This python package uses a large language model (currently OpenAI's gpt-3.5-turbo) to extract and then visualise (using the brilliant d3-milestones package) historical time lines from pdf documents.
Here are some example outputs:
- The life and works of Lord Byron extracted from the National Dictionary of Biography.
- A history of the world since the American War of Independence extracted from the last thirteen chapters of H G Wells A Short History of the World.
As with all uses of LLMs there is the possibility of hallucinations and so it is crucial that the information is verifiable. You can do this by clicking on each event on the timeline. You will then be taken to the page of the pdf from which the event was extracted.
Install using pip either directly from github...
pip install timeline_ai@git+https://github.com/coppeliaMLA/timeline_ai
Or if you have cloned the repository simply run pip install . from the repository root.
You will need an to run the package. Instructions for obtaining the key are here. Note the service is not free! Once you have the key you will need to add it to your environment variables. See this notebook for an example of how to do this in python.
A single line of code will generate the timeline.
import timeline_ai as ta
ta.build_timeline(pdf_file="byron_bio.pdf",
output_file = "byron.html",
timeline_title = "Byron's life and works",
start_year = 1780,
end_year = 1830)
The arguments to the function are as follows:
pdf_file(str): The path to the PDF file containing the timeline data.output_file(str): The path to save the generated timeline diagram.timeline_title(str): The title of the timeline.start_year(int): The starting year of the timeline.end_year(int): The ending year of the timeline.useful_info(str, optional): Additional information to pass to the LLM prompt.
You can find a jupyter notebook for generating the above examples here