mirror of
https://github.com/NVIDIA/TensorRT-LLM.git
synced 2026-01-14 06:27:45 +08:00
812b1abf86
Signed-off-by: Arthur Rasmusson <47877520+arthurrasmusson@users.noreply.github.com.> Co-authored-by: Robin Kobus <19427718+Funatiq@users.noreply.github.com> Co-authored-by: Aurelien Chartier <2567591+achartier@users.noreply.github.com>
81 lines
3.5 KiB
C++
81 lines
3.5 KiB
C++
/*
|
|
* Copyright (c) 2022-2024, 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/batch_manager/kvCacheManager.h"
|
|
#include "tensorrt_llm/executor/types.h"
|
|
#include "tensorrt_llm/runtime/bufferManager.h"
|
|
#include "tensorrt_llm/runtime/cudaEvent.h"
|
|
|
|
namespace tr = tensorrt_llm::runtime;
|
|
|
|
#pragma once
|
|
|
|
namespace tensorrt_llm::batch_manager::kv_cache_manager
|
|
{
|
|
|
|
// The TransferManager accelerates transfers to/from the GPU by overlapping HtoD and DtoH transfers, and tracks ongoing
|
|
// transfers in order to avoid race conditions. It is functionally equivalent to the prior approach of putting all
|
|
// transfers into the forward pass stream. This is only ever used as a component of a KVCacheManager.
|
|
class KVCacheTransferManager
|
|
{
|
|
public:
|
|
explicit KVCacheTransferManager(tr::BufferManager const& bufferManager);
|
|
|
|
//! \brief Onboard a block to gpu memory.
|
|
void onboard(BlockPtr const& offloadBlock, BlockPtr const& block, std::vector<KVCacheBlockPool> const& pools,
|
|
int numTokensToCopy = 0, executor::KvCacheTransferMode mode = executor::KvCacheTransferMode::DRAM,
|
|
std::optional<std::string> directory = std::nullopt);
|
|
|
|
//! \brief Offload a block to cpu memory.
|
|
void offload(BlockPtr const& block, BlockPtr const& offloadBlock, std::vector<KVCacheBlockPool> const& pools,
|
|
int numTokensToCopy = 0, executor::KvCacheTransferMode mode = executor::KvCacheTransferMode::DRAM,
|
|
std::optional<std::string> directory = std::nullopt);
|
|
|
|
//! \brief Synchronize the offload/onboard streams with the bufferManager stream.
|
|
void syncTransfers();
|
|
|
|
private:
|
|
//! \brief Get pointer to pool specified by cache block.
|
|
static tr::ITensor::SharedPtr computeBlockPointer(
|
|
BlockPtr const& block, std::vector<KVCacheBlockPool> const& pools, size_t poolIdx);
|
|
|
|
/*!
|
|
* \brief The key method that copies the src block to the dst block.
|
|
*
|
|
* \param src Source block
|
|
* \param dst Destination block
|
|
* \param pools Pools describing memory layout for KV blocks
|
|
* \param isOffload true => GPU->CPU/file, false => CPU/file->GPU
|
|
* \param numTokensToCopy if > 0, partial copy is done
|
|
* \param mode See \ref executor::KvCacheTransferMode
|
|
* \param directory Directory to save the file if mode is GDS or POSIX_DEBUG_FALLBACK
|
|
*
|
|
* The default param is set to executor::KvCacheTransferMode::DRAM.
|
|
*/
|
|
void copyBlock(BlockPtr const& src, BlockPtr const& dst, std::vector<KVCacheBlockPool> const& pools, bool isOffload,
|
|
int numTokensToCopy = 0, executor::KvCacheTransferMode mode = executor::KvCacheTransferMode::DRAM,
|
|
std::optional<std::string> directory = std::nullopt);
|
|
|
|
runtime::BufferManager mBufferManager;
|
|
runtime::BufferManager mOnboardManager;
|
|
runtime::BufferManager mOffloadManager;
|
|
|
|
// Track the block ids offloaded in this iteration.
|
|
std::unordered_map<int32_t, tr::CudaEvent> mPendingOffloads;
|
|
};
|
|
|
|
} // namespace tensorrt_llm::batch_manager::kv_cache_manager
|