o -iX@s,ddlmZmZddlZGdddeZdS))ABCabstractmethodNc@sVeZdZee d dejdededededejdBdeejejejffd d Z dS) AbstractEplbPolicyNweight num_replicas num_groups num_nodes num_ranksold_global_expert_indicesreturncCst)a  Entry point for expert-parallelism load balancer. Parameters: weight: [layers, num_logical_experts], the load statistics for all logical experts num_replicas: number of physical experts, must be a multiple of `num_ranks` num_groups: number of expert groups num_nodes: number of server nodes num_ranks: number of ranks, must be a multiple of `num_nodes` old_global_expert_indices: [layers, num_logical_experts], the old global expert indices. Used to avoid unnecessary weight copying for experts moving within one rank. Returns: physical_to_logical_map: [layers, num_replicas], the expert index of each replica logical_to_physical_map: [layers, num_logical_experts, X], the replica indices for each expert expert_count: [layers, num_logical_experts], number of physical replicas for each logical expert )NotImplementedError)clsrrrrr r rb/home/ubuntu/veenaModal/venv/lib/python3.10/site-packages/vllm/distributed/eplb/policy/abstract.pyrebalance_experts s!z$AbstractEplbPolicy.rebalance_experts)N) __name__ __module__ __qualname__ classmethodrtorchTensorinttuplerrrrrr s&r)abcrrrrrrrrs