Proof of Work
Last updated
Last updated
Proof of Work is a novel consensus mechanism for the iAgent protocol. It incentivizes and validates GPU resources' contribution towards game agent training.
Job initiation A job in the context of the iAgent ComputeHub represents a specific task related to the training or refining of an AI model. Job initiation occurs when a user (such as a gamer or an AI developer) requires intelligent model training. They submit their requirements to the network, which gets converted into a specific job (or series of jobs), with the parameters of the job being defined by the complexity and requirements of the model.
Job training Job-solving involves the actual computational work done by GPU resources in the network. When a job is initiated, it is distributed among the nodes (contributors to GPU resources) in the network. These nodes work on solving the job - training the intelligent model based on the requirements specified during job initiation. The nodes' computational resources are used to run the AI training algorithms, effectively "solving" the job.
Job validation Once a job is completed (i.e., the intelligent model is trained), the results need to be verified to ensure they meet the requirements specified in the job initiation phase. This could involve running the trained model through a series of tests or evaluating its performance in a virtual environment. The results of this validation process determine whether the job has been solved.
Job finalization It involves protocol nodes and recording the completed job results on the blockchain. Proof generation includes the details of the intelligent model, the GPUs that contributed to its training, the protocol nodes that will perform the approval process, and the distribution of $AGNT tokens as rewards and final confirmation of the model. By recording this on the blockchain, the process is transparent, auditable, and resistant to tampering. This ensures the protocol's fairness and reliability.