mirror of
https://github.com/NVIDIA/TensorRT-LLM.git
synced 2026-01-14 06:27:45 +08:00
d879430b04
* Update TensorRT-LLM --------- Co-authored-by: Shixiaowei02 <39303645+Shixiaowei02@users.noreply.github.com>
37 lines
1.5 KiB
C++
37 lines
1.5 KiB
C++
/*
|
|
* SPDX-FileCopyrightText: Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
|
|
* SPDX-License-Identifier: LicenseRef-NvidiaProprietary
|
|
*
|
|
* NVIDIA CORPORATION, its affiliates and licensors retain all intellectual
|
|
* property and proprietary rights in and to this material, related
|
|
* documentation and any modifications thereto. Any use, reproduction,
|
|
* disclosure or distribution of this material and related documentation
|
|
* without an express license agreement from NVIDIA CORPORATION or
|
|
* its affiliates is strictly prohibited.
|
|
*/
|
|
|
|
#pragma once
|
|
|
|
#include "tensorrt_llm/runtime/common.h"
|
|
#include "tensorrt_llm/runtime/gptModelConfig.h"
|
|
#include "tensorrt_llm/runtime/iTensor.h"
|
|
#include "tensorrt_llm/runtime/worldConfig.h"
|
|
|
|
namespace tensorrt_llm::runtime::lora
|
|
{
|
|
|
|
SizeType constexpr kLORA_CONFIG_ROW_SIZE = 3;
|
|
SizeType constexpr kLORA_CONFIG_MODULE_OFF = 0;
|
|
SizeType constexpr kLORA_CONFIG_LAYER_OFF = 1;
|
|
SizeType constexpr kLORA_CONFIG_ADAPTER_SIZE_OFF = 2;
|
|
|
|
SizeType constexpr kLORA_NUM_WEIGHTS_POINTERS = 2;
|
|
|
|
void loraValidateRequestTensorDims(const std::optional<ITensor::SharedPtr>& optReqLoraWeights,
|
|
const std::optional<ITensor::SharedPtr>& optReqLoraConfig);
|
|
|
|
void loraValidateRequestTensors(const std::optional<ITensor::SharedPtr>& optReqLoraWeights,
|
|
const std::optional<ITensor::SharedPtr>& optReqLoraConfig, runtime::GptModelConfig const& modelConfig,
|
|
runtime::WorldConfig const& worldConfig);
|
|
} // namespace tensorrt_llm::runtime::lora
|