Meta Lingua: a lean, efficient, and easy-to-hack codebase to research LLMs.

"LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding", Accepted to ACL 2024

Distill a Small Static Model from any Sentence Transformer

Prompt Engineering | Prompt Versioning | Use GPT or other prompt based models to get structured output. Join our discord for Prompt-Engineering, LLMs and other latest research

TorchFix - a linter for PyTorch-using code with autofix support

🤖 The free, Open Source alternative to OpenAI, Claude and others. Self-hosted and local-first. Drop-in replacement for OpenAI, running on consumer-grade hardware. No GPU required. Runs gguf, transf…

Mattermost Copilot plugin supporting multiple LLMs

DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads

Arena-Hard-Auto: An automatic LLM benchmark.

Scalable data pre processing and curation toolkit for LLMs

Chat with your data, modify it, visualize it, create and test machine learning models all in plain English. DataHorse makes data analysis and data science conversational using LLMs.

Optimized primitives for collective multi-GPU communication

Real-time video and audio processing on Streamlit

AllTalk is based on the Coqui TTS engine, similar to the Coqui_tts extension for Text generation webUI, however supports a variety of advanced features, such as a settings page, low VRAM support, D…

A Fundamental End-to-End Speech Recognition Toolkit and Open Source SOTA Pretrained Models, Supporting Speech Recognition, Voice Activity Detection, Text Post-processing etc.

Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.

Streamlit searchbox that dynamically updates and provides a list of suggestions based on a provided function

An educational machine learning library.

S2ORC: The Semantic Scholar Open Research Corpus: https://www.aclweb.org/anthology/2020.acl-main.447/

Fast, correct Python JSON library supporting dataclasses, datetimes, and numpy

🦉 Modern high-performance serialization utilities for Python (JSON, MessagePack, Pickle)

Scalable Computation of Hessian Diagonals

Examples of using sparse attention, as in "Generating Long Sequences with Sparse Transformers"

LLM training in simple, raw C/CUDA

Recipes for learning, fine-tuning, and adapting ColPali to your multimodal RAG use cases. 👨🏻‍🍳