Project to recreate your favourite block game for the Wii (Beta 1.7.3)

The best ChatGPT that $100 can buy.

gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

Distributed Compiler based on Triton for Parallel Systems

Async RL Training at Scale

QuickReduce is a performant all-reduce library designed for AMD ROCm that supports inline compression.

Unofficial description of the CUDA assembly (SASS) instruction sets.

A tool for recompiling Xbox 360 games to native executables.

build-once run-anywhere c library

PCCL (Prime Collective Communications Library) implements fault tolerant collective communications over IP

A snapshot of our vulkan backend code

Installer from Photoshop CC 2021 to 2022 on linux with a GUI

Nvidia Instruction Set Specification Generator

NVIDIA Linux open GPU with P2P support

The Kotlin Programming Language.

LLM inference in C/C++

Mirage Persistent Kernel: Compiling LLMs into a MegaKernel

Assembler for NVIDIA Volta and Turing GPUs

Mamba SSM architecture

Development repository for the Triton language and compiler

CUDA Templates and Python DSLs for High-Performance Linear Algebra

StableLM: Stability AI Language Models

ChatLLaMA 📢 Open source implementation for LLaMA-based ChatGPT runnable in a single GPU. 15x faster training process than ChatGPT

Fork of Paper which adds regionised multithreading to the dedicated server.

Locally run an Instruction-Tuned Chat-Style LLM

Instruct-tune LLaMA on consumer hardware

RWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RN…

The simplest, fastest repository for training/finetuning medium-sized GPTs.