DSPy: The framework for programming—not prompting—foundation models

Inference code for CodeLlama models

Mamba SSM architecture

LLMs build upon Evol Insturct: WizardLM, WizardCoder, WizardMath

Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.

Segment Anything in High Quality [NeurIPS 2023]

Rift: an AI-native language server for your personal AI software engineer

4 bits quantization of LLaMA using GPTQ

Pronterface, Pronsole, and Printcore - Pure Python 3d printing host software

ReFT: Representation Finetuning for Language Models

Customizable implementation of the self-instruct paper.

fast vector database made in numpy

A multi-purpose LLM framework for RAG and data creation.

Code for "AutoRecon: Automated 3D Object Discovery and Reconstruction" CVPR 2023 (Highlight)

Extract structured text from pdfs quickly

Lord of LLMS

tsl: a PyTorch library for processing spatiotemporal data.

Generate Synthetic Data Using OpenAI, MistralAI or AnthropicAI

A Python toolbox for conformal prediction research on deep learning models, using PyTorch.

DNXFIREWALL® and DAD'S NEXT-GEN FIREWALL™, a C/CPython hybrid next generation firewall built on top of Linux and bound to kernel/ netfilter hooks for packet control.

Efficient and general syntactical decoding for Large Language Models

Code to reproduce "Transformers Can Do Arithmetic with the Right Embeddings", McLeish et al (2024)

SFACT the New Skeinforge

app test for satgpt updates (langchain + panel)

A transpiler for GdScript (targeting C# and c++) written in Python

Residual Prompt Tuning: a method for faster and better prompt tuning.

Inference Llama 2 in one file of zero-dependency, zero-unsafe Rust

A framework to train agents in a Unity environment using a genetic algorithm

Software and stuff for RepRap World Tour participants

Train Alpaca-LLaMA with the zero-order MeZO optimizer