Typings

class rizemind.strategies.compensation.typings.SupportsDistribute(*args, **kwargs)

Bases: Protocol

Protocol for objects that can distribute compensation to federated learning trainers.

This protocol defines the interface for reward distribution mechanisms in a federated learning system. Implementers are responsible for distributing compensation to trainers based on their participation scores.

The protocol supports blockchain-based reward distribution where trainers are identified by their Ethereum addresses.

distribute(trainer_scores: list[tuple[ChecksumAddress, float]]) → str

Distribute compensation to trainers based on their scores.

Parameters:

• trainer_scores – List of tuples containing trainer addresses and their

• Ethereum (corresponding compensation scores. Addresses must be valid)

• addresses. (checksum)

Returns:

Transaction hash or confirmation string of the next round operation.