AI Configuration

There are 2 possible ways of configuring AI to help you tag and summarise books.

Chatgpt

To use OpenAI’s ChatGPT features, set up your openAI API key and desired model in the .env.local file:

OPEN_AI_API_KEY=XXXXXXXXX
OLLAMA_MODEL=gpt-3.5-turbo

Please refer to OpenAI’s documentation to get your API key and model.

If no key is set, ChatGPT will not be used in the application.

Ollama

Tip

This feature was tested with succes on mistral-nemo and llama3.2. Do not hesitate to share your best experience with other models!

To use a local instance of oLLama, set up the URL and model of your server in the .env.local file:

OLLAMA_URL=http://localhost:11434/api/
OLLAMA_MODEL=mistral-nemo

If no url is provided, Ollama will not be used.

Configure your prompts

Tip

Share your best prompts here!

When your favourite LLM is configured, fill in the prompts that will be used to generate the summaries and tags.

The application will replace the {book} placeholder with the book’s title, author and series.

The application will provide a base system prompt and add some formatting to the prompt.

Base prompt is defined in the code and is currently:

As a highly skilled and experienced librarian AI model, I'm here to provide you with deep insights and practical recommendations
based on my vast experience in the field of literature and knowledge organization. Upon requesting books related to a specific
topic or query, I will compile an extensive list of relevant titles accompanied by brief descriptions for your reference. If you
require more information about a particular book, I can provide a detailed description of its content and structure, helping you
decide if it's the right fit for your needs. For any specific chapter, part, or section within a book, my sophisticated algorithms
will generate an exhaustive outline accompanied by examples for each point to ensure clarity and comprehensiveness.
To enhance your experience even further, if you ask me to narrate a particular chapter or section, I will do my best to narrate
it as if I were the author of the book, taking care not to miss out on any important details. However, due to the intricacies
of the text, this could result in very lengthy responses as I aim to provide a faithful rendition of the content without
summarization. In general, I will refine your questions internally, so I will strive to offer more insights and beneficial
recommendations related to your request. If necessary, I will not hesitate to deliver very large responses up to 2000
tokens to ensure clarity and comprehensiveness. I will communicate with you primarily using your preferred language,
as it is assumed that this is how you're most comfortable interacting. However, when referencing titles of books or
other literature, I will maintain their original names in their respective languages to preserve accuracy and respect for these works.

For tags

For tag, the following will always be appended to your prompt:

The output must be only valid JSON format. It must be an object with one key named "genres" containing an array of genres and tags for this book in strings. Do not add anything else than json. Do not add any other text or comment.

Here are some example prompts:

Can you give 5 classifying keywords about the book {book} in a list without explanation in french

Give me a list of genres or tags for {book}. The first one must be about target audience (example: for teenagers), the second one must
be about the location of the main story (example: Switzerland), the next one are topical and must help the user find similar books.

For summaries

Can you make a factual summary of the book {book} in around 150 words in french

Run it for one book

On a book page, you can click on the “generate summary” or “generate tags” buttons to show propositions and accept them or not.

Run it through your whole library

You can run the following command:

docker compose exec biblioteca bin/console books:ai <tags|summary|both>

It will tag all your books that currently don’t have tags.

If you want to use a user’s configured prompts:

docker compose exec biblioteca bin/console books:ai <tags|summary|both> <userid>

If you want to use it on a specific book:

docker compose exec biblioteca bin/console books:ai <tags|summary|both> -b <book-id>

Add context

There are currently 2 ways to add context for better results. Both can be enabled at the same time.

From wikipedia

Set your WIKIPEDIA_API_TOKEN in your .env.local file. You must register in wikimedia API for a personal API Token.

From Amazon

Set AI_CONTEXT_AMAZON_ENABLED to 1 in your .env.local file to scrape results from Amazon. This should be used with parsimony.