sglang_v0.5.2/pytorch_2.8.0/third_party/fbgemm/test/GConvTest.cc

646 lines
23 KiB
C++

/*
* Copyright (c) Meta Platforms, Inc. and affiliates.
* All rights reserved.
*
* This source code is licensed under the BSD-style license found in the
* LICENSE file in the root directory of this source tree.
*/
#include <algorithm>
#include <chrono>
#include <cmath>
#include <random>
#include <vector>
#ifdef _OPENMP
#include <omp.h>
#endif
#include <gtest/gtest.h>
#include "./QuantizationHelpers.h"
#include "./TestUtils.h"
#include "bench/BenchUtils.h"
#include "fbgemm/Fbgemm.h"
#include "src/RefImplementations.h"
using namespace std;
using namespace fbgemm;
vector<matrix_op_t> transposeVals{
matrix_op_t::NoTranspose,
matrix_op_t::Transpose};
vector<QuantizationGranularity> qGranularityVals{
QuantizationGranularity::TENSOR,
QuantizationGranularity::GROUP,
QuantizationGranularity::OUT_CHANNEL};
namespace {
// class fbgemmGConvAcc32Test
// : public testing::TestWithParam<tuple<matrix_op_t, matrix_op_t>> {};
class fbgemmGConvAcc32WithQuantGranularityTest
: public testing::TestWithParam<tuple<
matrix_op_t,
matrix_op_t,
QuantizationGranularity,
bool,
bool>> {};
class fbgemmGConvPackTest : public testing::TestWithParam<matrix_op_t> {};
}; // namespace
// INSTANTIATE_TEST_CASE_P(
// InstantiationName,
// fbgemmGConvAcc32Test,
// ::testing::Combine(
// ::testing::Values(matrix_op_t::NoTranspose),
// ::testing::ValuesIn(transposeVals)));
INSTANTIATE_TEST_CASE_P(
InstantiationName,
fbgemmGConvAcc32WithQuantGranularityTest,
::testing::Combine(
::testing::Values(matrix_op_t::NoTranspose),
::testing::ValuesIn(transposeVals),
::testing::ValuesIn(qGranularityVals),
::testing::Bool(), // A symmetric
::testing::Bool())); // B symmetric
INSTANTIATE_TEST_CASE_P(
InstantiationName,
fbgemmGConvPackTest,
::testing::ValuesIn(transposeVals));
/**
* @brief 3D Shapes for unit test.
*/
template <int SPATIAL_DIM>
static typename std::enable_if<SPATIAL_DIM == 3, vector<conv_param_t<3>>>::type
GetShapes_() {
// clang-format off
vector<conv_param_t<3>> shapes = {
// MB, IC, OC, {IT, IH, IW}, G, {KT, KH, KW},
// {stride_t, stride_h, stride_w}, {pad_p, pad_t, pad_l,
// pad_n, pad_b, pad_r}
conv_param_t<3>(1, 16, 16, {5, 5, 5}, 8, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 5, 5}, 4, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 5, 5}, 2, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 5, 5}, 1, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 5, 5}, 8, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 5, 5}, 4, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 5, 5}, 2, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 5, 5}, 1, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {4, 4, 4}, 8, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {4, 4, 4}, 4, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {4, 4, 4}, 2, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {4, 4, 4}, 1, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {6, 6, 6}, 8, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {6, 6, 6}, 4, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {6, 6, 6}, 2, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {6, 6, 6}, 1, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
// batch size > 1
conv_param_t<3>(2, 16, 16, {4, 4, 4}, 8, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(2, 16, 16, {4, 4, 4}, 4, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(2, 16, 16, {4, 4, 4}, 2, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(2, 16, 16, {4, 4, 4}, 1, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(2, 16, 16, {6, 6, 6}, 8, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(2, 16, 16, {6, 6, 6}, 4, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(2, 16, 16, {6, 6, 6}, 2, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(2, 16, 16, {6, 6, 6}, 1, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 4, 4}, 8, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 4, 4}, 4, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 4, 4}, 2, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 4, 4}, 1, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 6, 6}, 8, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 6, 6}, 4, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 6, 6}, 2, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 6, 6}, 1, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
// unequal stride
conv_param_t<3>(1, 16, 16, {1, 6, 6}, 8, {3, 3, 3},
{1, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 6, 6}, 4, {3, 3, 3},
{1, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 6, 6}, 2, {3, 3, 3},
{1, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {1, 6, 6}, 1, {3, 3, 3},
{1, 2, 2}, {1, 1, 1, 1, 1, 1}),
// Small H and W corner cases
conv_param_t<3>(1, 16, 16, {5, 5, 2}, 1, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 2, 5}, 1, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 5, 1}, 1, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 1, 5}, 1, {3, 3, 3},
{1, 1, 1}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 5, 6}, 8, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 6, 5}, 8, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 5, 3}, 8, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 3, 5}, 8, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
conv_param_t<3>(1, 16, 16, {5, 3, 3}, 8, {3, 3, 3},
{2, 2, 2}, {1, 1, 1, 1, 1, 1}),
};
return shapes;
// clang-format off
}
/**
* @brief 2D Shapes for unit test.
*/
template <int SPATIAL_DIM = 2>
static typename std::enable_if<SPATIAL_DIM == 2, vector<conv_param_t<2>>>::type
GetShapes_() {
vector<conv_param_t<>> shapes = {
// MB, IC, OC, {IH, IW}, G, {KH, KW}, {stride_h, stride_w},
// {pad_t, pad_l, pad_b, pad_r}
conv_param_t<>(1, 16, 16, {5, 5}, 8, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 5}, 4, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 5}, 2, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 5}, 1, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 5}, 8, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 5}, 4, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 5}, 2, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 5}, 1, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 32, 32, {3, 3}, 8, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 32, 32, {4, 4}, 8, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 32, 32, {3, 5}, 8, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 32, 32, {5, 3}, 8, {3, 3}, {1, 1}, {1, 1, 1, 1}),
// fix from 8 to 16 to address G_together > G for avx512
conv_param_t<>(1, 16, 16, {5, 5}, 2, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 128, 128, {56, 48}, 32, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 128, 128, {48, 56}, 32, {3, 3}, {1, 1}, {1, 1, 1, 1}),
// the line below is from resnext101-32x4d
conv_param_t<>(1, 128, 128, {56, 56}, 32, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(2, 128, 128, {56, 56}, 32, {3, 3}, {1, 1}, {1, 1, 1, 1}),
// Small H and W corner cases
conv_param_t<>(1, 16, 16, {5, 2}, 1, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {2, 5}, 1, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 1}, 1, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {1, 5}, 1, {3, 3}, {1, 1}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 6}, 8, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {6, 5}, 8, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 4}, 8, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {4, 5}, 8, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {4, 4}, 8, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {5, 2}, 8, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {2, 2}, 8, {3, 3}, {2, 2}, {1, 1, 1, 1}),
conv_param_t<>(1, 16, 16, {1, 1}, 8, {3, 3}, {2, 2}, {1, 1, 1, 1}),
// The following lines are commented to reduce test time but still valid
// when we want more extensive testings.
// conv_param_t<>(1, 64, 64, {3, 3}, 8, {3, 3}, {1, 1}, {1, 1, 1, 1}),
// conv_param_t<>(1, 64, 64, {4, 4}, 8, {3, 3}, {1, 1}, {1, 1, 1, 1}),
// conv_param_t<>(1, 64, 64, {3, 5}, 8, {3, 3}, {1, 1}, {1, 1, 1, 1}),
// conv_param_t<>(1, 64, 64, {5, 3}, 8, {3, 3}, {1, 1}, {1, 1, 1, 1}),
// conv_param_t<>(1, 16, 16, {5, 5}, 2, {3, 3}, {1, 1}, {1, 1, 1, 1}),
// conv_param_t<>(1, 256, 256, {56, 48}, 32, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(1, 256, 256, {48, 56}, 32, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(1, 256, 256, {56, 56}, 32, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(2, 256, 256, {56, 56}, 32, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(1, 128, 128, {3, 3}, 8, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(1, 128, 128, {4, 4}, 8, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(1, 128, 128, {3, 5}, 8, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(1, 128, 128, {5, 3}, 8, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(1, 32, 32, {5, 5}, 2, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(1, 512, 512, {56, 48}, 32, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(1, 512, 512, {48, 56}, 32, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(1, 512, 512, {56, 56}, 32, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
// conv_param_t<>(2, 512, 512, {56, 56}, 32, {3, 3}, {1, 1},
// {1, 1, 1, 1}),
};
return shapes;
}
/**
* @brief Unit test for uint8 activations, int8 weights, and 32-bit
* accumulation. Output processing: requantization -> nothing
*/
template <int SPATIAL_DIM = 2>
void runRequantizeTest(matrix_op_t /* unused */,
matrix_op_t btrans,
QuantizationGranularity q_granularity,
bool a_symmetric, bool b_symmetric) {
vector<conv_param_t<SPATIAL_DIM>> shapes(GetShapes_<SPATIAL_DIM>());
for (auto conv_p : shapes) {
int T = SPATIAL_DIM <= 2 ? 1 : conv_p.K[SPATIAL_DIM - 3];
int R = SPATIAL_DIM == 1 ? 1 : conv_p.K[SPATIAL_DIM - 2];
int S = conv_p.K[SPATIAL_DIM - 1];
int G = conv_p.G;
int OC = conv_p.OC;
int IT = SPATIAL_DIM <= 2 ? 1 : conv_p.IN_DIM[SPATIAL_DIM - 3];
int IH = SPATIAL_DIM == 1 ? 1 : conv_p.IN_DIM[SPATIAL_DIM - 2];
int IW = conv_p.IN_DIM[SPATIAL_DIM - 1];
int OT = SPATIAL_DIM <= 2 ? 1 : conv_p.OUT_DIM[SPATIAL_DIM - 3];
int OH = SPATIAL_DIM == 1 ? 1 : conv_p.OUT_DIM[SPATIAL_DIM - 2];
int OW = conv_p.OUT_DIM[SPATIAL_DIM - 1];
int IC_per_G = conv_p.IC / conv_p.G;
int OC_per_G = conv_p.OC / conv_p.G;
// activations
aligned_vector<uint8_t> Aint8(
conv_p.MB * IT * IH *IW * conv_p.IC, 0);
// weights
// when btrans == Transpose, the weight matrix is
// in layout G K/G (T R S C/G) instead of G (T R S C/G) K/G
aligned_vector<int8_t> Bint8(T * R * S * G * IC_per_G * OC_per_G, 0);
aligned_vector<int8_t> Bint8_tr(Bint8.size(), 0);
aligned_vector<int32_t> Cint32_ref(conv_p.MB *OT *OH * OW * OC, 0);
aligned_vector<int32_t> Cint32_fb(Cint32_ref.size(), 0);
aligned_vector<uint8_t> Cint8_ref(Cint32_ref.size(), 0);
aligned_vector<uint8_t> Cint8_fb(Cint32_ref.size(), 0);
randFill<uint8_t>(Aint8, 0, 5);
int32_t Aint8_zero_point = a_symmetric ? 0 : 4;
randFill<int8_t>(Bint8, -4, 4);
// computing column offset
vector<int32_t> col_offsets(G * OC_per_G);
int ncols_per_quant_group = G * OC_per_G;
if (q_granularity == QuantizationGranularity::GROUP) {
ncols_per_quant_group = OC_per_G;
} else if (q_granularity == QuantizationGranularity::OUT_CHANNEL) {
ncols_per_quant_group = 1;
}
aligned_vector<int32_t> Bint8_zero_point(
G * OC_per_G / ncols_per_quant_group);
if (b_symmetric) {
randFill(Bint8_zero_point, 0, 0);
} else {
randFill(Bint8_zero_point, -3, -1);
}
// matrix dimensions after im2col for each GEMM.
// For each group, there is one GEMM of the following dimensions
int MDim = conv_p.MB * OT * OH * OW;
int NDim = OC_per_G;
int KDim = T * R * S * IC_per_G;
vector<uint8_t> Aint8_im2col(MDim * KDim * G);
im2col_ref(conv_p, Aint8.data(), Aint8_zero_point, Aint8_im2col.data());
vector<int32_t> row_offsets(MDim);
aligned_vector<float> C_multiplier(Bint8_zero_point.size());
randFill(C_multiplier, 0.1234f / 2, 0.1234f * 3 / 2);
int32_t C_zero_pt = 5;
// reference implementation
// conv_ref expects weights to be in G (T R S C/G) K/G
int8_t* rightBData = Bint8.data();
if (btrans == matrix_op_t::Transpose) {
transposeConvWeights(conv_p, Bint8.data(), Bint8_tr.data());
rightBData = Bint8_tr.data();
}
for (int g = 0; g < G; ++g) {
col_offsets_with_zero_pt_s8acc32_ref(
R * S * IC_per_G,
OC_per_G,
OC_per_G,
rightBData + g * R * S * IC_per_G * OC_per_G,
Bint8_zero_point.data() + g * OC_per_G / ncols_per_quant_group,
col_offsets.data() + g * OC_per_G,
ncols_per_quant_group);
}
conv_ref(
conv_p, Aint8.data(), Aint8_zero_point, rightBData, Cint32_ref.data());
for (int g = 0; g < G; ++g) {
row_offsets_u8acc32_ref(
MDim,
KDim,
KDim * G,
Aint8_im2col.data() + g * KDim,
row_offsets.data());
requantize_u8acc32_ref(
MDim,
NDim,
G * NDim,
Cint32_ref.data() + g * NDim,
Cint8_ref.data() + g * NDim,
C_multiplier.data() + g * NDim / ncols_per_quant_group,
C_zero_pt,
Aint8_zero_point,
Bint8_zero_point.data() + g * NDim / ncols_per_quant_group,
row_offsets.data(),
col_offsets.data() + g * NDim,
nullptr,
ncols_per_quant_group);
}
PackWeightMatrixForGConv<int8_t, int32_t, SPATIAL_DIM> packedWeights(
btrans, conv_p, Bint8.data(), nullptr);
#ifdef _OPENMP
#pragma omp parallel
#endif
{
vector<int32_t> row_offset_buf(rowOffsetBufferSizeGConv(conv_p));
DoNothing<> doNothingObj{};
int num_threads = fbgemm_get_num_threads();
int tid = fbgemm_get_thread_num();
if (q_granularity == QuantizationGranularity::TENSOR) {
ReQuantizeOutput<false, QuantizationGranularity::TENSOR> reqObj(
doNothingObj,
C_multiplier.data(),
C_zero_pt,
Aint8_zero_point,
Bint8_zero_point.data(),
Bint8_zero_point[0] ? row_offset_buf.data() : nullptr,
col_offsets.data(),
nullptr,
G * NDim,
G);
fbgemmGroupwiseConv(
conv_p,
Aint8.data(),
Aint8_zero_point,
Bint8_zero_point[0] ? row_offset_buf.data() : nullptr,
packedWeights,
Cint8_fb.data(),
Cint32_fb.data(),
reqObj,
tid,
num_threads);
} else if (q_granularity == QuantizationGranularity::GROUP) {
ReQuantizeOutput<false, QuantizationGranularity::GROUP> reqObj(
doNothingObj,
C_multiplier.data(),
C_zero_pt,
Aint8_zero_point,
Bint8_zero_point.data(),
row_offset_buf.data(),
col_offsets.data(),
nullptr,
G * NDim,
G);
fbgemmGroupwiseConv(
conv_p,
Aint8.data(),
Aint8_zero_point,
row_offset_buf.data(),
packedWeights,
Cint8_fb.data(),
Cint32_fb.data(),
reqObj,
tid,
num_threads);
} else {
ReQuantizeOutput<false, QuantizationGranularity::OUT_CHANNEL> reqObj(
doNothingObj,
C_multiplier.data(),
C_zero_pt,
Aint8_zero_point,
Bint8_zero_point.data(),
row_offset_buf.data(),
col_offsets.data(),
nullptr,
G * NDim,
G);
fbgemmGroupwiseConv(
conv_p,
Aint8.data(),
Aint8_zero_point,
row_offset_buf.data(),
packedWeights,
Cint8_fb.data(),
Cint32_fb.data(),
reqObj,
tid,
num_threads);
}
} // omp parallel
compare_validate_buffers(
Cint8_ref.data(),
Cint8_fb.data(),
MDim,
NDim * G,
NDim * G,
static_cast<uint8_t>(0));
} // for each shape
}
TEST_P(fbgemmGConvAcc32WithQuantGranularityTest, requantizeTest) {
matrix_op_t atrans, btrans;
QuantizationGranularity q_granularity;
bool a_symmetric, b_symmetric;
tie(atrans, btrans, q_granularity, a_symmetric, b_symmetric) = GetParam();
runRequantizeTest<2>(atrans, btrans, q_granularity, a_symmetric, b_symmetric);
runRequantizeTest<3>(atrans, btrans, q_granularity, a_symmetric, b_symmetric);
}
/**
* @brief Unit test for uint8 activations, int8 weights, and 32-bit
* accumulation. Output processing: nothing
*/
/*
TEST_P(fbgemmGConvAcc32Test, NoRequantizeTest) {
vector<conv_param_t<>> shapes(GetShapes_());
matrix_op_t atrans, btrans;
tie(atrans, btrans) = GetParam();
for (auto conv_p : shapes) {
int R = conv_p.K[0];
int S = conv_p.K[1];
int G = conv_p.G;
int OC = conv_p.OC;
int OH = conv_p.OUT_DIM[0];
int OW = conv_p.OUT_DIM[1];
int IC_per_G = conv_p.IC / conv_p.G;
int OC_per_G = conv_p.OC / conv_p.G;
// activations
aligned_vector<uint8_t> Aint8(
conv_p.MB * conv_p.IN_DIM[0] * conv_p.IN_DIM[1] * conv_p.IC, 0);
// weights
// when btrans == Transpose, the weight matrix is in layout G K/G (R S C/G)
// instead of G (R S C/G) K/G
aligned_vector<int8_t> Bint8(R * S * conv_p.G * IC_per_G * OC_per_G, 0);
aligned_vector<int8_t> Bint8_tr(R * S * conv_p.G * IC_per_G * OC_per_G, 0);
aligned_vector<int32_t> Cint32_ref(conv_p.MB * OH * OW * OC, 0);
aligned_vector<int32_t> Cint32_fb(Cint32_ref.size(), 0);
randFill<uint8_t>(Aint8, 0, 5);
int32_t Aint8_zero_point = 4;
randFill<int8_t>(Bint8, -4, 4);
// matrix dimensions after im2col for each GEMM.
// For each group, there is one GEMM of the following dimensions
int MDim = conv_p.MB * OH * OW;
int NDim = OC_per_G;
// int KDim = R * S * IC_per_G;
// reference implementation
// conv_ref expects weights to be in G (R S C/G) K/G
int8_t* rightBData = Bint8.data();
if (btrans == matrix_op_t::Transpose) {
transposeConvWeights(conv_p, Bint8.data(), Bint8_tr.data());
rightBData = Bint8_tr.data();
}
conv_ref(
conv_p, Aint8.data(), Aint8_zero_point, rightBData, Cint32_ref.data());
PackWeightMatrixForGConv<int8_t> packedWeights(
btrans, conv_p, Bint8.data(), nullptr);
// TODO: Uncomment once we support multiple threads in fbgemmGroupwiseConv
// #ifdef _OPENMP
// #pragma omp parallel
// #endif
{
vector<int32_t> row_offset_buf(rowOffsetBufferSizeGConv(conv_p));
DoNothing<int32_t, int32_t> doNothingObj{};
int num_threads = fbgemm_get_num_threads();
int tid = fbgemm_get_thread_num();
fbgemmGroupwiseConv(
conv_p,
Aint8.data(),
Aint8_zero_point,
row_offset_buf.data(),
packedWeights,
Cint32_fb.data(),
Cint32_fb.data(),
doNothingObj,
tid,
num_threads);
}
compare_validate_buffers(
Cint32_ref.data(),
Cint32_fb.data(),
MDim,
NDim * G,
NDim * G,
static_cast<int32_t>(0));
} // for each shape
}
*/
template <int SPATIAL_DIM = 2>
void runPackUnpackTest(matrix_op_t btrans) {
vector<conv_param_t<SPATIAL_DIM>> shapes(GetShapes_<SPATIAL_DIM>());
for (auto conv_p : shapes) {
int T = SPATIAL_DIM <= 2 ? 1 : conv_p.K[SPATIAL_DIM - 3];
int R = SPATIAL_DIM == 1 ? 1 : conv_p.K[SPATIAL_DIM - 2];
int S = conv_p.K[SPATIAL_DIM - 1];
int IC_per_G = conv_p.IC / conv_p.G;
int OC_per_G = conv_p.OC / conv_p.G;
// Weights -- test the packing/unpacking of only the weights
// when btrans == Transpose, the weight matrix is in
// layout G K/G (T R S C/G) instead of G (T R S C/G) K/G
int weight_len = T * R * S * conv_p.G * IC_per_G * OC_per_G;
aligned_vector<int8_t> Bint8(weight_len, 0);
// Random fill the weights
randFill<int8_t>(Bint8, -4, 4);
// Instantiate the object
PackWeightMatrixForGConv<int8_t, int32_t, SPATIAL_DIM> packedWeights(
btrans, conv_p, Bint8.data(), nullptr);
// Setup a buffer to get pack -> unpacked results
aligned_vector<int8_t> unpack_buf(weight_len, 0);
// START Actual pack-unpack operations
// Perform packing first. This should populate pdata_ of packedWeights
packedWeights.pack();
// Next perform unpacking
packedWeights.unpack(unpack_buf.data());
// END actual pack-unpack operations
// Sanity check
for (int i = 0; i < weight_len; ++i) {
EXPECT_EQ(unpack_buf.data()[i], Bint8.data()[i])
<< "Pack/Unpack results differ at index " << i
<< ", Reference: " << static_cast<int>(Bint8.data()[i])
<< ", Pack-Unpacked: " << static_cast<int>(unpack_buf.data()[i]);
}
} // for each shape
}
/**
* @brief Unit test for packing and unpacking the weight tensor
*/
TEST_P(fbgemmGConvPackTest, PackUnpackTest) {
matrix_op_t btrans = GetParam();
runPackUnpackTest<2>(btrans);
runPackUnpackTest<3>(btrans);
}