TensorRT-LLMs/tensorrt_llm/_torch/distributed/ops.py

import os
import threading
from typing import Optional, Tuple, Union

import torch
from torch import nn

from tensorrt_llm.functional import AllReduceParams, AllReduceStrategy
from tensorrt_llm.mapping import Mapping
from tensorrt_llm.plugin.plugin import CustomAllReduceHelper

_thread_local = threading.local()


def get_allreduce_workspace(mapping: Mapping) -> torch.LongTensor:
    if not hasattr(_thread_local, f'allreduce_workspaces_{mapping.pp_rank}'):
        setattr(_thread_local, f'allreduce_workspaces_{mapping.pp_rank}', {})

    allreduce_workspaces = getattr(_thread_local,
                                   f'allreduce_workspaces_{mapping.pp_rank}')
    if mapping not in allreduce_workspaces:
        ipc_buffers, workspace = CustomAllReduceHelper.allocate_allreduce_fusion_workspace(
            mapping,
            CustomAllReduceHelper.max_workspace_size_auto(
                mapping.tp_size, support_deterministic=False),
        )
        allreduce_workspaces[mapping] = (ipc_buffers, workspace)
    return allreduce_workspaces[mapping][1]


def allocate_low_presicion_allreduce_workspace(mapping: Mapping) -> None:
    if not hasattr(_thread_local, 'lowprecision_allreduce_workspaces'):
        _thread_local.lowprecision_allreduce_workspaces = {}
    lowprecision_allreduce_workspaces = _thread_local.lowprecision_allreduce_workspaces
    if mapping not in lowprecision_allreduce_workspaces:
        ipc_buffers, workspace = CustomAllReduceHelper.allocate_lowprecision_workspace(
            mapping,
            CustomAllReduceHelper.max_workspace_size_lowprecision(
                mapping.tp_size),
        )
        lowprecision_allreduce_workspaces[mapping] = (ipc_buffers, workspace)
        CustomAllReduceHelper.initialize_lowprecision_buffers(
            workspace, mapping.tp_size)
    return


def userbuffers_allreduce_finalize(
        input: torch.Tensor,
        force_applying_finalize: bool = False) -> torch.Tensor:
    output = torch.ops.trtllm.userbuffers_allreduce_finalize(
        input, force_applying_finalize)
    return output


def allgather(input: torch.Tensor,
              mapping: Mapping,
              gather_dim: int = -1) -> torch.Tensor:
    '''
    Add an operation that performs a collective all-gather.

    The input tensors in the different ranks must have the same shape.
    The output tensor will be replicated among the TP group.

    Given the 'section_size = input.shape[gather_dim]', each rank
    contributes a section of its input tensor that correspond to
    'rank*section_size:(rank+1)*section_size',
    and 'output.shape[gather_dim] = input.shape[gather_dim] * tp_group_size'.

    That operation is implemented using a torch op that wraps the NCCL all-gather
    collective operation. See
    https://docs.nvidia.com/deeplearning/nccl/user-guide/docs/usage/collectives.html#allgather
    for details.

    Args:
        input (Tensor): The input tensor.
        mapping (Mapping):  The parallel mapping.
        gather_dim (int): Gather along given dimension. By default -1.
    Returns:
        The gathered tensor.
    '''
    if mapping.tp_size == 1:
        return input

    output = torch.ops.trtllm.allgather(
        input,
        mapping.tp_group,
    )

    if gather_dim < 0:
        gather_dim += input.ndim

    output = torch.movedim(output, 0, gather_dim)
    input_shape = input.size()
    output = output.reshape(input_shape[:gather_dim] +
                            (mapping.tp_size * input_shape[gather_dim], ) +
                            input_shape[gather_dim + 1:])
    return output


def reducescatter(input: torch.Tensor,
                  mapping: Mapping,
                  scatter_dim: int = -1) -> torch.Tensor:
    if mapping.tp_size == 1:
        return input

    output = torch.ops.trtllm.reducescatter(
        input,
        mapping.tp_group,
    )

    if scatter_dim < 0:
        scatter_dim += input.ndim

    output = torch.movedim(output, 0, scatter_dim)
    input_shape = input.size()
    output = output.reshape(input_shape[:scatter_dim] +
                            (input_shape[scatter_dim] // mapping.tp_size, ) +
                            input_shape[scatter_dim + 1:])
    return output


class AllReduce(nn.Module):

    def __init__(self,
                 mapping: Mapping,
                 strategy: AllReduceStrategy = AllReduceStrategy.AUTO):
        super().__init__()
        """
        AllReduce is a module that performs an all-reduce operation on a tensor.

        Args:
            mapping (Mapping):  The parallel mapping config.
            strategy (AllReduceStrategy):
                The following all-reduce strategies are supported:

                - UB: AllReduce uses user-buffer based all-reduce kernel.

                - NCCL: Use NCCL allreduce.

                - MIN_LATENCY: AllReduce uses MIN_LATENCY mode kernel.

                - AUTO: AUTO chooses between NCCL and MIN_LATENCY mode based on a heuristic policy.

                - LOWPRECISION: AllReduce quantizes data to lower precision for transmission.
                  Should only be used on topologies with PCIe switches and without NVLink.
                  This strategy may result in some precision loss but can improve performance
                  on specific hardware configurations.

            All strategies support the following operations:
                - NONE (AllReduce only)
                - RESIDUAL_RMS_NORM
                - RESIDUAL_RMS_NORM_QUANT_FP8
                - RESIDUAL_RMS_NORM_QUANT_NVFP4
                - RESIDUAL_RMS_NORM_OUT_QUANT_FP8
                - RESIDUAL_RMS_NORM_OUT_QUANT_NVFP4

            Note: NCCL, UB, and LOWPRECISION strategies only support consequent kernel calls
        instead of fused operations.

        Note:
            For the reference implementation for each pattern, please refer to the following unit test:
            https://github.com/NVIDIA/TensorRT-LLM/blob/main/tests/unittest/_torch/multi_gpu/test_allreduce.py

            The LOWPRECISION strategy can be selected either by directly specifying it in the constructor
            or by setting the environment variable FORCE_LOW_PRECISION_ALL_REDUCE_STRATEGY when using
            the AUTO strategy.
        """

        self.mapping = mapping
        self.workspace = None
        self.strategy = strategy

        self.force_low_precision_env = os.environ.get(
            "FORCE_LOW_PRECISION_ALL_REDUCE_STRATEGY")
        if self.mapping.tp_size > 1:
            # When Strategy is UB, it is guaranteed that the workspace is not used.
            if self.strategy != AllReduceStrategy.UB:
                if self.strategy == AllReduceStrategy.LOWPRECISION or self.force_low_precision_env is not None:
                    allocate_low_presicion_allreduce_workspace(self.mapping)
                self.workspace = get_allreduce_workspace(self.mapping)

    def forward(
        self,
        input: torch.Tensor,
        *,
        all_reduce_params: Optional[AllReduceParams] = None,
    ) -> Union[torch.Tensor, Tuple[torch.Tensor, ...]]:
        '''
        The input tensors in the different ranks must have the same shape.
        The output tensor will have that same shape with the input tensor.
        The output tensor will be replicated among the TP group.
        Note that it is not an in-place operation like torch.distributed.all_reduce.

        That operation is implemented using a torch op that wraps the NCCL all-reduce
        collective operation and custom one-shot/two-shot allreduce kernels. See
        https://docs.nvidia.com/deeplearning/nccl/user-guide/docs/usage/collectives.html#allreduce
        for details.

        Args:
            input (Tensor): The input tensor.
            all_reduce_params (AllReduceParams): The parameters for the fused ops into the allreduce op.
        Returns:
            A tensor lists with different tensor outptus according to the fusion_op.
            NONE: [hidden_states]
            RESIDUAL_RMS_NORM: [hidden_states, residual]
            RESIDUAL_RMS_NORM_QUANT_FP8: [norm_quant, residual]
            RESIDUAL_RMS_NORM_OUT_QUANT_FP8: [norm, norm_quant, residual]
            RESIDUAL_RMS_NORM_QUANT_NVFP4: [norm_quant_fp4, scale_factor, residual]
            RESIDUAL_RMS_NORM_OUT_QUANT_NVFP4: [norm, norm_quant_fp4, scale_factor, residual]
        '''
        if self.mapping.tp_size == 1 or (all_reduce_params is not None
                                         and all_reduce_params.enable_allreduce
                                         == False):
            return input

        # Assume using no fusion allreduce here
        if all_reduce_params is None:
            all_reduce_params = AllReduceParams()

        output = torch.ops.trtllm.allreduce(
            input=input,
            residual=all_reduce_params.residual,
            norm_weight=all_reduce_params.norm_weight,
            scale=all_reduce_params.scale,
            bias=all_reduce_params.bias,
            workspace=self.workspace,
            group=self.mapping.tp_group,
            strategy=self.strategy,
            op=all_reduce_params.fusion_op,
            eps=all_reduce_params.eps,
        )

        return output if len(output) > 1 else output[0]


class MoEAllReduce(nn.Module):

    def __init__(self, mapping: Mapping):
        """
        MoEAllReduce is a module that performs a specific fused MoE reduction
        followed by a regular AR + RMS norm.

        Args:
            mapping (Mapping):  The parallel mapping config.

        Notes:
            Support pattern: MoE Reduction + Add + AR + ADD_RMS, see this torch reference implementation:
            expert_reduction = torch.sum(active_experts_token_input *
                                        scale.unsqueeze(-1),
                                        dim=0)
            output_add = expert_reduction + shared_expert_output
            output_residual = output_add + residual
            output_hidden_states = rms_norm(output_residual, norm_weight, eps)
        """
        super().__init__()
        self.mapping = mapping
        self.workspace = get_allreduce_workspace(self.mapping)

    def forward(
        self,
        residual: torch.Tensor,
        norm_weight: torch.Tensor,
        device_num_experts: torch.Tensor,
        scale_input: torch.Tensor,
        active_experts_token_input: torch.Tensor,
        token_input: torch.Tensor,
        eps: float,
    ) -> torch.Tensor:
        """
        Args:
            residual: residual tensor
            norm_weight: RMS norm weight
            device_num_experts: number of experts per device
            scale_input: experts to token score
            active_experts_token_input: per token per expert input
            token_input: per token input, shared expert output
            eps: epsilon for RMSNorm

        Output:
            hidden_states: hidden_states of the model
            residual: residual tensor
        """
        return torch.ops.trtllm.moe_allreduce(
            residual=residual,
            norm_weight=norm_weight,
            device_num_experts=device_num_experts,
            scale_input=scale_input,
            active_experts_token_input=active_experts_token_input,
            token_input=token_input,
            workspace=self.workspace,
            rank=self.mapping.tp_rank,
            nranks=self.mapping.tp_size,
            eps=eps,
        )