82 lines
3.5 KiB
C++
82 lines
3.5 KiB
C++
/**
|
||
* Copyright (c) 2025 Huawei Technologies Co., Ltd.
|
||
* This program is free software, you can redistribute it and/or modify it under the terms and conditions of
|
||
* CANN Open Software License Agreement Version 2.0 (the "License").
|
||
* Please refer to the License for details. You may not use this file except in compliance with the License.
|
||
* THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED,
|
||
* INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
|
||
* See LICENSE in the root of the software repository for the full text of the License.
|
||
*/
|
||
|
||
/*!
|
||
* \file lightning_indexer_def.cpp
|
||
* \brief
|
||
*/
|
||
#include <cstdint>
|
||
#include "register/op_def_registry.h"
|
||
|
||
namespace ops {
|
||
class LightningIndexer : public OpDef {
|
||
public:
|
||
explicit LightningIndexer(const char *name) : OpDef(name)
|
||
{
|
||
this->Input("query")
|
||
.ParamType(REQUIRED)
|
||
.DataType({ge::DT_BF16, ge::DT_FLOAT16, ge::DT_BF16, ge::DT_FLOAT16})
|
||
.FormatList({ge::FORMAT_ND})
|
||
.AutoContiguous();
|
||
this->Input("key")
|
||
.ParamType(REQUIRED)
|
||
.DataType({ge::DT_BF16, ge::DT_FLOAT16, ge::DT_BF16, ge::DT_FLOAT16})
|
||
.FormatList({ge::FORMAT_ND})
|
||
.AutoContiguous();
|
||
this->Input("weights")
|
||
.ParamType(REQUIRED)
|
||
.DataType({ge::DT_BF16, ge::DT_FLOAT16, ge::DT_FLOAT, ge::DT_FLOAT})
|
||
.FormatList({ge::FORMAT_ND})
|
||
.AutoContiguous();
|
||
this->Input("actual_seq_lengths_query")
|
||
.ParamType(OPTIONAL)
|
||
.DataTypeList({ge::DT_INT32})
|
||
.FormatList({ge::FORMAT_ND})
|
||
.AutoContiguous();
|
||
this->Input("actual_seq_lengths_key")
|
||
.ParamType(OPTIONAL)
|
||
.DataTypeList({ge::DT_INT32})
|
||
.FormatList({ge::FORMAT_ND})
|
||
.AutoContiguous();
|
||
this->Input("block_table")
|
||
.ParamType(OPTIONAL)
|
||
.DataTypeList({ge::DT_INT32})
|
||
.FormatList({ge::FORMAT_ND})
|
||
.AutoContiguous();
|
||
this->Output("sparse_indices")
|
||
.ParamType(REQUIRED)
|
||
.DataTypeList({ge::DT_INT32})
|
||
.FormatList({ge::FORMAT_ND});
|
||
this->Output("sparse_values")
|
||
.ParamType(REQUIRED)
|
||
.DataType({ge::DT_BF16, ge::DT_FLOAT16, ge::DT_BF16, ge::DT_FLOAT16})
|
||
.FormatList({ge::FORMAT_ND});
|
||
this->Attr("layout_query").AttrType(OPTIONAL).String("BSND");
|
||
this->Attr("layout_key").AttrType(OPTIONAL).String("BSND");
|
||
this->Attr("sparse_count").AttrType(OPTIONAL).Int(2048); // 2048:默认值，筛选前2048
|
||
this->Attr("sparse_mode").AttrType(OPTIONAL).Int(3); // 3:默认值，只计算下三角
|
||
this->Attr("pre_tokens").AttrType(OPTIONAL).Int(INT64_MAX);
|
||
this->Attr("next_tokens").AttrType(OPTIONAL).Int(INT64_MAX);
|
||
this->Attr("return_values").AttrType(OPTIONAL).Bool(false);
|
||
OpAICoreConfig aicore_config;
|
||
aicore_config.DynamicCompileStaticFlag(true)
|
||
.DynamicFormatFlag(true)
|
||
.DynamicRankSupportFlag(true)
|
||
.DynamicShapeSupportFlag(true)
|
||
.NeedCheckSupportFlag(false)
|
||
.PrecisionReduceFlag(true);
|
||
this->AICore().AddConfig("ascend910b", aicore_config);
|
||
this->AICore().AddConfig("ascend910_93", aicore_config);
|
||
this->AICore().AddConfig("ascend950", aicore_config);
|
||
}
|
||
};
|
||
OP_ADD(LightningIndexer);
|
||
} // namespace ops
|