Popular GitHub Repositories for Learned Optimizer

The largest collection of PyTorch image encoders / backbones. Including train, eval, inference, export scripts, and pretrained weights -- ResNet, ResNeXT, EfficientNet, NFNet, Vision Transformer (ViT), MobileNetV4, MobileNet-V3 & V2, RegNet, DPN, CSPNet, Swin Transformer, MaxViT, CoAtNet, ConvNeXt, and more

AdalFlow: The library to build & auto-optimize LLM applications.

torch-optimizer -- collection of optimizers for Pytorch

? Lion, new optimizer discovered by Google Brain using genetic algorithms that is purportedly better than Adam(w), in Pytorch

Adlik: Toolkit for Accelerating Deep Learning Inference

Adan: Adaptive Nesterov Momentum Algorithm for Faster Optimizing Deep Models

Implementation of MEGABYTE, Predicting Million-byte Sequences with Multiscale Transformers, in Pytorch

The nnsight package enables interpreting and manipulating the internals of deep learned models.

TensorFlow Implementation of Generative Adversarial Networks for Extreme Learned Image Compression

Object detection on multiple datasets with an automatically learned unified label space.

AI Image Signal Processing and Computational Photography. Official library for NTIRE (CVPR) and AIM (ICCV/ECCV) Challenges. You will find Learned ISPs, RAW Restoration-Upsampling-Reconstruction, Image Enhancement, Bokeh rendering and more!

Code for Adam-mini: Use Fewer Learning Rates To Gain More https://arxiv.org/abs/2406.16793

Model Compression Toolkit (MCT) is an open source project for neural network model optimization under efficient, constrained hardware. This project provides researchers, developers, and engineers advanced quantization and compression tools for deploying state-of-the-art neural networks.

Visualize Tensorflow's optimizers.

Effortless plugin and play Optimizer to cut model training costs by 50%. New optimizer that is 2x faster than Adam on LLMs.

Awesome-LLM-Prompt-Optimization: a curated list of advanced prompt optimization and tuning methods in Large Language Models

Deep learning library in plain Numpy.

Multi-hop knowledge graph reasoning learned via policy gradient with reward shaping and action dropout

Deep learned, NVIDIA-accelerated 3D object pose estimation

syftr is an agent optimizer that helps you find the best agentic workflows for your budget.