sputnik

Sparse matrix optimizer

A library of optimized GPU kernels for sparse matrix operations used in deep learning.

A library of GPU kernels for sparse matrix operations.

GitHub

248 stars
10 watching
51 forks
Language: C++
last commit: almost 6 years ago
Linked from 1 awesome list


Backlinks from these awesome lists:

Related projects:

RepositoryDescriptionStars
hiroyuki-kasai/sparsegdlibraryA collection of unconstrained optimization algorithms for sparse modeling in MATLAB53
neuralmagic/sparsemlEnables the creation of smaller neural network models through efficient pruning and quantization techniques2,083
gdikov/hypertunityA toolset for optimizing hyperparameters of machine learning models using Bayesian optimization and real-time visualization.136
gpleiss/efficient_densenet_pytorchAn implementation of DenseNets optimized to save GPU memory by using checkpointing1,524
google/jaxoptAn open-source project providing hardware accelerated, batchable and differentiable optimizers in JAX for deep learning.941
google-deepmind/optaxA gradient processing and optimization library designed to facilitate research and productivity in machine learning by providing building blocks for custom optimizers and gradient processing components.1,730
sparsemat/sprsA Rust library implementing sparse matrix data structures and linear algebra algorithms431
stanfordnqp/spins-bAn optimization library for photonic devices using gradient-based methods.291
james-bowman/sparseA package implementing various sparse matrix formats for efficient linear algebra operations160
dppalomar/sparseindextrackingA package for computing sparse portfolios of assets to track an index.51
mengrao/strAn optimized string class with an adaptive hash table for fast searching and comparisons126
indigits/sparse-plexA collection of algorithms and techniques for exploiting sparse data representations to solve various problems in signal processing and machine learning.99
google/llvm-propellerAn optimization tool for improving the performance of large-scale software recompilation using profiling and relinking techniques.381
datacanvasio/hypergbmAutomated machine learning tool for tabular data pipelines343
hyperopt/hyperopt-sklearnAutomates search for optimal parameters in machine learning algorithms.1,594