karpathy

followers

following

stars

Stanford

https://twitter.com/karpathy

Andrej's starred repositories

build-your-own-x

Master programming by recreating your favorite technologies from scratch.

256617 4976 609

langchain

🦜🔗 Build context-aware reasoning applications

Language:PythonMIT82602 645 6550

whisper

Robust Speech Recognition via Large-Scale Weak Supervision

Language:PythonMIT59946 5050

tesseract

Tesseract Open Source OCR Engine (main repository)

Language:C++Apache-2.057884 1681 2595

Open-Assistant

OpenAssistant is a chat-based assistant that understands tasks, can interact with third-party systems, and retrieve information dynamically to do so.

Language:PythonApache-2.036621 422 1643

llama_index

LlamaIndex is a data framework for your LLM applications

Language:PythonMIT30743 224 3636

stanford_alpaca

Code and documentation to train Stanford's Alpaca models, and generate the data.

Language:PythonApache-2.028743 336 266

diffusers

🤗 Diffusers: State-of-the-art diffusion models for image and audio generation in PyTorch and FLAX.

Language:PythonApache-2.022421 181 3467

vllm

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

Language:PythonApache-2.017934 178 2437

dalle-mini

DALL·E Mini - Generate images from a text prompt

Language:PythonApache-2.014632 111 155

mlx

MLX: An array framework for Apple silicon

Language:C++MIT13942 124 359

triton

Development repository for the Triton language and compiler

Language:C++MIT10895 177 1144

xformers

Hackable and optimized Transformers building blocks, supporting a composable construction.

Language:PythonNOASSERTION7522 74 463

TinyLlama

The TinyLlama project is an open endeavor to pretrain a 1.1B Llama model on 3 trillion tokens.

Language:PythonApache-2.06731 109 135

litgpt

Pretrain, finetune, deploy 20+ LLMs on your own data. Uses state-of-the-art techniques: flash attention, FSDP, 4-bit, LoRA, and more.

Language:PythonApache-2.06526 69 586

xv6-riscv

Xv6 for RISC-V

Language:CNOASSERTION6116 94 84

unsloth

2-5X faster 80% less memory LLM finetuning

Language:PythonApache-2.05780 49 226

lit-llama

Implementation of the LLaMA language model based on nanoGPT. Supports flash attention, Int8 and GPTQ 4bit quantization, LoRA and LLaMA-Adapter fine-tuning, pre-training. Apache 2.0-licensed.

Language:PythonApache-2.05779 68 264

gpt-fast

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

Language:PythonBSD-3-Clause5056 57 80

gemma_pytorch

The official PyTorch implementation of Google's Gemma models

Language:PythonApache-2.05009 38 32

AlphaCodium

Official implementation for the paper: "Code Generation with AlphaCodium: From Prompt Engineering to Flow Engineering""

Language:PythonAGPL-3.03020 43 14

cramming

Cramming the training of a (BERT-type) language model into limited compute.

Language:PythonMIT1231 22 32

hlb-CIFAR10

Train to 94% on CIFAR-10 in <6.3 seconds on a single A100. Or ~95.79% in ~110 seconds (or less!)

Language:PythonApache-2.01187 20 3

VQ-Diffusion

Official implementation of VQ-Diffusion

Language:PythonMIT831 11 34

yet-another-applied-llm-benchmark

A benchmark to evaluate language models on questions I've previously asked them to solve.

Language:PythonGPL-3.075900

fine-tune-mistral

Fine-tune mistral-7B on 3090s, a100s, h100s

Language:PythonMIT679 6 5

inbox_cleaner

A python script to help manage a Gmail inbox by filtering out promotional emails using GPT-3 or GPT-4.

Language:Python389 5 6

llm_rules

RuLES: a benchmark for evaluating rule-following in language models

Language:PythonApache-2.0191 1 1

bpeasy

Fast bare-bones BPE for modern tokenizer training

Language:PythonMIT125 10

vit-vqgan

JAX implementation ViT-VQGAN

Language:PythonMIT51 20