URL followed by the /callback path, e.g., https://line.the-walking.fish.com/callback, and click the "Update" button.
    • Click the "Verify" button to ensure the webhook is working.
    • Enable the "Use webhook" feature.
    • Disable the "Auto-reply messages" feature.
    • Disable the "Greeting messages" feature.

  • Line LIFF page settings

    • Go to the "LIFF" tab.
    • If the Endpoint URL was initially set to a placeholder, update it to the official LIFF page URL with the /liff path, e.g., https://line.the-walking.fish.com/liff.
    • Go back to the "LIFF" tab and copy the "LIFF URL".

  • Rich menu settings

    • Go to the "LINE Official Account Manager".
    • In the sidebar, find "Rich menu".
    • Click "Create".
    • Fill in the name, template, etc., as desired.
    • For "Action", choose "Link" and enter the LIFF URL copied in the previous step.
    • Click "Save".