OpenMetricLearning provides metric learning pipelines, models, and a model zoo.
Open source
24 alternativessomoclu
The somoclu project implements self-organizing maps in parallel across CPUs, GPUs, and clusters.
About somoclu
Somoclu is an open-source implementation of self-organizing maps that runs training in parallel. It uses OpenMP on multicore CPUs, MPI to distribute work across a cluster, and CUDA to run on GPUs. A sparse kernel is included for training maps on vector spaces produced by text mining. The program is run from the command line and takes a plain text input file containing either dense or sparse data.
The source notes several limits. The GPU and CPU kernels can produce different maps, because single-precision floats can create ties in distance that are resolved differently. Negative input values are not advised, particularly when the codebook is initialized from 0 to 1. On a single node, OpenMP is more efficient than launching MPI tasks to match the core count. The homepage excerpt is incomplete, so other documentation details are not covered here.
Key features
- Parallel training with OpenMP on CPUs, MPI across clusters, and CUDA on GPUs
- Planar and toroid map shapes
- Rectangular and hexagonal grid layouts
- Gaussian and bubble neighborhood functions
- Support for both dense and sparse input data
- Python, R, Julia, and MATLAB interfaces for dense CPU and GPU kernels
Best for: Developers and researchers who train self-organizing maps on large datasets and want to use multicore CPUs, GPUs, or clusters.
Good for
- Parallel execution on multicore CPUs, GPUs, and clusters
- Runs on Linux, macOS, and Windows
- Includes a sparse kernel for vector spaces generated in text mining
Watch out for
- Negative input values are not advised, especially with a 0-to-1 codebook initialization
- CPU and GPU kernels may produce different maps, since single-precision floats are used
- On a single node, MPI tasks are less efficient than OpenMP threads
somoclu at a glance
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Platforms
- Linux, macOS, Windows
- Open source
- Yes · Source code
- Language
- English
- Last checked
- Oct 6, 2026 · Online when checked
- Listed
- Oct 5, 2026
somoclu alternatives
Tools and sites you could use instead, ranked by how closely they match.
Hugging Face Accelerate launches and trains PyTorch models across devices and distributed configurations.
Open source
24 alternatives
Use Learnergy, a Python library for training and working with energy-based machine learning models.
Open source
24 alternatives
PyTorch Lightning extension that accelerates and enhances foundation model experimentation with flexible fine-tuning schedules.
Open source
24 alternatives
TensorFlow Model Optimization provides Keras and TensorFlow tools for model quantization and pruning.
Open source
24 alternatives
Karate Club offers algorithms for unsupervised learning on graph data in Python.
Open source
24 alternativesPopular in ML Training & Optimization
Well-known listings from the same topic.
Edwardedwardlib.orgAerosolveaerosolve.airbnb.techDataturedatature.ioTogether AItogether.aiOptunaoptuna.orgCatBoostcatboost.ai
somoclu FAQ
What input format does Somoclu accept?
It takes a plain text input file containing either dense or sparse data. Example files are included in the repository.
Does the input data need to be normalized?
With random initialization, the codebook is filled with numbers from 0 to 1. Either supply your own initial codebook or normalize your data to fall in that range.
Will the CPU and GPU versions produce the same map?
The source says the maps are likely to be different. Single-precision floats can produce identical distances, and the CPU picks the match with the lowest coordinates, which the GPU reduction cannot guarantee.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.