wehub-resource-sync
74 lines
2.7 KiB
C++
74 lines
2.7 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.
|
|
|
|
#include "paddle/phi/backends/xpu/enforce_xpu.h"
|
|
#include "paddle/phi/core/kernel_registry.h"
|
|
|
|
namespace phi {
|
|
namespace fusion {
|
|
|
|
template <typename T, typename Context>
|
|
void MaskAdaptiveXPUKernel(const Context& dev_ctx,
|
|
const DenseTensor& mask,
|
|
DenseTensor* length,
|
|
DenseTensor* seq_lod,
|
|
DenseTensor* pad_seq_len) {
|
|
const auto& mask_dims = mask.dims();
|
|
auto batch_size = mask_dims[0];
|
|
auto pad_seq_len_size = mask_dims[1];
|
|
std::vector<int> cpu_seq_lod{0};
|
|
std::vector<int64_t> cpu_seq_lens;
|
|
auto mask_ptr = mask.data<T>();
|
|
for (auto batch_idx = 0; batch_idx < batch_size; batch_idx++) {
|
|
int cur_batch_seq_len = 0;
|
|
for (auto seq_idx = 0; seq_idx < pad_seq_len_size; seq_idx++) {
|
|
if (mask_ptr[batch_idx * pad_seq_len_size + seq_idx] >
|
|
static_cast<T>(1e-7)) {
|
|
cur_batch_seq_len += 1;
|
|
} else {
|
|
break;
|
|
}
|
|
}
|
|
PADDLE_ENFORCE_GT(
|
|
cur_batch_seq_len,
|
|
0,
|
|
common::errors::InvalidArgument("cur_batch_seq_len [%d] is less than 0",
|
|
cur_batch_seq_len));
|
|
cpu_seq_lod.push_back(cpu_seq_lod.back() + cur_batch_seq_len);
|
|
cpu_seq_lens.push_back(cur_batch_seq_len);
|
|
}
|
|
auto* seq_lod_ptr = dev_ctx.template HostAlloc<int>(seq_lod);
|
|
memcpy(seq_lod_ptr, cpu_seq_lod.data(), cpu_seq_lod.size() * sizeof(int));
|
|
auto* seq_lens_ptr = dev_ctx.template HostAlloc<int64_t>(length);
|
|
memcpy(
|
|
seq_lens_ptr, cpu_seq_lens.data(), cpu_seq_lens.size() * sizeof(int64_t));
|
|
auto* pad_seq_len_ptr = dev_ctx.template HostAlloc<int>(pad_seq_len);
|
|
pad_seq_len_ptr[0] = pad_seq_len_size;
|
|
}
|
|
|
|
} // namespace fusion
|
|
} // namespace phi
|
|
|
|
PD_REGISTER_KERNEL(mask_adaptive_xpu,
|
|
XPU,
|
|
ALL_LAYOUT,
|
|
phi::fusion::MaskAdaptiveXPUKernel,
|
|
float,
|
|
phi::float16) {
|
|
kernel->InputAt(0).SetBackend(phi::Backend::CPU);
|
|
kernel->OutputAt(0).SetBackend(phi::Backend::CPU);
|
|
kernel->OutputAt(1).SetBackend(phi::Backend::CPU);
|
|
kernel->OutputAt(2).SetBackend(phi::Backend::CPU);
|
|
}
|