Li Dong (donglixp)

donglixp

Geek Repo

Company:Microsoft Research

Home Page:http://dong.li

Github PK Tool:Github PK Tool


Organizations
AGI-Team

Li Dong's starred repositories

generative-models

Generative Models by Stability AI

Language:PythonLicense:MITStargazers:21605Issues:226Issues:236
Language:PythonLicense:Apache-2.0Stargazers:8193Issues:92Issues:239

text-generation-inference

Large Language Model Text Generation Inference

Language:PythonLicense:NOASSERTIONStargazers:7365Issues:95Issues:958

chatgpt_system_prompt

A collection of GPT system prompts and various prompt injection/leaking knowledge.

Language:HTMLLicense:MITStargazers:7175Issues:75Issues:7

promptbase

All things prompt engineering

Language:PythonLicense:MITStargazers:4971Issues:55Issues:12

gpt-fast

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

Language:PythonLicense:BSD-3-ClauseStargazers:4970Issues:58Issues:77

alignment-handbook

Robust recipes to align language models with human and AI preferences

Language:PythonLicense:Apache-2.0Stargazers:3554Issues:110Issues:101

consistencydecoder

Consistency Distilled Diff VAE

Language:PythonLicense:MITStargazers:2042Issues:21Issues:18

llm-awq

AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration

Language:PythonLicense:MITStargazers:1686Issues:23Issues:130

TransformerEngine

A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper and Ada GPUs, to provide better performance with lower memory utilization in both training and inference.

Language:PythonLicense:Apache-2.0Stargazers:1356Issues:32Issues:178

AutoAWQ

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

Language:PythonLicense:MITStargazers:1083Issues:11Issues:250

MetaCLIP

ICLR2024 Spotlight: curation/training code, metadata, distribution and pre-trained models for MetaCLIP.

Language:PythonLicense:NOASSERTIONStargazers:965Issues:13Issues:19
Language:PythonLicense:Apache-2.0Stargazers:852Issues:7Issues:0

GPT-4V-Act

AI agent using GPT-4V(ision) capable of using a mouse/keyboard to interact with web UI

mesh-gpt

MeshGPT: Generating Triangle Meshes with Decoder-Only Transformers

Mini-Conf

Run a conference from your backyard.

Language:JavaScriptLicense:MITStargazers:534Issues:14Issues:49

faiss_tips

Some useful tips for faiss

Language:ShellLicense:MITStargazers:531Issues:8Issues:4

datablations

Scaling Data-Constrained Language Models

Language:Jupyter NotebookLicense:Apache-2.0Stargazers:278Issues:36Issues:5

lost-in-the-middle

Code and data for "Lost in the Middle: How Language Models Use Long Contexts"

Language:PythonLicense:MITStargazers:247Issues:5Issues:13

MMMU

This repo contains evaluation code for the paper "MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI"

Language:PythonLicense:Apache-2.0Stargazers:218Issues:3Issues:18

flash-fft-conv

FlashFFTConv: Efficient Convolutions for Long Sequences with Tensor Cores

Language:C++License:Apache-2.0Stargazers:195Issues:17Issues:18

InfiniteBench

100k+ Long-Context Benchmark for Large Language Models (paper upcoming)

Language:PythonLicense:MITStargazers:165Issues:7Issues:11

intercode

[NeurIPS 2023 D&B] Code repository for InterCode benchmark https://arxiv.org/abs/2306.14898

Language:PythonLicense:MITStargazers:162Issues:9Issues:12

causal-conv1d

Causal depthwise conv1d in CUDA, with a PyTorch interface

Language:CudaLicense:BSD-3-ClauseStargazers:120Issues:2Issues:11

catwalk

This project studies the performance and robustness of language models and task-adaptation methods.

Language:PythonLicense:Apache-2.0Stargazers:116Issues:6Issues:24

SmartPlay

SmartPlay is a benchmark for Large Language Models (LLMs). Uses a variety of games to test various important LLM capabilities as agents. SmartPlay is designed to be easy to use, and to support future development of LLMs.

Language:PythonLicense:CC-BY-4.0Stargazers:87Issues:4Issues:6

skill-it

Skill-It! A Data-Driven Skills Framework for Understanding and Training Language Models

Language:Jupyter NotebookLicense:Apache-2.0Stargazers:30Issues:13Issues:1
Language:RustLicense:Apache-2.0Stargazers:19Issues:5Issues:6