IlyasMoutawwakil

gpu-benches

collection of benchmarks to measure basic GPU capabilities

llm-perf-backend

The backend behind the LLM-Perf Leaderboard

py-txi

A Python wrapper around HuggingFace's TGI (text-generation-inference) and TEI (text-embedding-inference) servers.

optimum-amd

AMD related optimizations for transformer models

uBlock

uBlock Origin - An efficient blocker for Chromium and Firefox. Fast and lean.

insanely-fast-whisper

optimum-nvidia

scrape-open-llm-leaderboard

Scrape and export data from the Open LLM Leaderboard.

OmniQuant

[ICLR2024 spotlight] OmniQuant is a simple and powerful quantization technique for LLMs.

TensorRT-LLM

TensorRT-LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and build TensorRT engines that contain state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT-LLM also contains components to create Python and C++ runtimes that execute those TensorRT engines.

NEFTune

Official repository of NEFTune: Noisy Embeddings Improves Instruction Finetuning

GPU-Puzzles

Solve puzzles. Learn CUDA.

text-embeddings-inference

A blazing fast inference solution for text embeddings models

optimum-quanto

A pytorch quantization backend for optimum

attention_sinks

Extend existing LLMs way beyond the original training length with constant memory usage, without retraining

llm-vscode

LLM powered development for VSCode

diffusion-models-class

Materials for the Hugging Face Diffusion Models Course

cuda-python

CUDA Python Low-level Bindings

onnxscript

ONNX Script enables developers to naturally author ONNX functions and models using a subset of Python.

llvm-project

The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.

mteb

MTEB: Massive Text Embedding Benchmark

onnx

Open standard for machine learning interoperability

exllamav2

A fast inference library for running LLMs locally on modern consumer-class GPUs

AutoGPTQ

An easy-to-use LLMs quantization package with user-friendly apis, based on GPTQ algorithm.

private-gpt

Interact with your documents using the power of GPT, 100% privately, no data leaks

hydra-moe

nougat

Implementation of Nougat Neural Optical Understanding for Academic Documents

SpQR

AutoAWQ

AutoAWQ implements the AWQ algorithm for 4-bit quantization with a 2x speedup during inference. Documentation:

peft

🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.