381 lines
16 KiB
C++
381 lines
16 KiB
C++
// Copyright (c) 2023 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 <string>
|
|
|
|
#include "paddle/phi/kernels/quantize_linear_kernel.h"
|
|
|
|
#include "paddle/common/hostdevice.h"
|
|
#include "paddle/phi/backends/gpu/gpu_launch_config.h"
|
|
#include "paddle/phi/common/place.h"
|
|
#include "paddle/phi/core/tensor_utils.h"
|
|
#include "paddle/phi/kernels/cast_kernel.h"
|
|
#include "paddle/phi/kernels/funcs/fake_quantize_functor.h"
|
|
|
|
namespace phi {
|
|
|
|
template <typename Context, typename T>
|
|
struct DequantizeFunctor {
|
|
void operator()(const Context& dev_ctx,
|
|
const DenseTensor* in,
|
|
const DenseTensor* scale,
|
|
T max_range,
|
|
DenseTensor* out);
|
|
};
|
|
|
|
template <typename Context, typename T>
|
|
struct ChannelDequantizeFunctorV2 {
|
|
void operator()(const Context& dev_ctx,
|
|
const DenseTensor* in,
|
|
const DenseTensor** scales,
|
|
const int scale_num,
|
|
T max_range,
|
|
const int quant_axis,
|
|
DenseTensor* out);
|
|
};
|
|
|
|
template <typename T, typename Context, typename D>
|
|
void DeQuantizeLinearImpl(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const DenseTensor& scale,
|
|
int quant_axis,
|
|
int qmax,
|
|
bool only_observer,
|
|
DenseTensor* out) {
|
|
auto* in = &x;
|
|
|
|
auto in_tmp = Cast<T>(dev_ctx, *in, CppTypeToDataType<D>::Type());
|
|
|
|
dev_ctx.template Alloc<D>(out, out->numel() * sizeof(D));
|
|
|
|
if (only_observer) {
|
|
Copy(dev_ctx, *in, dev_ctx.GetPlace(), false, out);
|
|
return;
|
|
}
|
|
|
|
if (quant_axis < 0) {
|
|
float max_range = qmax;
|
|
DequantizeFunctor<Context, D>()(
|
|
dev_ctx, &in_tmp, &scale, static_cast<D>(max_range), out);
|
|
} else {
|
|
PADDLE_ENFORCE_EQ(
|
|
scale.numel(),
|
|
in_tmp.dims()[quant_axis],
|
|
common::errors::PreconditionNotMet(
|
|
"The number of first scale values must be the same with "
|
|
"quant_axis dimension value of Input(X) when the `scale` has "
|
|
"only one element, but %ld != %ld here.",
|
|
scale.numel(),
|
|
in_tmp.dims()[quant_axis]));
|
|
int max_range = qmax;
|
|
|
|
ChannelDequantizeFunctorV2<Context, D>()(
|
|
dev_ctx, &in_tmp, &scale, static_cast<D>(max_range), quant_axis, out);
|
|
}
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void DeQuantizeLinearKernel(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const optional<DenseTensor>& in_scale,
|
|
const DenseTensor& zero_point,
|
|
const optional<DenseTensor>& in_accum,
|
|
const optional<DenseTensor>& in_state,
|
|
int quant_axis,
|
|
int bit_length,
|
|
int qmin,
|
|
int qmax,
|
|
int round_type,
|
|
bool is_test,
|
|
bool only_observer,
|
|
DenseTensor* out,
|
|
DenseTensor* out_state,
|
|
DenseTensor* out_accum,
|
|
DenseTensor* out_scale) {
|
|
PADDLE_ENFORCE_NE(in_scale.get_ptr(),
|
|
nullptr,
|
|
common::errors::PreconditionNotMet(
|
|
"in_scale can't be nullptr in DeQuantizeLinearKernel"));
|
|
auto scale = in_scale.get();
|
|
switch (scale.dtype()) {
|
|
case DataType::FLOAT64:
|
|
DeQuantizeLinearImpl<T, Context, double>(
|
|
dev_ctx, x, scale, quant_axis, qmax, only_observer, out);
|
|
break;
|
|
case DataType::FLOAT32:
|
|
DeQuantizeLinearImpl<T, Context, float>(
|
|
dev_ctx, x, scale, quant_axis, qmax, only_observer, out);
|
|
break;
|
|
case DataType::FLOAT16:
|
|
DeQuantizeLinearImpl<T, Context, float16>(
|
|
dev_ctx, x, scale, quant_axis, qmax, only_observer, out);
|
|
break;
|
|
default:
|
|
PADDLE_THROW(common::errors::Unimplemented(
|
|
"In DeQuantizeLinearKernel, "
|
|
"data type %d for scale/output is not supported ",
|
|
scale.dtype()));
|
|
break;
|
|
}
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void QuantizeLinearTrainKernel(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const optional<DenseTensor>& scale,
|
|
const DenseTensor& zero_point,
|
|
const optional<DenseTensor>& in_accum,
|
|
const optional<DenseTensor>& in_state,
|
|
int quant_axis,
|
|
int bit_length,
|
|
int qmin,
|
|
int qmax,
|
|
int round_type,
|
|
bool only_observer,
|
|
DenseTensor* out,
|
|
DenseTensor* out_state,
|
|
DenseTensor* out_accum,
|
|
DenseTensor* out_scale) {
|
|
PADDLE_ENFORCE_NE(scale.get_ptr(),
|
|
nullptr,
|
|
common::errors::PreconditionNotMet(
|
|
"in_scale can't be nullptr in DeQuantizeLinearKernel"));
|
|
auto* in = &x;
|
|
dev_ctx.template Alloc<float>(out);
|
|
|
|
if (quant_axis < 0) {
|
|
// training
|
|
DenseTensor tmp_scale;
|
|
tmp_scale.Resize(common::make_dim(1));
|
|
T* cur_scale_data = dev_ctx.template Alloc<T>(&tmp_scale);
|
|
|
|
funcs::FindAbsMaxFunctor<Context, T>()(
|
|
dev_ctx, in->data<T>(), in->numel(), cur_scale_data);
|
|
|
|
dev_ctx.template Alloc<T>(out_state);
|
|
dev_ctx.template Alloc<T>(out_accum);
|
|
dev_ctx.template Alloc<T>(out_scale);
|
|
|
|
funcs::FindMovingAverageAbsMaxFunctor<Context, T>()(dev_ctx,
|
|
in_accum.get(),
|
|
in_state.get(),
|
|
cur_scale_data,
|
|
0.9,
|
|
out_state,
|
|
out_accum,
|
|
out_scale);
|
|
if (only_observer) {
|
|
Copy<Context>(dev_ctx, *in, dev_ctx.GetPlace(), false, out);
|
|
} else {
|
|
funcs::ClipAndFakeQuantFunctor<Context, T>()(
|
|
dev_ctx, *in, *out_scale, qmax, round_type, out);
|
|
}
|
|
} else {
|
|
T* out_scale_data = dev_ctx.template Alloc<T>(out_scale);
|
|
funcs::FindChannelAbsMaxFunctor<Context, T>()(
|
|
dev_ctx, *in, quant_axis, out_scale_data);
|
|
if (only_observer) {
|
|
Copy<Context>(dev_ctx, *in, dev_ctx.GetPlace(), false, out);
|
|
} else {
|
|
funcs::ChannelClipAndFakeQuantFunctor<Context, T>()(
|
|
dev_ctx, *in, *out_scale, qmax, round_type, quant_axis, out);
|
|
}
|
|
}
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void QuantizeLinearInferKernel(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const optional<DenseTensor>& scale,
|
|
const DenseTensor& zero_point,
|
|
int quant_axis,
|
|
int bit_length,
|
|
int qmin,
|
|
int qmax,
|
|
int round_type,
|
|
bool only_observer,
|
|
DenseTensor* out) {
|
|
PADDLE_ENFORCE_NE(scale.get_ptr(),
|
|
nullptr,
|
|
common::errors::PreconditionNotMet(
|
|
"in_scale can't be nullptr in DeQuantizeLinearKernel"));
|
|
auto* in = &x;
|
|
auto* in_scale = scale.get_ptr();
|
|
dev_ctx.template Alloc<float>(out);
|
|
|
|
if (quant_axis < 0) {
|
|
if (only_observer) {
|
|
Copy<Context>(dev_ctx, *in, dev_ctx.GetPlace(), false, out);
|
|
} else {
|
|
funcs::ClipAndFakeQuantFunctor<Context, T>()(
|
|
dev_ctx, *in, *in_scale, qmax, round_type, out);
|
|
}
|
|
} else {
|
|
if (only_observer) {
|
|
Copy<Context>(dev_ctx, *in, dev_ctx.GetPlace(), false, out);
|
|
} else {
|
|
funcs::ChannelClipAndFakeQuantFunctor<Context, T>()(
|
|
dev_ctx, *in, *in_scale, qmax, round_type, quant_axis, out);
|
|
}
|
|
}
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void QuantizeLinearKernel(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const optional<DenseTensor>& scale,
|
|
const DenseTensor& zero_point,
|
|
const optional<DenseTensor>& in_accum,
|
|
const optional<DenseTensor>& in_state,
|
|
int quant_axis,
|
|
int bit_length,
|
|
int qmin,
|
|
int qmax,
|
|
int round_type,
|
|
bool is_test,
|
|
bool only_observer,
|
|
DenseTensor* out,
|
|
DenseTensor* out_state,
|
|
DenseTensor* out_accum,
|
|
DenseTensor* out_scale) {
|
|
if (!is_test) {
|
|
QuantizeLinearTrainKernel<T, Context>(dev_ctx,
|
|
x,
|
|
scale,
|
|
zero_point,
|
|
in_accum,
|
|
in_state,
|
|
quant_axis,
|
|
bit_length,
|
|
qmin,
|
|
qmax,
|
|
round_type,
|
|
only_observer,
|
|
out,
|
|
out_state,
|
|
out_accum,
|
|
out_scale);
|
|
} else {
|
|
QuantizeLinearInferKernel<T, Context>(dev_ctx,
|
|
x,
|
|
scale,
|
|
zero_point,
|
|
quant_axis,
|
|
bit_length,
|
|
qmin,
|
|
qmax,
|
|
round_type,
|
|
only_observer,
|
|
out);
|
|
}
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void QuantizeLinearDeprecatedTrainKernel(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const DenseTensor& in_scale,
|
|
const DenseTensor& zero_point,
|
|
const optional<DenseTensor>& in_accum,
|
|
const optional<DenseTensor>& in_state,
|
|
int quant_axis,
|
|
int bit_length,
|
|
int qmin,
|
|
int qmax,
|
|
int round_type,
|
|
bool only_observer,
|
|
DenseTensor* out,
|
|
DenseTensor* out_state,
|
|
DenseTensor* out_accum,
|
|
DenseTensor* out_scale) {
|
|
optional<DenseTensor> scale = paddle::make_optional<DenseTensor>(in_scale);
|
|
QuantizeLinearTrainKernel<T, Context>(dev_ctx,
|
|
x,
|
|
scale,
|
|
zero_point,
|
|
in_accum,
|
|
in_state,
|
|
quant_axis,
|
|
bit_length,
|
|
qmin,
|
|
qmax,
|
|
round_type,
|
|
only_observer,
|
|
out,
|
|
out_state,
|
|
out_accum,
|
|
out_scale);
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void QuantizeLinearDeprecatedInferKernel(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const DenseTensor& in_scale,
|
|
const DenseTensor& zero_point,
|
|
int quant_axis,
|
|
int bit_length,
|
|
int qmin,
|
|
int qmax,
|
|
int round_type,
|
|
bool only_observer,
|
|
DenseTensor* out) {
|
|
optional<DenseTensor> scale = paddle::make_optional<DenseTensor>(in_scale);
|
|
QuantizeLinearInferKernel<T, Context>(dev_ctx,
|
|
x,
|
|
scale,
|
|
zero_point,
|
|
quant_axis,
|
|
bit_length,
|
|
qmin,
|
|
qmax,
|
|
round_type,
|
|
only_observer,
|
|
out);
|
|
}
|
|
|
|
template <typename T, typename Context>
|
|
void DeQuantizeLinearDeprecatedKernel(const Context& dev_ctx,
|
|
const DenseTensor& x,
|
|
const DenseTensor& in_scale,
|
|
const DenseTensor& zero_point,
|
|
int quant_axis,
|
|
int bit_length,
|
|
int qmin,
|
|
int qmax,
|
|
int round_type,
|
|
bool only_observer,
|
|
DenseTensor* out) {
|
|
optional<DenseTensor> scale = paddle::make_optional<DenseTensor>(in_scale);
|
|
DeQuantizeLinearKernel<T, Context>(dev_ctx,
|
|
x,
|
|
scale,
|
|
zero_point,
|
|
nullptr,
|
|
nullptr,
|
|
quant_axis,
|
|
bit_length,
|
|
qmin,
|
|
qmax,
|
|
round_type,
|
|
true,
|
|
only_observer,
|
|
out,
|
|
nullptr,
|
|
nullptr,
|
|
nullptr);
|
|
}
|
|
|
|
} // namespace phi
|