disTensor is a platform for decentralized distributed training of large-scale machine learning models, leveraging heterogeneous consumer compute resources. It supports explicit 4D parallelism and PyTorch semantics, making it suitable for ML engineers seeking scalable, cost-effective training solutions.
disTensor sits in PulseGate's Data science & ML workbench category. It focuses on enabling scalable, decentralized distributed training for large-scale machine learning on heterogeneous consumer hardware. disTensor is a B2B product aimed at machine learning engineers. The product ships for the web and the command line.
disTensor first shipped in 2025. The project is developed in the open on GitHub with 314 commits in the last 90 days. Across PulseGate's embedding index, disTensor has few near neighbours, marking it as relatively distinct. Among its 5 catalogued features are distributed training, decentralized compute, and pyTorch semantics.
Latest indexed changes and source events
distensor.com discovered by the PulseGate indexer
Other apps tracked under the same category.
