How this analysis was created: This summary and feature list were written by an AI model that read the project's README and public documentation pages. Each feature links to the documentation it came from; stars, license and language come straight from the GitHub API. The model does not read the source code, and the analysis is refreshed when the project is re-analysed. Learn more on our About page.
Code for the NeurIPS 2022 paper "Optimal Brain Compression: A Framework for Accurate Post-Training Quantization and Pruning".
The main features of ist-daslab/obc are: Weight Pruning.
Open-source alternatives to ist-daslab/obc include: tencent/pocketflow — PocketFlow is an integrated toolkit for deep learning model compression, distributed training, and mobile format… chrundle/biprop — This method identifies a binary weight or binary weight and activation subnetwork within a randomly initialized… dchiji-ntt/iterand — by Daiki Chijiwa\*, Shin’ya Yamaguchi, Yasutoshi Ida, Kenji Umakoshi, Tomohiro Inoue. dem123456789/pruning-deep-neural-networks-from-a-sparsity-perspective — [ICLR 2023] Pruning Deep Neural Networks from a Sparsity Perspective. densoitlab/bitprune — This is the official repo for ICLR 2023 Paper "Bit-Pruning: A Sparse Multiplication-Less Dot-Product" Yusuke Sekikawa… boone891214/sanity-check-lth — Sample code use for NeurIPS 2021 paper: Sanity Checks for Lottery Tickets: Does Your Winning Ticket Really Win the…
PocketFlow is an integrated toolkit for deep learning model compression, distributed training, and mobile format optimization. It provides a system for reducing the size and complexity of neural networks to improve inference efficiency, featuring a dedicated engine for knowledge distillation and a mobile model optimizer. The framework differentiates itself through an automated hyperparameter tuning system that uses reinforcement learning and statistical models to determine optimal compression ratios and layer-wise bit allocation. It also includes a distributed training system that utilizes mu
This method identifies a binary weight or binary weight and activation subnetwork within a randomly initialized network that achieves performance comparable to, and sometimes better than, a weight-optimized network. The resulting binarized and pruned networks that achieve comparable performance…
by Daiki Chijiwa\*, Shin’ya Yamaguchi, Yasutoshi Ida, Kenji Umakoshi, Tomohiro Inoue
Sample code use for NeurIPS 2021 paper: Sanity Checks for Lottery Tickets: Does Your Winning Ticket Really Win the Jackpot?