IlyasMoutawwakil

pytest-xdist

pytest plugin for distributed testing and loop-on-failures testing modes.

chug

Minimal sharded dataset loaders, decoders, and utils for multi-modal document, image, and text datasets.

lovely-tensors

Tensors, for human consumption

Triton-Puzzles

Puzzles for learning Triton

grok-1

Grok open release

lectures

Material for cuda-mode lectures

llm-swarm

Manage scalable open LLM inference endpoints in Slurm clusters

docker-py

A Python library for the Docker Engine API

uv

An extremely fast Python package installer and resolver, written in Rust.

rye

a Hassle-Free Python Experience

datatrove

Freeing data processing from scripting madness by providing a set of platform-agnostic customizable pipeline processing blocks.

nanotron

Minimalistic large language model 3D-parallelism training

triton

Development repository for the Triton language and compiler

lighteval

LightEval is a lightweight LLM evaluation suite that Hugging Face has been using internally with the recently released LLM data processing library datatrove and LLM training library nanotron.

cutlass

CUDA Templates for Linear Algebra Subroutines

ruff

An extremely fast Python linter and code formatter, written in Rust.

llama-cpp-python

Python bindings for llama.cpp

KVQuant

KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantization

float8_experimental

This repository contains the experimental PyTorch native float8 training UX

alphageometry

vllm

A high-throughput and memory-efficient inference and serving engine for LLMs

amdsmi

AMD SMI

hqq

Official implementation of Half-Quadratic Quantization (HQQ)

exllama

A more memory-efficient rewrite of the HF transformers implementation of Llama for use with quantized weights.

AutoAWQ_kernels

marlin

FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.

codecarbon

Track emissions from Compute and recommend ways to reduce their impact on the environment.

optimum-habana

Easy and lightning fast training of 🤗 Transformers on Habana Gaudi processor (HPU)

gpt-fast

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

gpu-benches

collection of benchmarks to measure basic GPU capabilities