Structured Text Generation

BAML is a language that helps you get structured data from LLMs, with the best DX possible. Works with all languages. Check out the promptfiddle.com playground

Resource list for generating JSON using LLMs via function calling, tools, CFG. Libraries, Models, Notebooks, etc.

Efficient, Flexible and Portable Structured Generation

Experimental Code for StructuredRAG: JSON Response Formatting with Large Language Models

Query language for blending SQL logic and LLM reasoning across structured + unstructured data. [Findings of ACL 2024]

Extract structured data from local or remote LLM models

TensorRT-LLM server with Structured Outputs (JSON) built with Rust

The developper starter pack for document processing

Structured Generation Evals

A guide to structured generation using constrained decoding

Use `outlines` generators with Haystack.

Speculative grammar backtracking algorithm for LLM decoding conforming to some lark context-free grammar (CFG)

NLP tasks with zero- and few-shot models.

Presentación en la conferencia IADevs 2024: Potenciando la Generación Aumentada usando Recuperación con Grafos de Conocimiento

Word alignment of multilingual sentences using structured generation

Presentation at PyData Global: Building Knowledge Graph-Based Agents with Structured Text Generation and Open-Weights Models

Mistral AI Fine-tuning Hackathon

Examples of code that use Outlines to enable structured text generation for LLMs running on Modal

A gentle tutorial on structured generation using Anthropic and Pydantic