LLM_for_isotope_separation

This project leverages a Large Language Model (LLM) to extract key settings and experiment results from an extensive volume of literature, enabling the design of experiments tailored to specific research targets

Workflow:

1. Extract related sections from research papers with model1 (or keyword filter).
2. Extract experiment setting and results from the selected section from model2.
3. Create a dataset comprising experiment settings.
4. Train/Fine-tune an inference model3 with the dataset, using it to design experiments

Useful Lib

Llama index: https://github.com/jerryjliu/llama_index

Installation

Usage

Data

Result