Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

add new format of quantization #41041

Merged
merged 17 commits into from
Apr 5, 2022
Merged
3 changes: 2 additions & 1 deletion paddle/fluid/operators/CMakeLists.txt
Original file line number Diff line number Diff line change
Expand Up @@ -102,10 +102,11 @@ endif()

set(OP_HEADER_DEPS ${OP_HEADER_DEPS} phi phi_api_utils gather_scatter_kernel)

register_operators(EXCLUDES py_layer_op py_func_op warpctc_op dgc_op load_combine_op lstm_op run_program_op eye_op
register_operators(EXCLUDES py_layer_op py_func_op warpctc_op dgc_op load_combine_op lstm_op run_program_op eye_op quantize_linear_op
recurrent_op save_combine_op sparse_attention_op sync_batch_norm_op spectral_op ${OP_MKL_DEPS} DEPS ${OP_HEADER_DEPS})

op_library(run_program_op SRCS run_program_op.cc run_program_op.cu.cc DEPS executor_cache ${OP_HEADER_DEPS})
op_library(quantize_linear_op DEPS cast_kernel)
op_library(save_combine_op DEPS string_array)
op_library(load_combine_op DEPS string_array)

Expand Down
135 changes: 1 addition & 134 deletions paddle/fluid/operators/fake_dequantize_op.cu
Original file line number Diff line number Diff line change
Expand Up @@ -12,142 +12,9 @@ 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/fluid/operators/fake_dequantize_op.cu.h"
#include "paddle/fluid/operators/fake_dequantize_op.h"

namespace paddle {
namespace operators {

template <typename T>
__global__ void KeDequantize(const T* in, const T* scale, T max_range, int num,
T* out) {
const int idx = threadIdx.x + blockIdx.x * blockDim.x;
if (idx < num) {
out[idx] = in[idx] * scale[0] / max_range;
}
}

template <typename T>
struct DequantizeFunctor<platform::CUDADeviceContext, T> {
void operator()(const platform::CUDADeviceContext& dev_ctx,
const framework::Tensor* in, const framework::Tensor* scale,
T max_range, framework::Tensor* out) {
const T* in_data = in->data<T>();
const T* scale_factor = scale->data<T>();
T* out_data = out->mutable_data<T>(dev_ctx.GetPlace());

int num = in->numel();
int block = 512;
int grid = (num + block - 1) / block;

KeDequantize<T><<<grid, block, 0, dev_ctx.stream()>>>(
in_data, scale_factor, max_range, num, out_data);
}
};

template <typename T>
__global__ void DequantizeOneScaleQuantAxis0(const T* in, const T* scale,
T max_range, int num, int channel,
T* out) {
int tid = threadIdx.x;
int channel_size = num / channel;
const T* in_c = in + blockIdx.x * channel_size;
T* out_c = out + blockIdx.x * channel_size;
for (int i = tid; i < channel_size; i += blockDim.x) {
out_c[i] = in_c[i] * scale[blockIdx.x] / max_range;
}
}

template <typename T>
__global__ void DequantizeOneScaleQuantAxisN(const T* in, const T* scale,
const T max_range,
const int64_t num,
const int n_scales,
const int quant_stride, T* out) {
int64_t idx = blockDim.x * blockIdx.x + threadIdx.x;
for (int64_t i = idx; i < num; i += blockDim.x * gridDim.x) {
T s = scale[(i / quant_stride) % n_scales];
out[i] = in[i] * s / max_range;
}
}

template <typename T>
__global__ void DequantizeTwoScale(const T* in, const T* scale_one,
const T* scale_two, T max_range, int num,
int iter_size, int channel, T* out) {
int tid = threadIdx.x;
int channel_size = num / (iter_size * channel);
int scale_index = blockIdx.x % channel;
const T* in_c = in + blockIdx.x * channel_size;
T* out_c = out + blockIdx.x * channel_size;
for (int i = tid; i < channel_size; i += blockDim.x) {
out_c[i] = in_c[i] * scale_one[scale_index] * scale_two[0] / max_range;
}
}

template <typename T>
struct ChannelDequantizeFunctor<platform::CUDADeviceContext, T> {
void operator()(const platform::CUDADeviceContext& dev_ctx,
const framework::Tensor* in, const framework::Tensor** scales,
const int scale_num, T max_range, const int quant_axis,
const int x_num_col_dims, framework::Tensor* out) {
auto in_dims = in->dims();
const T* in_data = in->data<T>();
T* out_data = out->mutable_data<T>(dev_ctx.GetPlace());
if (scale_num == 1) {
int64_t num = in->numel();
const T* scale_factor = scales[0]->data<T>();
if (quant_axis == 0) {
int grid = in_dims[0];
int block = 1024;
DequantizeOneScaleQuantAxis0<T><<<grid, block, 0, dev_ctx.stream()>>>(
in_data, scale_factor, max_range, num, in_dims[0], out_data);
} else {
int quant_stride = 1;
for (int i = quant_axis + 1; i < in_dims.size(); i++) {
quant_stride *= in_dims[i];
}

int64_t block_size = std::min(
num, static_cast<int64_t>(dev_ctx.GetMaxThreadsPerBlock() / 4));
int64_t max_threads =
dev_ctx.GetMaxPhysicalThreadCount(); // SM * block_per_SM
const int64_t max_blocks = std::max(
((max_threads - 1) / block_size + 1), static_cast<int64_t>(1));
const int64_t grid_size =
std::min(max_blocks, (num + block_size - 1) / block_size);

DequantizeOneScaleQuantAxisN<
T><<<grid_size, block_size, 0, dev_ctx.stream()>>>(
in_data, scale_factor, max_range, num, in_dims[quant_axis],
quant_stride, out_data);
}
} else if (scale_num == 2) {
// Not need to consider quant_axis
int num = in->numel();
int iter_size = 1;
for (int i = 0; i < x_num_col_dims; i++) {
iter_size *= in->dims()[i];
}
int channel = in->dims()[x_num_col_dims];
const T* scale_one = scales[0]->data<T>();
const T* scale_two = scales[1]->data<T>();
int block = 1024;
int grid = iter_size * channel;
DequantizeTwoScale<T><<<grid, block, 0, dev_ctx.stream()>>>(
in_data, scale_one, scale_two, max_range, num, iter_size, channel,
out_data);
}
}
};

template struct DequantizeFunctor<platform::CUDADeviceContext, float>;
template struct DequantizeFunctor<platform::CUDADeviceContext, double>;
template struct ChannelDequantizeFunctor<platform::CUDADeviceContext, float>;
template struct ChannelDequantizeFunctor<platform::CUDADeviceContext, double>;

} // namespace operators
} // namespace paddle

namespace ops = paddle::operators;
using CUDA = paddle::platform::CUDADeviceContext;
REGISTER_OP_CUDA_KERNEL(fake_dequantize_max_abs,
Expand Down
151 changes: 151 additions & 0 deletions paddle/fluid/operators/fake_dequantize_op.cu.h
Original file line number Diff line number Diff line change
@@ -0,0 +1,151 @@
/* Copyright (c) 2022 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. */

#ifndef PADDLE_FLUID_OPERATORS_FAKE_DEQUANTIZE_OP_CU_H_
#define PADDLE_FLUID_OPERATORS_FAKE_DEQUANTIZE_OP_CU_H_
#endif // PADDLE_FLUID_OPERATORS_FAKE_DEQUANTIZE_OP_CU_H_

#include "paddle/fluid/operators/fake_dequantize_op.h"

namespace paddle {
namespace operators {

template <typename T>
__global__ void KeDequantize(const T* in, const T* scale, T max_range,
int64_t num, T* out) {
int64_t idx = threadIdx.x + blockIdx.x * blockDim.x;
for (int64_t i = idx; i < num; i += blockDim.x * gridDim.x) {
out[i] = in[i] * scale[0] / max_range;
}
}

template <typename T>
struct DequantizeFunctor<platform::CUDADeviceContext, T> {
void operator()(const platform::CUDADeviceContext& dev_ctx,
const framework::Tensor* in, const framework::Tensor* scale,
T max_range, framework::Tensor* out) {
const T* in_data = in->data<T>();
const T* scale_factor = scale->data<T>();
T* out_data = out->mutable_data<T>(dev_ctx.GetPlace());

int64_t num = in->numel();
int64_t block_size = std::min(
num, static_cast<int64_t>(dev_ctx.GetMaxThreadsPerBlock() / 4));
int64_t max_threads =
dev_ctx.GetMaxPhysicalThreadCount(); // SM * block_per_SM
const int64_t max_blocks =
std::max(((max_threads - 1) / block_size + 1), static_cast<int64_t>(1));
const int64_t grid_size =
std::min(max_blocks, (num + block_size - 1) / block_size);
KeDequantize<T><<<grid_size, block_size, 0, dev_ctx.stream()>>>(
in_data, scale_factor, max_range, num, out_data);
}
};

template <typename T>
__global__ void DequantizeOneScaleQuantAxis0(const T* in, const T* scale,
T max_range, int num, int channel,
T* out) {
int tid = threadIdx.x;
int channel_size = num / channel;
const T* in_c = in + blockIdx.x * channel_size;
T* out_c = out + blockIdx.x * channel_size;
for (int i = tid; i < channel_size; i += blockDim.x) {
out_c[i] = in_c[i] * scale[blockIdx.x] / max_range;
}
}

template <typename T>
__global__ void DequantizeOneScaleQuantAxisN(const T* in, const T* scale,
const T max_range,
const int64_t num,
const int n_scales,
const int quant_stride, T* out) {
int64_t idx = blockDim.x * blockIdx.x + threadIdx.x;
for (int64_t i = idx; i < num; i += blockDim.x * gridDim.x) {
T s = scale[(i / quant_stride) % n_scales];
out[i] = in[i] * s / max_range;
}
}

template <typename T>
__global__ void DequantizeTwoScale(const T* in, const T* scale_one,
const T* scale_two, T max_range, int num,
int iter_size, int channel, T* out) {
int tid = threadIdx.x;
int channel_size = num / (iter_size * channel);
int scale_index = blockIdx.x % channel;
const T* in_c = in + blockIdx.x * channel_size;
T* out_c = out + blockIdx.x * channel_size;
for (int i = tid; i < channel_size; i += blockDim.x) {
out_c[i] = in_c[i] * scale_one[scale_index] * scale_two[0] / max_range;
}
}

template <typename T>
struct ChannelDequantizeFunctor<platform::CUDADeviceContext, T> {
void operator()(const platform::CUDADeviceContext& dev_ctx,
const framework::Tensor* in, const framework::Tensor** scales,
const int scale_num, T max_range, const int quant_axis,
const int x_num_col_dims, framework::Tensor* out) {
auto in_dims = in->dims();
const T* in_data = in->data<T>();
T* out_data = out->mutable_data<T>(dev_ctx.GetPlace());
if (scale_num == 1) {
int64_t num = in->numel();
const T* scale_factor = scales[0]->data<T>();
int64_t block_size = std::min(
num, static_cast<int64_t>(dev_ctx.GetMaxThreadsPerBlock() / 4));
int64_t max_threads =
dev_ctx.GetMaxPhysicalThreadCount(); // SM * block_per_SM
const int64_t max_blocks = std::max(((max_threads - 1) / block_size + 1),
static_cast<int64_t>(1));
const int64_t grid_size =
std::min(max_blocks, (num + block_size - 1) / block_size);

int quant_stride = 1;
for (int i = quant_axis + 1; i < in_dims.size(); i++) {
quant_stride *= in_dims[i];
}

DequantizeOneScaleQuantAxisN<
T><<<grid_size, block_size, 0, dev_ctx.stream()>>>(
in_data, scale_factor, max_range, num, in_dims[quant_axis],
quant_stride, out_data);
} else if (scale_num == 2) {
// Not need to consider quant_axis
int num = in->numel();
int iter_size = 1;
for (int i = 0; i < x_num_col_dims; i++) {
iter_size *= in->dims()[i];
}
int channel = in->dims()[x_num_col_dims];
const T* scale_one = scales[0]->data<T>();
const T* scale_two = scales[1]->data<T>();
int block = 1024;
int grid = iter_size * channel;
DequantizeTwoScale<T><<<grid, block, 0, dev_ctx.stream()>>>(
in_data, scale_one, scale_two, max_range, num, iter_size, channel,
out_data);
}
}
};

template struct DequantizeFunctor<platform::CUDADeviceContext, float>;
template struct DequantizeFunctor<platform::CUDADeviceContext, double>;
template struct ChannelDequantizeFunctor<platform::CUDADeviceContext, float>;
template struct ChannelDequantizeFunctor<platform::CUDADeviceContext, double>;

} // namespace operators
} // namespace paddle
Loading