A multi-level tensor algebra superoptimizer

离线证件水印助手

Example fasthtml applications demonstrating a range of web programming techniques

The fastest way to create an HTML app

Material for gpu-mode lectures

An efficient, flexible and full-featured toolkit for fine-tuning LLM (InternLM2, Llama3, Phi3, Qwen, Mistral, ...)

USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference

[USENIX ATC '24] Accelerating the Training of Large Language Models using Efficient Activation Rematerialization and Optimal Hybrid Parallelism

A collection of memory efficient attention operators implemented in the Triton language.

Official repository for LightSeq: Sequence Level Parallelism for Distributed Training of Long Context Transformers

[EMNLP 2024 Industry Track] This is the official PyTorch implementation of "LLMC: Benchmarking Large Language Model Quantization with a Versatile Compression Toolkit".

SGLang is a fast serving framework for large language models and vision language models.

FlagGems is an operator library for large language models implemented in Triton Language.

Official implementation of "Samba: Simple Hybrid State Space Models for Efficient Unlimited Context Language Modeling"

Awesome Papers related to Mamba.

The Fast Cross-Platform Package Manager

A generative speech model for daily dialogue.

Causal depthwise conv1d in CUDA, with a PyTorch interface

Mamba SSM architecture

Mora: More like Sora for Generalist Video Generation

CV-CUDA™ is an open-source, GPU accelerated library for cloud-scale image processing and computer vision.

JAX bindings for Flash Attention v2

Grok open release

Efficiently Fine-Tune 100+ LLMs in WebUI (ACL 2024)

GaLore: Memory-Efficient LLM Training by Gradient Low-Rank Projection

Code for our EMNLP 2023 Paper: "LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of Large Language Models"

Official implementation of "DoRA: Weight-Decomposed Low-Rank Adaptation"