Skip to content
This repository has been archived by the owner on Nov 17, 2023. It is now read-only.

[Improvement] Invoke mkldnn and cudnn BatchNorm when axis != 1 #18504

Merged
merged 21 commits into from
Jul 9, 2020
Merged

[Improvement] Invoke mkldnn and cudnn BatchNorm when axis != 1 #18504

Show file tree
Hide file tree
Changes from 20 commits
Commits
Show all changes
21 commits
Select commit Hold shift + click to select a range
fac3b42
fix batch norm when fix_gamma is True
wkcn Jun 5, 2020
34c6a3b
support gradient accumulation for batch norm
wkcn Jun 6, 2020
11c8e26
mkldnn batchnorm support grad add
wkcn Jun 6, 2020
830ba0d
unittest for bn
wkcn Jun 6, 2020
a9c91d2
fix bn arg
wkcn Jun 6, 2020
8594642
fix lint
wkcn Jun 6, 2020
f9995b0
fix mkldnn
wkcn Jun 6, 2020
d380443
fix mkldnn bn
wkcn Jun 6, 2020
eaeae21
fix grad when fixing gamma
wkcn Jun 6, 2020
25924c9
fix naive gpu bn
wkcn Jun 6, 2020
e3bb53c
fix lint
wkcn Jun 6, 2020
9c00567
invoke mkldnn and cudnn batchnorm when axis != 1
wkcn Jun 6, 2020
4cab1ed
backport 18500
wkcn Jun 8, 2020
4bf0000
Merge branch 'master' into invoke_mkldnn_cudnn_bn_when_axis_ne_1
wkcn Jun 8, 2020
b78b4cb
change condition
wkcn Jun 8, 2020
e411890
fix
wkcn Jun 9, 2020
aad431b
fix
wkcn Jun 9, 2020
939750f
add mkldnn_off for bn
wkcn Jun 11, 2020
25e4666
remove mkldnn_off
wkcn Jul 1, 2020
97c6746
recover save_000800.json
wkcn Jul 1, 2020
38da95f
cast
wkcn Jul 5, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 8 additions & 4 deletions src/operator/nn/batch_norm.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -420,10 +420,14 @@ static bool BatchNormType(const nnvm::NodeAttrs& attrs,

#if MXNET_USE_MKLDNN == 1
static inline bool SupportMKLDNNBN(const NDArray &input, const BatchNormParam &param) {
mxnet::TShape shape = input.shape();
return SupportMKLDNN(input) && shape.ndim() == 4
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

we are removing the check for ndim == 4 here, and another lighter check for ndim == 1 || ndim ==2 || ndim ==4 present in SupportMKLDNN.
Does that mean ndim can be anything >0 ? What are the allowed values for ndim?

Copy link
Member Author

@wkcn wkcn Jul 3, 2020 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, ndim coulld be anything > 0.

If the shape of input A is shape, the input will be shaped into (prod(shape[0:axis]), shape[axis], 1, prod(shape[axis+1:len(shape)]) ).

ndim shape axis=0 axis=1 axis=2 axis=3 axis=4
1 (N,) (1, N, 1, 1) x x x x
2 (N,C) (1, N, 1, C) (N, C, 1, 1) x x x
3 (N,C,H) (1, N, 1, CH) (N, C, 1, H) (NC, H, 1, 1) x x
4 (N,C,H,W) (1, N, 1, CHW) (N, C, 1, HW) (NC, H, 1, W) (NCH, W, 1, 1) x
5 (N,D,C,H,W) (1, N, 1, DCHW) (N, D, 1, CHW) (ND,C, 1, HW) (NDC, H, 1, W) (NDCH, W, 1, 1)

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for the explanation! Does the table continue further for ndims > 5? Or should we place a check for that?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, it supports ndim > 5 too.

&& param.axis == mxnet::op::batchnorm::DEFAULT_AXIS
&& !mxnet::op::batchnorm::disable_mkl;
if (mxnet::op::batchnorm::disable_mkl) return false;
const mxnet::TShape shape = input.shape();
const int ndim = shape.ndim();
if (ndim == 0 || shape.Size() == 0) return false;
const int dtype = input.dtype();
return (dtype == mshadow::kFloat32 ||
dtype == mshadow::kBfloat16) &&
SupportStorageMKLDNN(input.storage_type());
}

void BatchNormComputeExCPU(const nnvm::NodeAttrs &attrs,
Expand Down
6 changes: 2 additions & 4 deletions src/operator/nn/batch_norm.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -698,8 +698,7 @@ void BatchNormCompute<gpu>(const nnvm::NodeAttrs& attrs,

param.axis = mxnet::op::batchnorm::GetRealAxis(shape, param.axis);
#if MXNET_USE_CUDNN == 1
if (!param.use_global_stats && !param.cudnn_off
&& param.axis == mxnet::op::batchnorm::DEFAULT_AXIS) {
if (!param.use_global_stats && !param.cudnn_off) {
MSHADOW_REAL_TYPE_SWITCH(dtype, DType, {
GetCuDNNOp<DType>(param).Forward(ctx, in_data, req, outputs, aux_states);
})
Expand Down Expand Up @@ -727,8 +726,7 @@ void BatchNormGradCompute<gpu>(const nnvm::NodeAttrs& attrs,

param.axis = mxnet::op::batchnorm::GetRealAxis(shape, param.axis);
#if MXNET_USE_CUDNN == 1
if (!param.use_global_stats && !param.cudnn_off
&& param.axis == mxnet::op::batchnorm::DEFAULT_AXIS) {
if (!param.use_global_stats && !param.cudnn_off) {
MSHADOW_REAL_TYPE_SWITCH(dtype, DType, {
GetCuDNNOp<DType>(param).Backward(ctx, inputs, req, outputs);
})
Expand Down
25 changes: 18 additions & 7 deletions src/operator/nn/cudnn/cudnn_batch_norm-inl.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -262,15 +262,26 @@ class CuDNNBatchNormOp {

private:
void Init(const TBlob &in_data) {
if (in_data.ndim() == 4) {
for (int i = 0; i < 4; ++i)
shape_[i] = in_data.shape_[i];
CHECK_GE(param_.axis, 0);
CHECK_LT(param_.axis, in_data.ndim());
if (param_.axis == 1) {
if (in_data.ndim() == 4) {
for (int i = 0; i < 4; ++i)
shape_[i] = in_data.shape_[i];
} else {
// when in_data.ndim() != 4
shape_[0] = in_data.shape_[0];
shape_[1] = in_data.ndim() > 1 ? in_data.shape_[1] : 1;
shape_[2] = 1;
shape_[3] = in_data.shape_.ProdShape(2, in_data.ndim());
}
} else {
// when in_data.ndim() != 4
shape_[0] = in_data.shape_[0];
shape_[1] = in_data.ndim() > 1 ? in_data.shape_[1] : 1;
// reshape to (N, C, 1, D), C is the `param_.axis` dimension
shape_[0] = static_cast<dim_t>(in_data.shape_.ProdShape(0, param_.axis));
shape_[1] = in_data.shape_[param_.axis];
shape_[2] = 1;
shape_[3] = in_data.shape_.ProdShape(2, in_data.ndim());
shape_[3] = static_cast<dim_t>(in_data.shape_.ProdShape(param_.axis + 1,
static_cast<int>(in_data.ndim())));
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

why do we need static_cast here? why cant we do it like line 276?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The return dtype of ProdShape is size_t, a unsigned int, but the dtype of shape_[i] is dim_t, namely int64_t.
I do not know whether the static_cast is neceseary to avoid the potential compiler's warning.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we add static_cast to line 276 as well?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

in_data.ndims() shouldn’t need a static cast though, right?

Copy link
Member Author

@wkcn wkcn Jul 3, 2020 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, the dtype of .ndim() is int32_t.

variable dtype
axis int32_t
ProdShape size_t
shape_[i] dim_t, int64_t
ndim() int32_t

The signature of ProdShape is size_t ProdShape(int dimstart, int dimend) const.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hi @mseth10 , do you have any suggestion about whether to use static_cast ?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Imo you can ignore the compiler warnings (if any) for int32_t to int,
you can keep static_cast for size_t to dim_t

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I see. I have updated it : )

}

CUDNN_CALL(cudnnSetTensor4dDescriptor(io_desc_,
Expand Down
44 changes: 39 additions & 5 deletions src/operator/nn/mkldnn/mkldnn_batch_norm-inl.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -157,7 +157,25 @@ void MKLDNNBatchNormForward(const nnvm::NodeAttrs &attrs, const OpContext &ctx,
const std::vector<NDArray> &inputs, const std::vector<OpReqType> &req,
const std::vector<NDArray> &outputs, bool fuse_relu) {
const BatchNormParam &param = nnvm::get<BatchNormParam>(attrs.parsed);
const std::vector<NDArray> in_data(inputs.begin(), inputs.begin() + batchnorm::kInMovingMean);
std::vector<NDArray> in_data(inputs.begin(), inputs.begin() + batchnorm::kInMovingMean);

mxnet::TShape shape = inputs[batchnorm::kData].shape();
const int real_axis = mxnet::op::batchnorm::GetRealAxis(shape, param.axis);
CHECK_LT(real_axis, shape.ndim());
NDArray out = outputs[batchnorm::kOut];
if (param.axis != 1 || shape.ndim() != 4) {
// reshape to (N, C, 1, D)
mxnet::TShape new_shape{
static_cast<dim_t>(shape.ProdShape(0, real_axis)),
shape[real_axis],
1,
static_cast<dim_t>(shape.ProdShape(real_axis + 1,
static_cast<int>(shape.ndim())))
};
in_data[batchnorm::kData] = in_data[batchnorm::kData].Reshape(new_shape);
out = out.Reshape(new_shape);
}

const std::vector<NDArray> aux_states(inputs.begin() + batchnorm::kInMovingMean, inputs.end());
TmpMemMgr::Get()->Init(ctx.requested[batchnorm::kTempSpace]);
mkldnn::normalization_flags flags = _GetFlags(in_data,
Expand All @@ -166,7 +184,6 @@ void MKLDNNBatchNormForward(const nnvm::NodeAttrs &attrs, const OpContext &ctx,
fuse_relu);
const NDArray &data = in_data[batchnorm::kData];
auto &fwd = GetBNForward<DType>(param, ctx, data, flags);
const NDArray &out = outputs[batchnorm::kOut];

// for output memory
auto out_mem = const_cast<NDArray &>(out).CreateMKLDNNData(fwd.GetPd().dst_desc());
Expand Down Expand Up @@ -325,9 +342,9 @@ void MKLDNNBatchNormBackward(const nnvm::NodeAttrs &attrs, const OpContext &ctx,
ctx.is_train && !param.use_global_stats,
fuse_relu);

const NDArray &data = in_data[batchnorm::kData];
const NDArray &diff = out_grad[batchnorm::kOut];
const NDArray &gradIn = in_grad[batchnorm::kData];
NDArray data = in_data[batchnorm::kData];
NDArray diff = out_grad[batchnorm::kOut];
NDArray gradIn = in_grad[batchnorm::kData];
const NDArray &moving_mean = aux_states[batchnorm::kMovingMean];
const NDArray &moving_var = aux_states[batchnorm::kMovingVar];
const NDArray &out_mean = out_data[batchnorm::kMean];
Expand All @@ -338,6 +355,23 @@ void MKLDNNBatchNormBackward(const nnvm::NodeAttrs &attrs, const OpContext &ctx,
CHECK(moving_mean.IsDefaultData());
CHECK(moving_var.IsDefaultData());

mxnet::TShape shape = data.shape();
const int real_axis = mxnet::op::batchnorm::GetRealAxis(shape, param.axis);
CHECK_LT(real_axis, shape.ndim());
if (param.axis != 1 || shape.ndim() != 4) {
// reshape to (N, C, 1, D)
mxnet::TShape new_shape{
static_cast<dim_t>(shape.ProdShape(0, real_axis)),
shape[real_axis],
1,
static_cast<dim_t>(shape.ProdShape(real_axis + 1,
static_cast<int>(shape.ndim())))
};
data = data.Reshape(new_shape);
diff = diff.Reshape(new_shape);
gradIn = gradIn.Reshape(new_shape);
}

auto data_mem = data.GetMKLDNNData();
auto diff_mem = diff.GetMKLDNNData();
// MKLDNN batchnorm should run on special layouts. If one of them isn't, we
Expand Down