A high-performance, lightweight, and cross-platform QUIC library

The largest collection of PyTorch image encoders / backbones. Including train, eval, inference, export scripts, and pretrained weights -- ResNet, ResNeXT, EfficientNet, NFNet, Vision Transformer (V…

Material for cuda-mode lectures

Scalable, Portable and Distributed Gradient Boosting (GBDT, GBRT or GBM) Library, for Python, R, Java, Scala, C++ and more. Runs on single machine, Hadoop, Spark, Dask, Flink and DataFlow

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

LLM inference in C/C++

Development repository for the Triton language and compiler

A subset of PyTorch's neural network modules, written in Python using OpenAI's Triton.

Inference Llama 2 in one file of pure C

An implementation of model parallel autoregressive transformers on GPUs, based on the DeepSpeed library.

Source code behind the python-patterns.guide site by Brandon Rhodes

Implementing a ChatGPT-like LLM from scratch, step by step

A library for mechanistic interpretability of GPT-style language models

Kolmogorov Arnold Networks

RWKV is an RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). So it's combining the best of RNN and transformer - great performance, fast inference,…

Minimal, clean code for the Byte Pair Encoding (BPE) algorithm commonly used in LLM tokenization.

DeepSpeed is a deep learning optimization library that makes distributed training easy, efficient, and effective.