在 马来西亚 購買 τemplar (SN3)

Templar, also known as Subnet 3 or SN3, is a decentralized project operating within the Bittensor ecosystem. Its primary mission is to establish a permissionless and incentivized platform for the distributed training of large scale artificial intelligence models. By leveraging the infrastructure of the Bittensor network, Templar aims to provide an open source alternative to the centralized computing power typically controlled by large corporate laboratories. The core functionality of Templar revolves around decentralized pre training. In this system, participants known as miners contribute their own computing resources to train foundational machine learning models. These miners submit gradients, which are essentially updates that improve the performance of a model, to the network. Another group of participants, called validators, then assesses these contributions to ensure they actually enhance the model. This process uses incentive mechanisms designed to reward high quality work, effectively turning AI training into a competitive, global effort similar to how decentralized mining secures other blockchains. Technologically, the project incorporates advanced research to manage the challenges of training models over a distributed internet connection. It utilizes methods such as SparseLoCo for communication efficient pre training, which helps reduce the amount of data that needs to be transferred between different parts of the network. The project also explores using hosting services for verifiable timestamps to coordinate the submission of training data across a global group of miners. The vision behind Templar is to democratize access to high tier intelligence by allowing a diverse community to pool resources and co own the resulting models. This approach is intended to foster an anti fragile and sovereign AI ecosystem that is not dependent on any single entity. Instead of being locked behind proprietary walls, the models developed on this subnet are meant to be shared resources that can be used or monetized by the community members who help build and secure them. In essence, the project seeks to transform AI development from a centralized corporate activity into a transparent, community driven digital commodity. 瞭解更多