Tensorplex

Tensorplex Labs develops applications, infrastructure, and tools intended to accelerate AI development and usage through the application of decentralised technologies. The organisation is currently working on projects such as a parameter-efficient checkpoint merging method that utilises metrics-weighted averaging to improve model performance. Their work aims to drive progress in open-source AI by providing the underlying technical foundations required for various AI use cases. The focus remains on building robust infrastructure rather than offering direct user-facing applications at this stage.