TensorRT-LLMs/cpp/tensorrt_llm/thop/userbuffersFinalizeOp.cpp
Kaiyu Xie 2631f21089
Update (#2978)
Signed-off-by: Kaiyu Xie <26294424+kaiyux@users.noreply.github.com>
2025-03-23 16:39:35 +08:00

64 lines
2.3 KiB
C++

/*
* Copyright (c) 2025, NVIDIA CORPORATION. All rights reserved.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "tensorrt_llm/common/dataType.h"
#include "tensorrt_llm/kernels/userbuffers/ub_interface.h"
#include "tensorrt_llm/runtime/torchUtils.h"
#include "tensorrt_llm/thop/thUtils.h"
#include <torch/extension.h>
namespace torch_ext
{
torch::Tensor userbuffers_allreduce_finalize(torch::Tensor input, bool force_applying_finalize)
{
#if ENABLE_MULTI_DEVICE
auto stream = at::cuda::getCurrentCUDAStream(input.get_device());
auto output = torch::empty_like(input);
size_t size = input.numel();
int hidden_size = input.size(-1);
TLLM_CHECK_WITH_INFO(tensorrt_llm::runtime::ub::ub_is_initialized(), "UserBuffer has not been initialized!");
auto ub_buffer1 = tensorrt_llm::runtime::ub::ub_get(1);
auto ub_comm = tensorrt_llm::runtime::ub::ub_comm();
auto const dtype = tensorrt_llm::runtime::TorchUtils::dataType(input.scalar_type());
tensorrt_llm::kernels::ub::allgather2_userbuff_residual_launcher(
ub_buffer1.handle, 0, size, hidden_size, input.data_ptr(), dtype, ub_comm, stream, force_applying_finalize);
auto const elem_size = tensorrt_llm::common::getDTypeSize(dtype);
// Force memcpy to avoid ub tensor exposed to torch graph.
TLLM_CUDA_CHECK(
cudaMemcpyAsync(output.data_ptr(), ub_buffer1.addr, size * elem_size, cudaMemcpyDeviceToDevice, stream));
return output;
#else
return input;
#endif // ENABLE_MULTI_DEVICE
}
} // namespace torch_ext
TORCH_LIBRARY_FRAGMENT(trtllm, m)
{
m.def("userbuffers_allreduce_finalize(Tensor input, bool force_applying_finalize) -> Tensor");
}
TORCH_LIBRARY_IMPL(trtllm, CUDA, m)
{
m.impl("userbuffers_allreduce_finalize", &torch_ext::userbuffers_allreduce_finalize);
}