🧑‍🏫 60+ Implementations/tutorials of deep learning papers with side-by-side notes 📝; including transformers (original, xl, switch, feedback, vit, ...), optimizers (adam, adabelief, sophia, ...), ga…

Nano vLLM

Accessible large language models via k-bit quantization for PyTorch.

DFlash: Block Diffusion for Flash Speculative Decoding

[ECCV2024] IDM-VTON : Improving Diffusion Models for Authentic Virtual Try-on in the Wild

[CAAI AIR'24] Bilateral Reference for High-Resolution Dichotomous Image Segmentation

[MLSys 2024 Best Paper Award] AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration

Visual Causal Flow

MarS: a Financial Market Simulation Engine Powered by Generative Foundation Model

Image Test Time Augmentation with PyTorch!

A lightweight inference engine supporting speculative speculative decoding (SSD).

⚡ Fastest way to serve open source ML models to millions

ARC Relay — WebSocket relay server for agent remote control by Axolotl AI

wubuwizard: The best inference engine

Python package for processing deep learning models

cpm is command line tool, it's use to create c++ project and manage

RAG model implemented from scratch given a pdf it will answer question based on the given context

PyTorch adapted implementation of "Translating Videos to Commands for Robotic Manipulation with Deep Recurrent Neural Networks" in ICRA 2018

ML papers Implementation

vision language model

Small scale distributed training of sequential deep learning models, built on Numpy and MPI.

A bot with gemma api that talks likes wrote in streamlit

video to music

Trained a GPT model from scratch on shakespeare's plays just generates random Play scripts it will need fit it and some post processing to deploy

High-performance Rust extensions for Axolotl (no OOM for large datasets) - drop-in acceleration for existing installations.

Mixture-of-Experts (MoE) Router Optimization project. This is an excellent project that touches on cutting-edge LLM optimization techniques

Transformer from scratch