Your AI second brain. Self-hostable. Get answers from the web or your docs. Build custom agents, schedule automations, do deep research. Turn any online or local LLM into your personal, autonomous …

A PyTorch native platform for training generative AI models

Fast BM25 search in Python, powered by Numpy and Numba

LMDeploy is a toolkit for compressing, deploying, and serving LLMs.

A high-throughput and memory-efficient inference and serving engine for LLMs

A more memory-efficient rewrite of the HF transformers implementation of Llama for use with quantized weights.

Running large language models on a single GPU for throughput-oriented scenarios.

An easy-to-use LLMs quantization package with user-friendly apis, based on GPTQ algorithm.

LLM inference in C/C++

A fast llama2 decoder in pure Rust.

🦦 Otter, a multi-modal model based on OpenFlamingo (open-sourced version of DeepMind's Flamingo), trained on MIMIC-IT and showcasing improved instruction-following and in-context learning ability.

[EMNLP 2023] The CoT Collection: Improving Zero-shot and Few-shot Learning of Language Models via Chain-of-Thought Fine-Tuning

GRiT: A Generative Region-to-text Transformer for Object Understanding (ECCV2024)

🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.

[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.

MultimodalC4 is a multimodal extension of c4 that interleaves millions of images with text.

An open-source framework for training large multimodal models.

OCRmyPDF adds an OCR text layer to scanned PDF files, allowing them to be searched

EVA Series: Visual Representation Fantasies from BAAI

A modular RL library to fine-tune language models to human preferences

Corpus to accompany: "Do Androids Laugh at Electric Sheep? Humor "Understanding" Benchmarks from The New Yorker Caption Contest"

VIsually-Pivoted Audio and(N) Text

Code, data, models for the Sherlock corpus

🚀 A simple way to launch, train, and use PyTorch models on almost any device and distributed configuration, automatic mixed precision (including fp8), and easy-to-configure FSDP and DeepSpeed support

Repro is a library for easily running code from published papers via Docker.

A Python scikit for building and analyzing recommender systems

WordNet in JSON format.