The unofficial DSPy framework. Build LLM powered Agents and "Agentic workflows" based on the Stanford DSP paper.

TypeScript to C++ compiler.

@mention people in a textarea

Train a tiny Llama3 model with parquet datasets, using JavaScript.

The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery 🧑‍🔬

Efficient Triton Kernels for LLM Training

Simple and fast low-bit matmul kernels in CUDA / Triton

A MLX port of FLUX based on the Huggingface Diffusers implementation.

Official repository of Evolutionary Optimization of Model Merging Recipes

Low code framework to build and launch a crew of AI agents with shared state. Built with https://axllm.dev.

Fast and memory-efficient exact attention

A plugin for Jupyter Notebook to run CUDA C/C++ code

LLM101n: Let's build a Storyteller

LLM training in simple, raw C/CUDA

Distributed pretraining of large language models (LLMs) on cloud TPU slices, with Jax and Equinox.

Efficient GPU kernels for block-sparse matrix multiplication and convolution

LightLLM is a Python-based LLM (Large Language Model) inference and serving framework, notable for its lightweight design, easy scalability, and high-speed performance.

Implementation of Diffusion Transformer (DiT) in JAX

CUDA Learning guide

Sparse autoencoders

GPU programming related news and material links

Make triton easier

Gemma 2B with 10M context length using Infini-attention.

The simplest way to set up your tsconfig.json

Large language models (LLMs) made easy, EasyLM is a one stop solution for pre-training, finetuning, evaluating and serving LLMs in JAX/Flax.

Simple and readable code for training and sampling from diffusion models

Top-down 2D RPG