Machine Learning · GitHub

Machine learning Guide. Learn all about Machine Learning Tools, Libraries, Frameworks, Large Language Models (LLMs), and Training Models.

docTR (Document Text Recognition) - a seamless, high-performing & accessible library for OCR-related tasks powered by Deep Learning. Ongoing development and maintenance by t2k.

Companion code for Machine Learning From Scratch — 10 core ML algorithms built from scratch with NumPy, compared with Scikit-learn and PyTorch.

This repository contains the best Data Science free hand-picked resources to equip you with all the industry-driven skills and interview preparation kit.

Fast and Accurate One-Stage Space-Time Video Super-Resolution (accepted in CVPR 2020)

Code for Tensorflow Machine Learning Cookbook

Transformer related optimization, including BERT, GPT

SOMA: Unifying Parametric Human Body Models

🌊 Online machine learning in Python

• Linear models, with a wide array of optimizers
• Decision trees and random forests

GammaCV is a WebGL accelerated Computer Vision library for browser

This repository contains a hand-curated resources for Prompt Engineering with a focus on Generative Pre-trained Transformer (GPT), ChatGPT, PaLM etc

The official implementation for ICLR2023 spotlight paper "DIFFormer: Scalable (Graph) Transformers Induced by Energy Constrained Diffusion"

The official implementation for ICLR2023 paper "GNNSafe: Energy-based Out-of-Distribution Detection for Graph Neural Networks"

The official implementation of NeurIPS2022 spotlight paper "NodeFormer: A Scalable Graph Structure Learning Transformer for Node Classification"

• All-Pair Message Passing on Layer-specific Adaptive Structures. The feature propagation per layer is operated on a latent graph that potentially connect all the nodes, in contrast with the local propagation design of GNNs that only aggregates the embeddings of neighbored nodes.
• Efficient End-to-End Learning for Latent Structures. The optimization for the latent structures is allowed for end-to-end training with the model, making the whole learning process simple and efficient. E.g., the training on Cora only requires 1-2 minutes, while on OGBN-Proteins requires 1-2 hours in one run.

NanoDet-Plus⚡Super fast and lightweight anchor-free object detection model. 🔥Only 980 KB(int8) / 1.8MB (fp16) and run 97FPS on cellphone🔥

Zero-shot clone-tuner for Kokoro-82M. Generate stock-compatible custom voice packs from 5-seconds of reference audio in less than 1-second

Scientific machine learning (SciML) benchmarks, AI for science, and (differential) equation solvers. Covers Julia, Python (PyTorch, Jax), MATLAB, R

Statistical learning methods, 统计学习方法(第2版)[李航] [笔记, 代码, notebook, 参考文献, Errata, lihang]

Silero Models: pre-trained text-to-speech models made embarrassingly simple