27 lines
863 B
Markdown
27 lines
863 B
Markdown
# Exam: Structured Output with Pydantic
|
||
|
||
This repository contains a small CLI tool that demonstrates how to extract structured data from raw text using LangChain and Pydantic.
|
||
|
||
## Features
|
||
- Two Pydantic models: `PersonInfo` and `MeetingNotes`.
|
||
- Automatic schema selection based on input type.
|
||
- One‑shot extraction via LangChain’s `PydanticOutputParser`.
|
||
- Simple command line interface with example texts.
|
||
|
||
## Usage
|
||
```bash
|
||
python -m cli --text "Анна, 28 лет, Python-разработчик. Навыки: FastAPI, Docker."
|
||
```
|
||
or
|
||
```bash
|
||
python -m cli --text "Встреча 2026-05-30: Участники: А., Б.; Темы: проект X; Решения: продлить срок."
|
||
```
|
||
|
||
## Requirements
|
||
- Python 3.10+
|
||
- `langchain-core`, `langchain-openai` (or `langchain-ollama`) and `pydantic`.
|
||
|
||
```bash
|
||
pip install -r requirements.txt
|
||
```
|