wehub-resource-sync
108 lines
3.5 KiB
C++
108 lines
3.5 KiB
C++
// Copyright (c) 2024 PaddlePaddle Authors. 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.
|
|
|
|
#pragma once
|
|
#include <vector>
|
|
#include "paddle/phi/backends/xpu/enforce_xpu.h"
|
|
#include "paddle/phi/common/memory_utils.h"
|
|
#include "xfa/flash_api.h"
|
|
|
|
namespace xfa = baidu::xpu::xfa;
|
|
namespace phi {
|
|
|
|
using XPUTypeFP16 = typename XPUTypeTrait<phi::float16>::Type;
|
|
using XPUTypeBF16 = typename XPUTypeTrait<phi::bfloat16>::Type;
|
|
|
|
enum XPU_FA_DTYPE {
|
|
FA_FLOAT = 0,
|
|
FA_TFLOAT32,
|
|
FA_FLOAT16,
|
|
FA_BFLOAT16,
|
|
FA_INT32,
|
|
};
|
|
|
|
inline bool get_env_flag(const char* env_name) {
|
|
const char* env = std::getenv(env_name);
|
|
return env != nullptr && std::strcmp(env, "1") == 0;
|
|
}
|
|
|
|
template <typename T> // for bfloat16 and float16
|
|
XPU_FA_DTYPE get_flash_attn_tgemm() {
|
|
if (get_env_flag("XPU_PADDLE_FA_TGEMM_FLOAT16")) {
|
|
return XPU_FA_DTYPE::FA_FLOAT16;
|
|
} else if (get_env_flag("XPU_PADDLE_FA_TGEMM_FLOAT")) {
|
|
return XPU_FA_DTYPE::FA_FLOAT;
|
|
} else {
|
|
return XPU_FA_DTYPE::FA_TFLOAT32;
|
|
}
|
|
}
|
|
|
|
template <> // for float
|
|
inline XPU_FA_DTYPE get_flash_attn_tgemm<float>() {
|
|
if (get_env_flag("XPU_PADDLE_FA_TGEMM_FLOAT")) {
|
|
return XPU_FA_DTYPE::FA_FLOAT;
|
|
} else {
|
|
return XPU_FA_DTYPE::FA_TFLOAT32;
|
|
}
|
|
}
|
|
|
|
template <typename T> // for bfloat16 and float16
|
|
XPU_FA_DTYPE get_flash_attn_taccum() {
|
|
if (get_env_flag("XPU_PADDLE_FA_TACCUM_FLOAT16")) {
|
|
return XPU_FA_DTYPE::FA_FLOAT16;
|
|
} else {
|
|
return XPU_FA_DTYPE::FA_FLOAT;
|
|
}
|
|
}
|
|
|
|
template <> // for float
|
|
inline XPU_FA_DTYPE get_flash_attn_taccum<float>() {
|
|
return XPU_FA_DTYPE::FA_FLOAT;
|
|
}
|
|
|
|
static void GenerateRNGState(const XPUContext& dev_ctx,
|
|
const optional<DenseTensor>& fixed_seed_offset,
|
|
int64_t* seed_offset_data,
|
|
const std::string& rng_name,
|
|
const int64_t batch_size,
|
|
const int64_t num_heads) {
|
|
if (fixed_seed_offset.get_ptr()) {
|
|
if ((fixed_seed_offset->place()).GetType() == AllocationType::XPU) {
|
|
memory_utils::Copy(CPUPlace(),
|
|
seed_offset_data,
|
|
fixed_seed_offset->place(),
|
|
fixed_seed_offset->data<int64_t>(),
|
|
sizeof(int64_t) * 2);
|
|
} else {
|
|
const int64_t* fixed_seed_offset_data =
|
|
fixed_seed_offset->data<int64_t>();
|
|
seed_offset_data[0] = fixed_seed_offset_data[0];
|
|
seed_offset_data[1] = fixed_seed_offset_data[1];
|
|
}
|
|
} else {
|
|
std::pair<uint64_t, uint64_t> seed_offset_pair;
|
|
uint64_t inc = batch_size * num_heads * 32;
|
|
if (rng_name != "") {
|
|
auto gen = phi::GetRandomSeedGenerator(rng_name);
|
|
seed_offset_pair = gen->IncrementOffset(inc);
|
|
} else {
|
|
auto* gen = dev_ctx.GetGenerator();
|
|
seed_offset_pair = gen->IncrementOffset(inc);
|
|
}
|
|
seed_offset_data[0] = static_cast<int64_t>(seed_offset_pair.first);
|
|
seed_offset_data[1] = static_cast<int64_t>(seed_offset_pair.second);
|
|
}
|
|
}
|
|
} // namespace phi
|