Train and run models with OneFlow, a scalable deep learning framework for distributed machine learning workloads.
Open source
24 alternativesSinga
Apache SINGA supports distributed deep learning through a system and library for model training.
About Singa
Apache SINGA is a distributed deep learning system for training deep learning and machine learning models. It offers an easy-to-use Python interface, records the computation graph to apply back propagation automatically after forward propagation, and supports data parallel training across multiple GPUs on a single node or across different nodes. Memory optimization and time profiling of graph operators are implemented in the system, and several popular optimizers such as stochastic gradient descent with momentum, Adam, RMSProp and AdaGrad are available.
Trained models can be queried as stored procedures of a relational database, and the project provides domain-specific models in its GitHub repository and on Google Colab. SINGA can load ONNX format models and save models defined with its APIs into ONNX format, which helps move models between libraries and tools. Installation options are Conda, Pip, Docker, or building from source. The homepage lists no pricing information and does not describe geographic eligibility limits.
Key features
- Data parallel training across multiple GPUs on one or several nodes
- Automatic gradient calculation from a recorded computation graph
- Installation via Conda, Pip, Docker, or from source
- Optimizers including SGD with momentum, Adam, RMSProp and AdaGrad
- Loading and saving models in ONNX format for interoperability
- Trained models queried as stored procedures in an RDBMS
Best for: Developers and researchers who train deep learning models across multiple GPUs or nodes and want a Python interface with database-queryable model output.
Good for
- Supports data parallel training across multiple GPUs on a single node or across different nodes
- Provides an easy-to-use Python interface over a structured software stack
- Offers several installation routes, including Conda, Pip, Docker and source builds
Watch out for
- Homepage content is provided in English only
- No pricing information is stated on the homepage
Singa at a glance
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Platforms
- Web
- Open source
- Not stated
- Language
- English
- Last checked
- Oct 5, 2026 · Online when checked
- Listed
- Oct 5, 2026
Singa alternatives
Tools and sites you could use instead, ranked by how closely they match.
Flexible, high-performance serving system for machine learning models.
Open source
24 alternatives
DJL is a Java framework for running deep learning models across different engines.
Open source
24 alternatives
DeepDetect provides a deep learning server and command-line interface for Torch and TensorRT.
dstack coordinates AI compute and supports training and inference across GPU clouds, Kubernetes, VMs, and bare metal.
Open source
24 alternatives
ORMB packages machine learning models as OCI artifacts, in a workflow described as Docker for ML models.
Open source
24 alternativesPopular in AI Infrastructure & MLOps
Well-known listings from the same topic.
Guild AIguild.aiFlow AIflow-ai.comsystemprompt.iosystemprompt.ioCoreWeavecoreweave.comKili Technologykili-technology.comAPI docsdocs.x.ai
Singa FAQ
How can Apache SINGA be installed?
The project lists installation using Conda, Pip, Docker, or from source.
Can SINGA models be used with other libraries and tools?
SINGA can load ONNX format models and can save models defined with SINGA APIs into ONNX format, which supports use across different libraries and tools.
Does SINGA support training on multiple GPUs?
Yes. SINGA supports data parallel training across multiple GPUs, either on a single node or across different nodes.
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.