Repository History
5 repositories tagged with Hugging Face
Observers: A Lightweight Library for AI Observability in Python
Observers is a Python library designed for AI observability, enabling developers to track and store interactions with generative AI APIs. It provides a flexible framework with various observers for popular LLM providers and multiple storage backends. This tool helps in monitoring, debugging, and analyzing AI model behavior effectively.
Jsonformer: Bulletproof Structured JSON Generation from Language Models
Jsonformer is a powerful library designed to generate syntactically correct and schema-conforming JSON from language models. It addresses the common challenge of unreliable JSON output by focusing on generating only content tokens, making the process more efficient and robust. This approach ensures bulletproof structured data generation for various applications.
LMQL: A Language for Constraint-Guided and Efficient LLM Programming
LMQL is an innovative programming language that extends Python, designed for efficient and constraint-guided programming of Large Language Models (LLMs). It allows developers to interweave traditional programming logic with native LLM calls, offering advanced control over model behavior through features like constraints, rich control flow, and optimized runtimes. This makes it easier to build complex LLM-powered applications with greater precision and efficiency.
CSM: A Conversational Speech Generation Model by SesameAILabs
CSM (Conversational Speech Model) is an advanced speech generation model from SesameAILabs, designed to create RVQ audio codes from text and audio inputs. It leverages a Llama backbone and a smaller audio decoder for Mimi audio codes, enabling high-quality, context-aware speech synthesis. The model is now natively available in Hugging Face Transformers, making it accessible for researchers and developers.
Picotron: Minimalistic 4D-Parallelism Framework for LLM Training Education
Picotron is a minimalistic and hackable distributed training framework designed for educational purposes. Inspired by NanoGPT, it focuses on pre-training Llama-like models using 4D Parallelism, making complex concepts accessible. Its simple and readable codebase, with core files under 300 lines, provides an excellent tool for learning and experimentation in distributed machine learning.