Browse Source
fix(mgb/gopt): fix global layout transform
add a special opr_format modify function for concat operators to modify concat axis when input's layout has been changed
GitOrigin-RevId: 4094208057
release-1.7
Megvii Engine Team
3 years ago
14 changed files with 535 additions and 40 deletions
Unified View
Diff Options
-
+2 -2dnn/src/common/convolution.cpp
-
+2 -0src/gopt/impl/framework.cpp
-
+0 -1src/gopt/impl/global_layout_transform/layout_transform_context.cpp
-
+7 -7src/gopt/impl/global_layout_transform/layout_transform_pass.cpp
-
+63 -5src/gopt/impl/global_layout_transform/opr_format_modifier.cpp
-
+20 -1src/gopt/impl/global_layout_transform/opr_format_modifier.h
-
+24 -2src/gopt/impl/global_layout_transform/opr_tensor_formats_config.cpp
-
+41 -10src/gopt/impl/global_layout_transform/profiler_impl.cpp
-
+22 -4src/gopt/impl/global_layout_transform/reformat_manager.cpp
-
+2 -1src/gopt/impl/global_layout_transform/utils.h
-
+7 -0src/gopt/include/megbrain/gopt/reformat_manager.h
-
+214 -0src/gopt/test/cache_data.h
-
+21 -7src/gopt/test/embed_cache.py
-
+110 -0src/gopt/test/layout_transform_pass.cpp
+ 2
- 2
dnn/src/common/convolution.cpp
View File
| @@ -831,7 +831,7 @@ typename ConvolutionBase<Parameter>::CanonizedFilterMeta ConvolutionBase<Paramet | |||||
| dst[4] = 32; | dst[4] = 32; | ||||
| } else if (param().format == Param::Format::NCHW88) { | } else if (param().format == Param::Format::NCHW88) { | ||||
| megdnn_assert( | megdnn_assert( | ||||
| src.ndim == 5 || src.ndim == 4, | |||||
| src.ndim == 5 || (src.ndim == 4 && src[1] <= 8), | |||||
| "invalid src ndim for NCHW88, expected=5 or 4, got=%zu", src.ndim); | "invalid src ndim for NCHW88, expected=5 or 4, got=%zu", src.ndim); | ||||
| dst.ndim = 5; | dst.ndim = 5; | ||||
| dst[0] = src[0]; | dst[0] = src[0]; | ||||
| @@ -854,7 +854,7 @@ typename ConvolutionBase<Parameter>::CanonizedFilterMeta ConvolutionBase<Paramet | |||||
| param().format == Param::Format::NCHW44 || | param().format == Param::Format::NCHW44 || | ||||
| param().format == Param::Format::NCHW44_DOT) { | param().format == Param::Format::NCHW44_DOT) { | ||||
| megdnn_assert( | megdnn_assert( | ||||
| src.ndim == 5 || src.ndim == 4, | |||||
| src.ndim == 5 || (src.ndim == 4 && src[1] <= 4), | |||||
| "invalid src ndim for NCHW44, expected=5 or 4, got=%zu", src.ndim); | "invalid src ndim for NCHW44, expected=5 or 4, got=%zu", src.ndim); | ||||
| dst.ndim = 5; | dst.ndim = 5; | ||||
| dst[0] = src[0]; | dst[0] = src[0]; | ||||
+ 2
- 0
src/gopt/impl/framework.cpp
View File
| @@ -840,7 +840,9 @@ const GraphOptimizer& GraphOptimizer::add_passes_for_graph_tuning_options( | |||||
| if (need_param_fuse) { | if (need_param_fuse) { | ||||
| add_pass<ParamFusePass>(); | add_pass<ParamFusePass>(); | ||||
| add_pass<ParamMergePass>(); | |||||
| } | } | ||||
| return *this; | return *this; | ||||
| } | } | ||||
+ 0
- 1
src/gopt/impl/global_layout_transform/layout_transform_context.cpp
View File
| @@ -66,7 +66,6 @@ std::unique_ptr<LayoutTransformContext> make_cuda_ctx( | |||||
| ctx->add_opr_config( | ctx->add_opr_config( | ||||
| opr::ConvBiasForward::typeinfo(), | opr::ConvBiasForward::typeinfo(), | ||||
| {OprFormatConfigID::NCHW, OprFormatConfigID::NHWC, | {OprFormatConfigID::NCHW, OprFormatConfigID::NHWC, | ||||
| OprFormatConfigID::NCHW4_NCHW32, OprFormatConfigID::NCHW32_NCHW4, | |||||
| OprFormatConfigID::NCHW4, OprFormatConfigID::NCHW32, | OprFormatConfigID::NCHW4, OprFormatConfigID::NCHW32, | ||||
| OprFormatConfigID::NCHW64, OprFormatConfigID::CHWN4}) | OprFormatConfigID::NCHW64, OprFormatConfigID::CHWN4}) | ||||
| .add_opr_config( | .add_opr_config( | ||||
+ 7
- 7
src/gopt/impl/global_layout_transform/layout_transform_pass.cpp
View File
| @@ -18,6 +18,7 @@ | |||||
| #include "megbrain/gopt/solver.h" | #include "megbrain/gopt/solver.h" | ||||
| #include "megbrain/opr/dnn/pooling.h" | #include "megbrain/opr/dnn/pooling.h" | ||||
| #include "megbrain/opr/imgproc.h" | #include "megbrain/opr/imgproc.h" | ||||
| #include "megbrain/opr/tensor_manip.h" | |||||
| #include "megbrain/serialization/sereg.h" | #include "megbrain/serialization/sereg.h" | ||||
| #include "megbrain/utils/hash_ct.h" | #include "megbrain/utils/hash_ct.h" | ||||
| @@ -64,11 +65,6 @@ void LayoutTransformPass::apply(OptState& opt) const { | |||||
| auto&& base_cfg_id = m_ctx->attribute().base_config_id; | auto&& base_cfg_id = m_ctx->attribute().base_config_id; | ||||
| auto&& reformat_attribute = m_ctx->attribute().reformat_attribute; | auto&& reformat_attribute = m_ctx->attribute().reformat_attribute; | ||||
| ThinHashMap<VarNode*, TensorFormats> var2fmts; | ThinHashMap<VarNode*, TensorFormats> var2fmts; | ||||
| static ThinHashSet<Typeinfo*> format_aware_oprs = { | |||||
| #define cb(_Opr) opr::_Opr::typeinfo(), | |||||
| FOREACH_FORMAT_AWARE_OPR(cb) | |||||
| #undef cb | |||||
| }; | |||||
| auto rewriter = opt.graph().make_rewriter(); | auto rewriter = opt.graph().make_rewriter(); | ||||
| auto on_opr = [&opr_configs, &base_fmt, &base_cfg_id, &reformat_attribute, | auto on_opr = [&opr_configs, &base_fmt, &base_cfg_id, &reformat_attribute, | ||||
| &rewriter, &solution, &var2fmts, | &rewriter, &solution, &var2fmts, | ||||
| @@ -141,8 +137,12 @@ void LayoutTransformPass::apply(OptState& opt) const { | |||||
| new_inp[i] = new_var; | new_inp[i] = new_var; | ||||
| } | } | ||||
| VarNode* new_out; | VarNode* new_out; | ||||
| if (format_aware_oprs.count(opr->dyn_typeinfo()) > 0) { | |||||
| new_out = intl::modify_opr_format(opr_fmt.val(), new_inp, opr); | |||||
| if (intl::has_opr_format_modifier(opr)) { | |||||
| intl::OprFormatInfo opr_format_info; | |||||
| opr_format_info.opr_format = opr_fmt.val(); | |||||
| opr_format_info.tensor_formats = { | |||||
| base_fmt, opr_format_to_tensor_formats(opr_fmt.val())}; | |||||
| new_out = intl::modify_opr_format(opr_format_info, new_inp, opr); | |||||
| } else { | } else { | ||||
| new_out = serialization::copy_opr_shallow(*opr, new_inp, opr->config()) | new_out = serialization::copy_opr_shallow(*opr, new_inp, opr->config()) | ||||
| ->output(0); | ->output(0); | ||||
+ 63
- 5
src/gopt/impl/global_layout_transform/opr_format_modifier.cpp
View File
| @@ -11,10 +11,12 @@ | |||||
| */ | */ | ||||
| #include "./opr_format_modifier.h" | #include "./opr_format_modifier.h" | ||||
| #include "./utils.h" | |||||
| #include "megbrain/opr/dnn/convolution.h" | #include "megbrain/opr/dnn/convolution.h" | ||||
| #include "megbrain/opr/dnn/pooling.h" | #include "megbrain/opr/dnn/pooling.h" | ||||
| #include "megbrain/opr/imgproc.h" | #include "megbrain/opr/imgproc.h" | ||||
| #include "megbrain/opr/io.h" | #include "megbrain/opr/io.h" | ||||
| #include "megbrain/opr/tensor_manip.h" | |||||
| #include "megbrain/serialization/sereg.h" | #include "megbrain/serialization/sereg.h" | ||||
| #include "midout.h" | #include "midout.h" | ||||
| @@ -201,6 +203,37 @@ INST(ConvolutionBackwardData) | |||||
| INST(PoolingForward) | INST(PoolingForward) | ||||
| #undef APPLY | #undef APPLY | ||||
| #undef INST | #undef INST | ||||
| VarNode* modify_concat_opr_format( | |||||
| gopt::intl::OprFormatInfo::TensorFormatsInfo tensor_formats, | |||||
| const VarNodeArray& i, const cg::OperatorNodeBase* opr) { | |||||
| auto base_format = tensor_formats.from; | |||||
| auto tensor_format = tensor_formats.to; | |||||
| int axis = opr->cast_final_safe<Concat>().axis(); | |||||
| /// modify axis | |||||
| using Dimension = megdnn::Dimension; | |||||
| static constexpr uint32_t UNDETERMINED_EXTENT = Dimension::UNDETERMINED_EXTENT; | |||||
| auto orig_shape = tensor_formats_to_named_tensor_shape(base_format); | |||||
| auto target_shape = tensor_formats_to_named_tensor_shape(tensor_format); | |||||
| mgb_assert( | |||||
| static_cast<size_t>(axis) < orig_shape.ndim, | |||||
| "invalid axis of concat opr(axis:%d,shp:%s)", axis, | |||||
| orig_shape.to_string().c_str()); | |||||
| if (orig_shape[axis].extent() != UNDETERMINED_EXTENT) | |||||
| return nullptr; | |||||
| auto axis_name = orig_shape[axis].name(); | |||||
| int new_axis = target_shape.ndim; | |||||
| for (size_t i = 0; i < target_shape.ndim; ++i) { | |||||
| if (target_shape[i].name() == axis_name && | |||||
| target_shape[i].extent() == UNDETERMINED_EXTENT) { | |||||
| new_axis = i; | |||||
| break; | |||||
| } | |||||
| } | |||||
| if (static_cast<size_t>(new_axis) >= target_shape.ndim) | |||||
| return nullptr; | |||||
| return opr::Concat::make(i, new_axis, opr->config()).node(); | |||||
| } | |||||
| } // namespace | } // namespace | ||||
| namespace mgb { | namespace mgb { | ||||
| @@ -275,13 +308,16 @@ INST(Resize, 2); | |||||
| #undef INST | #undef INST | ||||
| VarNode* modify_opr_format( | VarNode* modify_opr_format( | ||||
| opr::ConvBias::Param::Format opr_format, const VarNodeArray& i, | |||||
| OprFormatInfo opr_format_info, const VarNodeArray& i, | |||||
| const cg::OperatorNodeBase* opr) { | const cg::OperatorNodeBase* opr) { | ||||
| #define cb(_Opr) \ | |||||
| if (opr->dyn_typeinfo() == _Opr::typeinfo()) { \ | |||||
| return OprFormatModifier<_Opr>::make(opr_format, i, opr); \ | |||||
| #define cb(_Opr) \ | |||||
| if (opr->dyn_typeinfo() == _Opr::typeinfo()) { \ | |||||
| return OprFormatModifier<_Opr>::make(opr_format_info.opr_format, i, opr); \ | |||||
| } else | } else | ||||
| FOREACH_FORMAT_AWARE_OPR(cb) { | |||||
| FOREACH_FORMAT_AWARE_OPR(cb) | |||||
| if (opr->dyn_typeinfo() == opr::Concat::typeinfo()) { | |||||
| return modify_concat_opr_format(opr_format_info.tensor_formats, i, opr); | |||||
| } else { | |||||
| mgb_throw( | mgb_throw( | ||||
| InternalError, "invalid format aware operator(got:%s)", | InternalError, "invalid format aware operator(got:%s)", | ||||
| opr->dyn_typeinfo()->name); | opr->dyn_typeinfo()->name); | ||||
| @@ -302,6 +338,28 @@ bool has_available_algo(const VarNodeArray& i, const cg::OperatorNodeBase* opr) | |||||
| InternalError, "invalid multi-algo operator(got:%s)", | InternalError, "invalid multi-algo operator(got:%s)", | ||||
| opr->dyn_typeinfo()->name); | opr->dyn_typeinfo()->name); | ||||
| } | } | ||||
| #undef cb | |||||
| } | |||||
| bool has_opr_format(const cg::OperatorNodeBase* opr) { | |||||
| bool ret = false; | |||||
| #define cb(_Opr) ret |= opr->dyn_typeinfo() == _Opr::typeinfo(); | |||||
| FOREACH_FORMAT_AWARE_OPR(cb) | |||||
| #undef cb | |||||
| return ret; | |||||
| } | |||||
| bool has_opr_format_modifier(const cg::OperatorNodeBase* opr) { | |||||
| bool ret = false; | |||||
| #define cb(_Opr) ret |= opr->dyn_typeinfo() == _Opr::typeinfo(); | |||||
| FOREACH_MODIFY_OPR_FORMAT_OPR(cb) | |||||
| #undef cb | |||||
| return ret; | |||||
| } | |||||
| bool allow_aligned_layout(const cg::OperatorNodeBase* opr) { | |||||
| return opr->dyn_typeinfo() != opr::Concat::typeinfo() && | |||||
| opr->dyn_typeinfo() != opr::Reduce::typeinfo(); | |||||
| } | } | ||||
| } // namespace intl | } // namespace intl | ||||
+ 20
- 1
src/gopt/impl/global_layout_transform/opr_format_modifier.h
View File
| @@ -16,17 +16,36 @@ | |||||
| namespace mgb { | namespace mgb { | ||||
| namespace gopt { | namespace gopt { | ||||
| enum class TensorFormats : uint32_t; | |||||
| namespace intl { | namespace intl { | ||||
| #define FOREACH_FORMAT_AWARE_OPR(cb) \ | #define FOREACH_FORMAT_AWARE_OPR(cb) \ | ||||
| cb(Convolution) cb(ConvBiasForward) cb(ConvolutionBackwardData) cb(PoolingForward) \ | cb(Convolution) cb(ConvBiasForward) cb(ConvolutionBackwardData) cb(PoolingForward) \ | ||||
| cb(WarpPerspective) cb(Resize) | cb(WarpPerspective) cb(Resize) | ||||
| #define FOREACH_MODIFY_OPR_FORMAT_OPR(cb) FOREACH_FORMAT_AWARE_OPR(cb) cb(Concat) | |||||
| bool has_available_algo(const VarNodeArray& i, const cg::OperatorNodeBase* opr); | bool has_available_algo(const VarNodeArray& i, const cg::OperatorNodeBase* opr); | ||||
| struct OprFormatInfo { | |||||
| opr::Convolution::Param::Format opr_format; | |||||
| struct TensorFormatsInfo { | |||||
| TensorFormats from; | |||||
| TensorFormats to; | |||||
| }; | |||||
| TensorFormatsInfo tensor_formats; | |||||
| }; | |||||
| VarNode* modify_opr_format( | VarNode* modify_opr_format( | ||||
| opr::Convolution::Param::Format opr_format, const VarNodeArray& i, | |||||
| OprFormatInfo opr_format, const VarNodeArray& i, | |||||
| const cg::OperatorNodeBase* opr); | const cg::OperatorNodeBase* opr); | ||||
| bool has_opr_format(const cg::OperatorNodeBase* opr); | |||||
| bool has_opr_format_modifier(const cg::OperatorNodeBase* opr); | |||||
| bool allow_aligned_layout(const cg::OperatorNodeBase* opr); | |||||
| } // namespace intl | } // namespace intl | ||||
| } // namespace gopt | } // namespace gopt | ||||
| } // namespace mgb | } // namespace mgb | ||||
+ 24
- 2
src/gopt/impl/global_layout_transform/opr_tensor_formats_config.cpp
View File
| @@ -625,7 +625,13 @@ struct ConvTensorFormatsDispatcherImpl<Opr, OprFormatConfigID::NCHW44_HYBRID> { | |||||
| TensorType tensor_type = i == 1 ? TensorType::WEIGHT : TensorType::FEATURE; | TensorType tensor_type = i == 1 ? TensorType::WEIGHT : TensorType::FEATURE; | ||||
| config.input_tensor_types.emplace_back(tensor_type); | config.input_tensor_types.emplace_back(tensor_type); | ||||
| } | } | ||||
| available &= check_dtype(opr->output(0)->dtype(), false); | |||||
| // FIXME: hack for nchw nchw44 hybrid mode | |||||
| static_assert( | |||||
| std::is_same<Opr, opr::ConvolutionForward>::value || | |||||
| std::is_same<Opr, opr::ConvBiasForward>::value, | |||||
| "nchw44 hybrid only support conv or conv_bias opr"); | |||||
| size_t in_channel = opr->input(0)->shape()[1]; | |||||
| available &= in_channel <= 4_z && check_dtype(opr->output(0)->dtype(), false); | |||||
| config.output_dtypes.emplace_back(opr->output(0)->dtype().enumv()); | config.output_dtypes.emplace_back(opr->output(0)->dtype().enumv()); | ||||
| available &= conv.param().sparse == Opr::Param::Sparse::DENSE; | available &= conv.param().sparse == Opr::Param::Sparse::DENSE; | ||||
| config.input_tensor_formats = { | config.input_tensor_formats = { | ||||
| @@ -696,7 +702,14 @@ struct ConvTensorFormatsDispatcherImpl<Opr, OprFormatConfigID::NCHW88_HYBRID> { | |||||
| TensorType tensor_type = i == 1 ? TensorType::WEIGHT : TensorType::FEATURE; | TensorType tensor_type = i == 1 ? TensorType::WEIGHT : TensorType::FEATURE; | ||||
| config.input_tensor_types.emplace_back(tensor_type); | config.input_tensor_types.emplace_back(tensor_type); | ||||
| } | } | ||||
| available &= opr->output(0)->dtype().enumv() == DTypeEnum::Float32; | |||||
| // FIXME: hack for nchw nchw88 hybrid mode | |||||
| static_assert( | |||||
| std::is_same<Opr, opr::ConvolutionForward>::value || | |||||
| std::is_same<Opr, opr::ConvBiasForward>::value, | |||||
| "nchw nchw88 hybrid mode only support conv or conv_bias opr"); | |||||
| size_t in_channel = opr->input(0)->shape()[1]; | |||||
| available &= in_channel <= 8_z && | |||||
| opr->output(0)->dtype().enumv() == DTypeEnum::Float32; | |||||
| config.output_dtypes.emplace_back(opr->output(0)->dtype().enumv()); | config.output_dtypes.emplace_back(opr->output(0)->dtype().enumv()); | ||||
| available &= conv.param().sparse == Opr::Param::Sparse::DENSE; | available &= conv.param().sparse == Opr::Param::Sparse::DENSE; | ||||
| // setup tensor formats | // setup tensor formats | ||||
| @@ -783,6 +796,13 @@ struct ConvTensorFormatsDispatcherImpl<Opr, OprFormatConfigID::NCHW44_DOT_HYBRID | |||||
| opr->output(0)->dtype().enumv() == DTypeEnum::Quantized8Asymm; | opr->output(0)->dtype().enumv() == DTypeEnum::Quantized8Asymm; | ||||
| config.output_dtypes.emplace_back(opr->output(0)->dtype().enumv()); | config.output_dtypes.emplace_back(opr->output(0)->dtype().enumv()); | ||||
| available &= conv.param().sparse == Opr::Param::Sparse::DENSE; | available &= conv.param().sparse == Opr::Param::Sparse::DENSE; | ||||
| // FIXME: hack for nchw nchw44 dot hybrid mode | |||||
| static_assert( | |||||
| std::is_same<Opr, opr::ConvolutionForward>::value || | |||||
| std::is_same<Opr, opr::ConvBiasForward>::value, | |||||
| "nchw44 dot hybrid only support conv or conv_bias opr"); | |||||
| size_t in_channel = opr->input(0)->shape()[1]; | |||||
| available &= in_channel <= 4_z; | |||||
| // setup tensor formats | // setup tensor formats | ||||
| config.input_tensor_formats = { | config.input_tensor_formats = { | ||||
| TensorFormats::NCHW, TensorFormats::KRSCk4, TensorFormats::NCHWc4, | TensorFormats::NCHW, TensorFormats::KRSCk4, TensorFormats::NCHWc4, | ||||
| @@ -940,6 +960,8 @@ StaticData::StaticData() { | |||||
| OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW4); | OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW4); | ||||
| OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, CHWN4); | OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, CHWN4); | ||||
| OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW32); | OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW32); | ||||
| OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW32_NCHW4); | |||||
| OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW4_NCHW32); | |||||
| OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW64); | OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW64); | ||||
| OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW44); | OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW44); | ||||
| OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW88); | OPR_TENSOR_FORMATS_CONFIG_REG(ConvBias, NCHW88); | ||||
+ 41
- 10
src/gopt/impl/global_layout_transform/profiler_impl.cpp
View File
| @@ -19,6 +19,7 @@ | |||||
| #include "megbrain/opr/imgproc.h" | #include "megbrain/opr/imgproc.h" | ||||
| #include "megbrain/opr/io.h" | #include "megbrain/opr/io.h" | ||||
| #include "megbrain/opr/nn_int.h" | #include "megbrain/opr/nn_int.h" | ||||
| #include "megbrain/opr/tensor_manip.h" | |||||
| #include "megbrain/plugin/base.h" | #include "megbrain/plugin/base.h" | ||||
| #include "megbrain/serialization/sereg.h" | #include "megbrain/serialization/sereg.h" | ||||
| @@ -202,20 +203,43 @@ float ProfilerImpl::profile_operator( | |||||
| auto graph = ComputingGraph::make(); | auto graph = ComputingGraph::make(); | ||||
| graph->options().graph_opt_level = 0; | graph->options().graph_opt_level = 0; | ||||
| graph->options().var_sanity_check_first_run = false; | graph->options().var_sanity_check_first_run = false; | ||||
| OperatorNodeBase* new_opr; | |||||
| /// \note: Concat operators are specially treated. The reasons are as | |||||
| /// follows: | |||||
| /// 1. Padding the input varnodes of the | |||||
| /// Concat opr is not allowed. If we pad the input varnodes of Concat opr, the | |||||
| /// padding information of output varnode should be propagated in the layout | |||||
| /// selection algorithm. But this feature has not been implemented now. | |||||
| /// 2. The axis of concat operator should be modified, because the layouts of the | |||||
| /// input varnodes has been modified. So we handle the new axis in the OprMaker | |||||
| /// function. | |||||
| bool allow_aligned = intl::allow_aligned_layout(opr); | |||||
| VarNodeArray new_inps(opr->input().size()); | VarNodeArray new_inps(opr->input().size()); | ||||
| for (size_t i = 0; i < opr->input().size(); ++i) { | for (size_t i = 0; i < opr->input().size(); ++i) { | ||||
| auto&& var = opr->input(i); | auto&& var = opr->input(i); | ||||
| auto&& cn = var->comp_node(); | auto&& cn = var->comp_node(); | ||||
| auto&& dtype = var->dtype(); | auto&& dtype = var->dtype(); | ||||
| auto dval = std::make_shared<DeviceTensorND>(cn, dtype); | auto dval = std::make_shared<DeviceTensorND>(cn, dtype); | ||||
| auto aligned_tensor_shape = ReformatManager::make_aligned_tensor_shape( | |||||
| var, base_format, tensor_format, extra_attribute); | |||||
| dval->resize(aligned_tensor_shape); | |||||
| auto aligned_var = opr::VolatileSharedDeviceTensor::make(*graph, dval); | |||||
| new_inps[i] = aligned_var.node(); | |||||
| auto new_shape = ReformatManager::try_make_tensor_shape( | |||||
| var, base_format, tensor_format, extra_attribute, allow_aligned); | |||||
| if (new_shape.ndim == 0) | |||||
| return PROFILE_TIME_OUT; | |||||
| dval->resize(new_shape); | |||||
| auto new_var = opr::VolatileSharedDeviceTensor::make(*graph, dval); | |||||
| new_inps[i] = new_var.node(); | |||||
| } | |||||
| if (intl::has_opr_format_modifier(opr)) { | |||||
| intl::OprFormatInfo opr_format_info; | |||||
| opr_format_info.tensor_formats = {base_format, tensor_format}; | |||||
| auto new_var = intl::modify_opr_format(opr_format_info, new_inps, opr); | |||||
| if (new_var) | |||||
| new_opr = new_var->owner_opr(); | |||||
| else | |||||
| return PROFILE_TIME_OUT; | |||||
| } else { | |||||
| new_opr = serialization::copy_opr_shallow( | |||||
| *opr, new_inps, opr->config(), {graph.get()}); | |||||
| } | } | ||||
| auto new_opr = serialization::copy_opr_shallow( | |||||
| *opr, new_inps, opr->config(), {graph.get()}); | |||||
| if (!m_opr_filter(opr, new_opr)) | if (!m_opr_filter(opr, new_opr)) | ||||
| return PROFILE_TIME_OUT; | return PROFILE_TIME_OUT; | ||||
| auto y = new_opr->output(0); | auto y = new_opr->output(0); | ||||
| @@ -248,6 +272,8 @@ float ProfilerImpl::profile_operator( | |||||
| ReformatAttribute extra_attribute) const { | ReformatAttribute extra_attribute) const { | ||||
| auto graph = ComputingGraph::make(); | auto graph = ComputingGraph::make(); | ||||
| graph->options().graph_opt_level = 0; | graph->options().graph_opt_level = 0; | ||||
| graph->options().graph_opt.weight_preprocess = | |||||
| opr->owner_graph()->options().graph_opt.weight_preprocess; | |||||
| graph->options().var_sanity_check_first_run = false; | graph->options().var_sanity_check_first_run = false; | ||||
| VarNodeArray new_inps(opr->input().size()); | VarNodeArray new_inps(opr->input().size()); | ||||
| size_t i = 0; | size_t i = 0; | ||||
| @@ -274,8 +300,11 @@ float ProfilerImpl::profile_operator( | |||||
| config.input_tensor_formats[i], extra_attribute); | config.input_tensor_formats[i], extra_attribute); | ||||
| } | } | ||||
| dval->resize(aligned_shape); | dval->resize(aligned_shape); | ||||
| auto aligned_var = opr::VolatileSharedDeviceTensor::make(*graph, dval); | |||||
| new_inps[i] = aligned_var.node(); | |||||
| if (config.input_tensor_types[i] == TensorType::WEIGHT) { | |||||
| new_inps[i] = opr::SharedDeviceTensor::make_const(*graph, dval).node(); | |||||
| } else { | |||||
| new_inps[i] = opr::VolatileSharedDeviceTensor::make(*graph, dval).node(); | |||||
| } | |||||
| } | } | ||||
| for (; i < opr->input().size(); ++i) { | for (; i < opr->input().size(); ++i) { | ||||
| auto&& var = opr->input(i); | auto&& var = opr->input(i); | ||||
| @@ -291,7 +320,9 @@ float ProfilerImpl::profile_operator( | |||||
| auto imm = opr::ImmutableTensor::make(*graph, *hval); | auto imm = opr::ImmutableTensor::make(*graph, *hval); | ||||
| new_inps[i] = imm.node(); | new_inps[i] = imm.node(); | ||||
| } | } | ||||
| VarNode* y = mgb::gopt::intl::modify_opr_format(config.opr_format, new_inps, opr); | |||||
| intl::OprFormatInfo opr_format_info; | |||||
| opr_format_info.opr_format = config.opr_format; | |||||
| VarNode* y = mgb::gopt::intl::modify_opr_format(opr_format_info, new_inps, opr); | |||||
| static const ThinHashSet<Typeinfo*> multi_algo_oprs = { | static const ThinHashSet<Typeinfo*> multi_algo_oprs = { | ||||
| opr::Convolution::typeinfo(), | opr::Convolution::typeinfo(), | ||||
| opr::ConvBiasForward::typeinfo(), | opr::ConvBiasForward::typeinfo(), | ||||
+ 22
- 4
src/gopt/impl/global_layout_transform/reformat_manager.cpp
View File
| @@ -587,9 +587,9 @@ const ReformatManager& ReformatManager::instance() { | |||||
| return inst; | return inst; | ||||
| } | } | ||||
| TensorShape ReformatManager::make_aligned_tensor_shape( | |||||
| TensorShape ReformatManager::try_make_tensor_shape( | |||||
| const VarNode* var, TensorFormats orig_formats, TensorFormats target_formats, | const VarNode* var, TensorFormats orig_formats, TensorFormats target_formats, | ||||
| ReformatKey::Attribute extra_attribute) { | |||||
| ReformatKey::Attribute extra_attribute, bool allow_aligned) { | |||||
| using Dimension = megdnn::Dimension; | using Dimension = megdnn::Dimension; | ||||
| static constexpr uint32_t UNDETERMINED_EXTENT = Dimension::UNDETERMINED_EXTENT; | static constexpr uint32_t UNDETERMINED_EXTENT = Dimension::UNDETERMINED_EXTENT; | ||||
| auto orig_shape = tensor_formats_to_named_tensor_shape(orig_formats); | auto orig_shape = tensor_formats_to_named_tensor_shape(orig_formats); | ||||
| @@ -623,8 +623,17 @@ TensorShape ReformatManager::make_aligned_tensor_shape( | |||||
| : (orig_shape[idx] / target_shape[i]).extent(); | : (orig_shape[idx] / target_shape[i]).extent(); | ||||
| if (mul) | if (mul) | ||||
| tshp[i] = oshp[idx] * factor; | tshp[i] = oshp[idx] * factor; | ||||
| else | |||||
| tshp[i] = divup(oshp[idx], factor); | |||||
| else { | |||||
| if (allow_aligned) | |||||
| tshp[i] = divup(oshp[idx], factor); | |||||
| else if (!(oshp[idx] % factor)) { | |||||
| tshp[i] = oshp[idx] / factor; | |||||
| } else { | |||||
| return TensorShape{}; | |||||
| } | |||||
| } | |||||
| /// hack for nhwc auto padding | |||||
| if (name == Dimension::Name::C) { | if (name == Dimension::Name::C) { | ||||
| size_t channel_alignment = target_shape[i].stride(); | size_t channel_alignment = target_shape[i].stride(); | ||||
| size_t channels = tshp[i] * channel_alignment; | size_t channels = tshp[i] * channel_alignment; | ||||
| @@ -641,6 +650,15 @@ TensorShape ReformatManager::make_aligned_tensor_shape( | |||||
| return tshp; | return tshp; | ||||
| } | } | ||||
| TensorShape ReformatManager::make_aligned_tensor_shape( | |||||
| const VarNode* var, TensorFormats orig_formats, TensorFormats target_formats, | |||||
| ReformatKey::Attribute extra_attribute) { | |||||
| auto tshp = ReformatManager::try_make_tensor_shape( | |||||
| var, orig_formats, target_formats, extra_attribute); | |||||
| mgb_assert(tshp.ndim); | |||||
| return tshp; | |||||
| } | |||||
| TensorShape ReformatManager::make_aligned_weight_shape( | TensorShape ReformatManager::make_aligned_weight_shape( | ||||
| const VarNode* var, TensorFormats orig_formats, TensorFormats target_formats, | const VarNode* var, TensorFormats orig_formats, TensorFormats target_formats, | ||||
| TensorFormats extra_formats, ReformatKey::Attribute extra_attribute) { | TensorFormats extra_formats, ReformatKey::Attribute extra_attribute) { | ||||
+ 2
- 1
src/gopt/impl/global_layout_transform/utils.h
View File
| @@ -93,6 +93,8 @@ static inline TensorFormats opr_format_to_tensor_formats( | |||||
| return TensorFormats::NCHWc4; | return TensorFormats::NCHWc4; | ||||
| case OprFormat::NCHW8: | case OprFormat::NCHW8: | ||||
| return TensorFormats::NCHWc8; | return TensorFormats::NCHWc8; | ||||
| case OprFormat::NCHW44_DOT: | |||||
| return TensorFormats::NCHWc4; | |||||
| default: | default: | ||||
| mgb_throw( | mgb_throw( | ||||
| AssertionError, "format(%s) is not supported", | AssertionError, "format(%s) is not supported", | ||||
| @@ -171,7 +173,6 @@ static inline megdnn::NamedTensorShape tensor_formats_to_named_tensor_shape( | |||||
| static_cast<uint32_t>(format)); | static_cast<uint32_t>(format)); | ||||
| } | } | ||||
| } | } | ||||
| } // namespace gopt | } // namespace gopt | ||||
| } // namespace mgb | } // namespace mgb | ||||
+ 7
- 0
src/gopt/include/megbrain/gopt/reformat_manager.h
View File
| @@ -135,6 +135,13 @@ public: | |||||
| const VarNode* orig_var, const ReformatKey& key, | const VarNode* orig_var, const ReformatKey& key, | ||||
| const AlignmentDesc& extra_alignment = {}) const; | const AlignmentDesc& extra_alignment = {}) const; | ||||
| /// return empty shape, if shape of origin varnode does not satisfy the alignment | |||||
| /// requirement of the target tensor formats | |||||
| static TensorShape try_make_tensor_shape( | |||||
| const VarNode* var, TensorFormats orig_formats, | |||||
| TensorFormats target_formats, | |||||
| ReformatKey::Attribute extra_attribute = ReformatKey::Attribute::DEFAULT, | |||||
| bool allow_aligned = true); | |||||
| static TensorShape make_aligned_tensor_shape( | static TensorShape make_aligned_tensor_shape( | ||||
| const VarNode* var, TensorFormats orig_formats, | const VarNode* var, TensorFormats orig_formats, | ||||
| TensorFormats target_formats, | TensorFormats target_formats, | ||||
+ 214
- 0
src/gopt/test/cache_data.h
View File
| @@ -18104,3 +18104,217 @@ static const std::vector<uint8_t> TestLayoutTransform_Wide = { | |||||
| 108, 111, 97, 116, 51, 50, 124, 49, 54, 44, 49, 54, 44, 56, 44, 56, | 108, 111, 97, 116, 51, 50, 124, 49, 54, 44, 49, 54, 44, 56, 44, 56, | ||||
| 59, 70, 108, 111, 97, 116, 51, 50, 124, 49, 48, 4, 0, 0, 0, 233, | 59, 70, 108, 111, 97, 116, 51, 50, 124, 49, 48, 4, 0, 0, 0, 233, | ||||
| 38, 177, 64}; | 38, 177, 64}; | ||||
| static const std::vector<uint8_t> TestLayoutTransform_Concat = { | |||||
| 1, 0, 0, 0, 34, 0, 0, 0, 108, 97, 121, 111, 117, 116, 95, 116, | |||||
| 114, 97, 110, 115, 102, 111, 114, 109, 95, 112, 114, 111, 102, 105, 108, 101, | |||||
| 58, 112, 108, 97, 116, 61, 99, 117, 100, 97, 33, 0, 0, 0, 77, 0, | |||||
| 0, 0, 49, 54, 44, 50, 44, 51, 54, 56, 44, 54, 52, 48, 59, 81, | |||||
| 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 123, 78, 44, 67, 47, | |||||
| 47, 52, 44, 72, 44, 87, 44, 67, 37, 52, 125, 59, 123, 78, 44, 67, | |||||
| 44, 72, 44, 87, 125, 59, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, | |||||
| 100, 83, 56, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 4, | |||||
| 0, 0, 0, 28, 90, 20, 67, 89, 0, 0, 0, 47, 225, 79, 220, 129, | |||||
| 196, 236, 15, 1, 0, 0, 0, 49, 54, 44, 50, 44, 51, 54, 56, 44, | |||||
| 54, 52, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, | |||||
| 49, 54, 44, 50, 44, 51, 54, 56, 44, 54, 52, 48, 59, 81, 117, 97, | |||||
| 110, 116, 105, 122, 101, 100, 83, 56, 124, 49, 54, 44, 52, 44, 51, 54, | |||||
| 56, 44, 54, 52, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, | |||||
| 56, 124, 51, 48, 4, 0, 0, 0, 128, 150, 24, 75, 77, 0, 0, 0, | |||||
| 49, 54, 44, 52, 44, 49, 56, 52, 44, 51, 50, 48, 59, 81, 117, 97, | |||||
| 110, 116, 105, 122, 101, 100, 83, 56, 124, 123, 78, 44, 67, 47, 47, 52, | |||||
| 44, 72, 44, 87, 44, 67, 37, 52, 125, 59, 123, 78, 44, 67, 44, 72, | |||||
| 44, 87, 125, 59, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, | |||||
| 56, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 4, 0, 0, | |||||
| 0, 236, 81, 208, 66, 52, 0, 0, 0, 145, 22, 69, 252, 191, 174, 0, | |||||
| 1, 49, 54, 44, 51, 44, 55, 51, 54, 44, 49, 50, 56, 48, 59, 85, | |||||
| 105, 110, 116, 56, 124, 49, 54, 44, 51, 44, 55, 51, 54, 44, 49, 50, | |||||
| 56, 48, 59, 70, 108, 111, 97, 116, 51, 50, 124, 48, 48, 4, 0, 0, | |||||
| 0, 88, 57, 79, 68, 66, 0, 0, 0, 49, 54, 44, 51, 44, 55, 51, | |||||
| 54, 44, 49, 50, 56, 48, 59, 70, 108, 111, 97, 116, 51, 50, 124, 123, | |||||
| 78, 44, 67, 44, 72, 44, 87, 125, 59, 123, 78, 44, 67, 47, 47, 52, | |||||
| 44, 72, 44, 87, 44, 67, 37, 52, 125, 59, 48, 59, 70, 108, 111, 97, | |||||
| 116, 51, 50, 59, 70, 108, 111, 97, 116, 51, 50, 4, 0, 0, 0, 134, | |||||
| 75, 130, 69, 163, 0, 0, 0, 10, 151, 159, 190, 73, 14, 145, 194, 1, | |||||
| 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, | |||||
| 0, 0, 0, 1, 0, 0, 0, 2, 0, 0, 0, 2, 0, 0, 0, 1, | |||||
| 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, | |||||
| 0, 0, 0, 255, 255, 255, 255, 255, 255, 255, 255, 49, 54, 44, 56, 44, | |||||
| 49, 56, 52, 44, 51, 50, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, | |||||
| 100, 83, 56, 124, 49, 54, 44, 56, 44, 51, 44, 51, 59, 81, 117, 97, | |||||
| 110, 116, 105, 122, 101, 100, 83, 56, 124, 49, 44, 49, 54, 44, 49, 44, | |||||
| 49, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 51, 50, 124, 49, | |||||
| 54, 44, 49, 54, 44, 57, 50, 44, 49, 54, 48, 59, 81, 117, 97, 110, | |||||
| 116, 105, 122, 101, 100, 83, 56, 124, 51, 48, 4, 0, 0, 0, 68, 139, | |||||
| 104, 67, 60, 0, 0, 0, 49, 54, 44, 51, 44, 55, 51, 54, 44, 49, | |||||
| 50, 56, 48, 59, 85, 105, 110, 116, 56, 124, 123, 78, 44, 67, 47, 47, | |||||
| 52, 44, 72, 44, 87, 44, 67, 37, 52, 125, 59, 123, 78, 44, 67, 44, | |||||
| 72, 44, 87, 125, 59, 48, 59, 85, 105, 110, 116, 56, 59, 85, 105, 110, | |||||
| 116, 56, 4, 0, 0, 0, 201, 161, 135, 68, 64, 0, 0, 0, 218, 182, | |||||
| 120, 146, 221, 155, 179, 235, 11, 0, 0, 0, 49, 54, 44, 52, 44, 49, | |||||
| 56, 52, 44, 51, 50, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, | |||||
| 83, 56, 124, 49, 54, 44, 52, 44, 49, 56, 52, 44, 51, 50, 48, 59, | |||||
| 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 48, 48, 4, 0, | |||||
| 0, 0, 182, 243, 29, 66, 161, 0, 0, 0, 10, 151, 159, 190, 73, 14, | |||||
| 145, 194, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, | |||||
| 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 2, 0, 0, 0, 2, 0, | |||||
| 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, | |||||
| 0, 0, 0, 0, 0, 0, 255, 255, 255, 255, 255, 255, 255, 255, 49, 54, | |||||
| 44, 52, 44, 51, 54, 56, 44, 54, 52, 48, 59, 81, 117, 97, 110, 116, | |||||
| 105, 122, 101, 100, 83, 56, 124, 52, 44, 52, 44, 51, 44, 51, 59, 81, | |||||
| 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 49, 44, 52, 44, 49, | |||||
| 44, 49, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 51, 50, 124, | |||||
| 49, 54, 44, 52, 44, 49, 56, 52, 44, 51, 50, 48, 59, 81, 117, 97, | |||||
| 110, 116, 105, 122, 101, 100, 83, 56, 124, 48, 48, 4, 0, 0, 0, 128, | |||||
| 150, 24, 75, 60, 0, 0, 0, 49, 54, 44, 51, 44, 55, 51, 54, 44, | |||||
| 49, 50, 56, 48, 59, 85, 105, 110, 116, 56, 124, 123, 78, 44, 67, 44, | |||||
| 72, 44, 87, 125, 59, 123, 78, 44, 67, 47, 47, 52, 44, 72, 44, 87, | |||||
| 44, 67, 37, 52, 125, 59, 48, 59, 85, 105, 110, 116, 56, 59, 85, 105, | |||||
| 110, 116, 56, 4, 0, 0, 0, 203, 33, 219, 68, 68, 0, 0, 0, 218, | |||||
| 182, 120, 146, 221, 155, 179, 235, 16, 0, 0, 0, 49, 54, 44, 51, 44, | |||||
| 55, 51, 54, 44, 49, 50, 56, 48, 59, 70, 108, 111, 97, 116, 51, 50, | |||||
| 124, 49, 59, 70, 108, 111, 97, 116, 51, 50, 124, 49, 54, 44, 51, 44, | |||||
| 55, 51, 54, 44, 49, 50, 56, 48, 59, 70, 108, 111, 97, 116, 51, 50, | |||||
| 124, 48, 48, 4, 0, 0, 0, 205, 204, 161, 68, 77, 0, 0, 0, 49, | |||||
| 54, 44, 52, 44, 51, 54, 56, 44, 54, 52, 48, 59, 81, 117, 97, 110, | |||||
| 116, 105, 122, 101, 100, 83, 56, 124, 123, 78, 44, 67, 47, 47, 52, 44, | |||||
| 72, 44, 87, 44, 67, 37, 52, 125, 59, 123, 78, 44, 67, 44, 72, 44, | |||||
| 87, 125, 59, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, | |||||
| 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 4, 0, 0, 0, | |||||
| 200, 118, 174, 67, 163, 0, 0, 0, 10, 151, 159, 190, 73, 14, 145, 194, | |||||
| 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, | |||||
| 1, 0, 0, 0, 1, 0, 0, 0, 2, 0, 0, 0, 2, 0, 0, 0, | |||||
| 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, | |||||
| 0, 0, 0, 0, 255, 255, 255, 255, 255, 255, 255, 255, 49, 54, 44, 56, | |||||
| 44, 49, 56, 52, 44, 51, 50, 48, 59, 81, 117, 97, 110, 116, 105, 122, | |||||
| 101, 100, 83, 56, 124, 49, 54, 44, 56, 44, 51, 44, 51, 59, 81, 117, | |||||
| 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 49, 44, 49, 54, 44, 49, | |||||
| 44, 49, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 51, 50, 124, | |||||
| 49, 54, 44, 49, 54, 44, 57, 50, 44, 49, 54, 48, 59, 81, 117, 97, | |||||
| 110, 116, 105, 122, 101, 100, 83, 56, 124, 48, 48, 4, 0, 0, 0, 128, | |||||
| 150, 24, 75, 161, 0, 0, 0, 10, 151, 159, 190, 73, 14, 145, 194, 1, | |||||
| 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, | |||||
| 0, 0, 0, 1, 0, 0, 0, 2, 0, 0, 0, 2, 0, 0, 0, 1, | |||||
| 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, | |||||
| 0, 0, 0, 255, 255, 255, 255, 255, 255, 255, 255, 49, 54, 44, 52, 44, | |||||
| 51, 54, 56, 44, 54, 52, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, | |||||
| 100, 83, 56, 124, 52, 44, 52, 44, 51, 44, 51, 59, 81, 117, 97, 110, | |||||
| 116, 105, 122, 101, 100, 83, 56, 124, 49, 44, 52, 44, 49, 44, 49, 59, | |||||
| 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 51, 50, 124, 49, 54, 44, | |||||
| 52, 44, 49, 56, 52, 44, 51, 50, 48, 59, 81, 117, 97, 110, 116, 105, | |||||
| 122, 101, 100, 83, 56, 124, 51, 48, 4, 0, 0, 0, 138, 65, 164, 67, | |||||
| 77, 0, 0, 0, 49, 54, 44, 50, 44, 51, 54, 56, 44, 54, 52, 48, | |||||
| 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 123, 78, 44, | |||||
| 67, 44, 72, 44, 87, 125, 59, 123, 78, 44, 67, 47, 47, 52, 44, 72, | |||||
| 44, 87, 44, 67, 37, 52, 125, 59, 48, 59, 81, 117, 97, 110, 116, 105, | |||||
| 122, 101, 100, 83, 56, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, | |||||
| 56, 4, 0, 0, 0, 137, 65, 190, 67, 58, 0, 0, 0, 145, 22, 69, | |||||
| 252, 191, 174, 0, 1, 49, 54, 44, 51, 44, 55, 51, 54, 44, 49, 50, | |||||
| 56, 48, 59, 70, 108, 111, 97, 116, 51, 50, 124, 49, 54, 44, 51, 44, | |||||
| 55, 51, 54, 44, 49, 50, 56, 48, 59, 81, 117, 97, 110, 116, 105, 122, | |||||
| 101, 100, 83, 56, 124, 48, 48, 4, 0, 0, 0, 38, 177, 146, 68, 89, | |||||
| 0, 0, 0, 47, 225, 79, 220, 129, 196, 236, 15, 1, 0, 0, 0, 49, | |||||
| 54, 44, 50, 44, 51, 54, 56, 44, 54, 52, 48, 59, 81, 117, 97, 110, | |||||
| 116, 105, 122, 101, 100, 83, 56, 124, 49, 54, 44, 50, 44, 51, 54, 56, | |||||
| 44, 54, 52, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, | |||||
| 124, 49, 54, 44, 52, 44, 51, 54, 56, 44, 54, 52, 48, 59, 81, 117, | |||||
| 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 48, 48, 4, 0, 0, 0, | |||||
| 150, 67, 235, 66, 162, 0, 0, 0, 10, 151, 159, 190, 73, 14, 145, 194, | |||||
| 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, | |||||
| 1, 0, 0, 0, 1, 0, 0, 0, 2, 0, 0, 0, 2, 0, 0, 0, | |||||
| 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, | |||||
| 0, 0, 0, 0, 255, 255, 255, 255, 255, 255, 255, 255, 49, 54, 44, 51, | |||||
| 44, 55, 51, 54, 44, 49, 50, 56, 48, 59, 81, 117, 97, 110, 116, 105, | |||||
| 122, 101, 100, 83, 56, 124, 50, 44, 51, 44, 51, 44, 51, 59, 81, 117, | |||||
| 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 49, 44, 50, 44, 49, 44, | |||||
| 49, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 51, 50, 124, 49, | |||||
| 54, 44, 50, 44, 51, 54, 56, 44, 54, 52, 48, 59, 81, 117, 97, 110, | |||||
| 116, 105, 122, 101, 100, 83, 56, 124, 51, 48, 4, 0, 0, 0, 128, 150, | |||||
| 24, 75, 64, 0, 0, 0, 218, 182, 120, 146, 221, 155, 179, 235, 11, 0, | |||||
| 0, 0, 49, 54, 44, 50, 44, 51, 54, 56, 44, 54, 52, 48, 59, 81, | |||||
| 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 49, 54, 44, 50, 44, | |||||
| 51, 54, 56, 44, 54, 52, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, | |||||
| 100, 83, 56, 124, 51, 48, 4, 0, 0, 0, 227, 165, 243, 66, 77, 0, | |||||
| 0, 0, 49, 54, 44, 52, 44, 49, 56, 52, 44, 51, 50, 48, 59, 81, | |||||
| 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 123, 78, 44, 67, 44, | |||||
| 72, 44, 87, 125, 59, 123, 78, 44, 67, 47, 47, 52, 44, 72, 44, 87, | |||||
| 44, 67, 37, 52, 125, 59, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, | |||||
| 100, 83, 56, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 4, | |||||
| 0, 0, 0, 174, 71, 169, 66, 162, 0, 0, 0, 10, 151, 159, 190, 73, | |||||
| 14, 145, 194, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, | |||||
| 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 2, 0, 0, 0, 2, | |||||
| 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, | |||||
| 0, 0, 0, 0, 0, 0, 0, 255, 255, 255, 255, 255, 255, 255, 255, 49, | |||||
| 54, 44, 51, 44, 55, 51, 54, 44, 49, 50, 56, 48, 59, 81, 117, 97, | |||||
| 110, 116, 105, 122, 101, 100, 83, 56, 124, 50, 44, 51, 44, 51, 44, 51, | |||||
| 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 49, 44, 50, | |||||
| 44, 49, 44, 49, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 51, | |||||
| 50, 124, 49, 54, 44, 50, 44, 51, 54, 56, 44, 54, 52, 48, 59, 81, | |||||
| 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 48, 48, 4, 0, 0, | |||||
| 0, 128, 150, 24, 75, 66, 0, 0, 0, 49, 54, 44, 51, 44, 55, 51, | |||||
| 54, 44, 49, 50, 56, 48, 59, 70, 108, 111, 97, 116, 51, 50, 124, 123, | |||||
| 78, 44, 67, 47, 47, 52, 44, 72, 44, 87, 44, 67, 37, 52, 125, 59, | |||||
| 123, 78, 44, 67, 44, 72, 44, 87, 125, 59, 48, 59, 70, 108, 111, 97, | |||||
| 116, 51, 50, 59, 70, 108, 111, 97, 116, 51, 50, 4, 0, 0, 0, 196, | |||||
| 32, 63, 69, 89, 0, 0, 0, 47, 225, 79, 220, 129, 196, 236, 15, 1, | |||||
| 0, 0, 0, 49, 54, 44, 52, 44, 49, 56, 52, 44, 51, 50, 48, 59, | |||||
| 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 49, 54, 44, 52, | |||||
| 44, 49, 56, 52, 44, 51, 50, 48, 59, 81, 117, 97, 110, 116, 105, 122, | |||||
| 101, 100, 83, 56, 124, 49, 54, 44, 56, 44, 49, 56, 52, 44, 51, 50, | |||||
| 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 48, 48, | |||||
| 4, 0, 0, 0, 156, 196, 136, 66, 77, 0, 0, 0, 49, 54, 44, 49, | |||||
| 54, 44, 57, 50, 44, 49, 54, 48, 59, 81, 117, 97, 110, 116, 105, 122, | |||||
| 101, 100, 83, 56, 124, 123, 78, 44, 67, 47, 47, 52, 44, 72, 44, 87, | |||||
| 44, 67, 37, 52, 125, 59, 123, 78, 44, 67, 44, 72, 44, 87, 125, 59, | |||||
| 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 59, 81, 117, | |||||
| 97, 110, 116, 105, 122, 101, 100, 83, 56, 4, 0, 0, 0, 207, 247, 211, | |||||
| 66, 77, 0, 0, 0, 49, 54, 44, 56, 44, 49, 56, 52, 44, 51, 50, | |||||
| 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 123, 78, | |||||
| 44, 67, 47, 47, 52, 44, 72, 44, 87, 44, 67, 37, 52, 125, 59, 123, | |||||
| 78, 44, 67, 44, 72, 44, 87, 125, 59, 48, 59, 81, 117, 97, 110, 116, | |||||
| 105, 122, 101, 100, 83, 56, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, | |||||
| 83, 56, 4, 0, 0, 0, 54, 94, 58, 67, 64, 0, 0, 0, 218, 182, | |||||
| 120, 146, 221, 155, 179, 235, 11, 0, 0, 0, 49, 54, 44, 50, 44, 51, | |||||
| 54, 56, 44, 54, 52, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, | |||||
| 83, 56, 124, 49, 54, 44, 50, 44, 51, 54, 56, 44, 54, 52, 48, 59, | |||||
| 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 48, 48, 4, 0, | |||||
| 0, 0, 155, 196, 128, 66, 78, 0, 0, 0, 49, 54, 44, 51, 44, 55, | |||||
| 51, 54, 44, 49, 50, 56, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, | |||||
| 100, 83, 56, 124, 123, 78, 44, 67, 44, 72, 44, 87, 125, 59, 123, 78, | |||||
| 44, 67, 47, 47, 52, 44, 72, 44, 87, 44, 67, 37, 52, 125, 59, 48, | |||||
| 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 59, 81, 117, 97, | |||||
| 110, 116, 105, 122, 101, 100, 83, 56, 4, 0, 0, 0, 191, 74, 159, 68, | |||||
| 89, 0, 0, 0, 47, 225, 79, 220, 129, 196, 236, 15, 1, 0, 0, 0, | |||||
| 49, 54, 44, 52, 44, 49, 56, 52, 44, 51, 50, 48, 59, 81, 117, 97, | |||||
| 110, 116, 105, 122, 101, 100, 83, 56, 124, 49, 54, 44, 52, 44, 49, 56, | |||||
| 52, 44, 51, 50, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, | |||||
| 56, 124, 49, 54, 44, 56, 44, 49, 56, 52, 44, 51, 50, 48, 59, 81, | |||||
| 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 51, 48, 4, 0, 0, | |||||
| 0, 170, 241, 138, 66, 77, 0, 0, 0, 49, 54, 44, 52, 44, 51, 54, | |||||
| 56, 44, 54, 52, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, | |||||
| 56, 124, 123, 78, 44, 67, 44, 72, 44, 87, 125, 59, 123, 78, 44, 67, | |||||
| 47, 47, 52, 44, 72, 44, 87, 44, 67, 37, 52, 125, 59, 48, 59, 81, | |||||
| 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 59, 81, 117, 97, 110, 116, | |||||
| 105, 122, 101, 100, 83, 56, 4, 0, 0, 0, 79, 141, 147, 67, 77, 0, | |||||
| 0, 0, 49, 54, 44, 49, 54, 44, 57, 50, 44, 49, 54, 48, 59, 81, | |||||
| 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 123, 78, 44, 67, 44, | |||||
| 72, 44, 87, 125, 59, 123, 78, 44, 67, 47, 47, 52, 44, 72, 44, 87, | |||||
| 44, 67, 37, 52, 125, 59, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, | |||||
| 100, 83, 56, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 4, | |||||
| 0, 0, 0, 170, 241, 170, 66, 77, 0, 0, 0, 49, 54, 44, 56, 44, | |||||
| 49, 56, 52, 44, 51, 50, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, | |||||
| 100, 83, 56, 124, 123, 78, 44, 67, 44, 72, 44, 87, 125, 59, 123, 78, | |||||
| 44, 67, 47, 47, 52, 44, 72, 44, 87, 44, 67, 37, 52, 125, 59, 48, | |||||
| 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 59, 81, 117, 97, | |||||
| 110, 116, 105, 122, 101, 100, 83, 56, 4, 0, 0, 0, 203, 161, 25, 67, | |||||
| 64, 0, 0, 0, 218, 182, 120, 146, 221, 155, 179, 235, 11, 0, 0, 0, | |||||
| 49, 54, 44, 52, 44, 49, 56, 52, 44, 51, 50, 48, 59, 81, 117, 97, | |||||
| 110, 116, 105, 122, 101, 100, 83, 56, 124, 49, 54, 44, 52, 44, 49, 56, | |||||
| 52, 44, 51, 50, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, | |||||
| 56, 124, 51, 48, 4, 0, 0, 0, 102, 102, 22, 66, 78, 0, 0, 0, | |||||
| 49, 54, 44, 51, 44, 55, 51, 54, 44, 49, 50, 56, 48, 59, 81, 117, | |||||
| 97, 110, 116, 105, 122, 101, 100, 83, 56, 124, 123, 78, 44, 67, 47, 47, | |||||
| 52, 44, 72, 44, 87, 44, 67, 37, 52, 125, 59, 123, 78, 44, 67, 44, | |||||
| 72, 44, 87, 125, 59, 48, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, | |||||
| 83, 56, 59, 81, 117, 97, 110, 116, 105, 122, 101, 100, 83, 56, 4, 0, | |||||
| 0, 0, 54, 94, 72, 68}; | |||||
+ 21
- 7
src/gopt/test/embed_cache.py
View File
| @@ -20,7 +20,7 @@ | |||||
| # 2. 编译megbrain_test，并运行所有全局图优化相关测试： | # 2. 编译megbrain_test，并运行所有全局图优化相关测试： | ||||
| # ./megbrain_test --gtest_filter="*LayoutTransform*" | # ./megbrain_test --gtest_filter="*LayoutTransform*" | ||||
| # 3. 用这个脚本把所有的cache文件打包在一起 | # 3. 用这个脚本把所有的cache文件打包在一起 | ||||
| # python3 embed_cache.py -o cache_data.h -r $(ls /path/to/cache/*.cache) | |||||
| # python3 embed_cache.py -o cache_data.h -r -a $(ls /path/to/cache/*.cache) | |||||
| # 4. 将步骤1中的 define 语句改回原样，这样 profile 过程就会使用 cache 下来的数据。 | # 4. 将步骤1中的 define 语句改回原样，这样 profile 过程就会使用 cache 下来的数据。 | ||||
| # 5. 最后可以重新构建一下 megbrain_test ，确保测试结果正确。 | # 5. 最后可以重新构建一下 megbrain_test ，确保测试结果正确。 | ||||
| import os.path | import os.path | ||||
| @@ -44,9 +44,10 @@ def _u32(data): | |||||
| class CacheDataGenerator: | class CacheDataGenerator: | ||||
| _cache_files = None | _cache_files = None | ||||
| def __init__(self, cache_files, remove_plat_info = True): | |||||
| def __init__(self, cache_files, remove_plat_info=True, append_cache=True): | |||||
| self._cache_files = cache_files | self._cache_files = cache_files | ||||
| self._remove_plat_info = remove_plat_info | self._remove_plat_info = remove_plat_info | ||||
| self._append_cache = append_cache | |||||
| def _get_hash(self): | def _get_hash(self): | ||||
| return _u32(self._hash.digest()[:4]) | return _u32(self._hash.digest()[:4]) | ||||
| @@ -71,9 +72,10 @@ class CacheDataGenerator: | |||||
| return ','.join(ret) | return ','.join(ret) | ||||
| def gen_cache_data_header(self, fout, src_map): | def gen_cache_data_header(self, fout, src_map): | ||||
| fout.write('// generated embed_cache.py\n') | |||||
| fout.write('#include <vector>\n') | |||||
| fout.write('#include <stdint.h>\n') | |||||
| if not self._append_cache: | |||||
| fout.write('// generated embed_cache.py\n') | |||||
| fout.write('#include <vector>\n') | |||||
| fout.write('#include <stdint.h>\n') | |||||
| for k, v in sorted(src_map.items()): | for k, v in sorted(src_map.items()): | ||||
| fout.write(""" | fout.write(""" | ||||
| static const std::vector<uint8_t> {} = {{ | static const std::vector<uint8_t> {} = {{ | ||||
| @@ -89,7 +91,11 @@ static const std::vector<uint8_t> {} = {{ | |||||
| assert ext == ".cache", "ext: {}, fname {}".format(ext, fname) | assert ext == ".cache", "ext: {}, fname {}".format(ext, fname) | ||||
| assert base not in fname2cache_data, "duplicated kernel: " + base | assert base not in fname2cache_data, "duplicated kernel: " + base | ||||
| fname2cache_data[base] = self.gen_cache_data(fname) | fname2cache_data[base] = self.gen_cache_data(fname) | ||||
| with open(output, 'w') as fout: | |||||
| if self._append_cache: | |||||
| mode = 'a' | |||||
| else: | |||||
| mode = 'w' | |||||
| with open(output, mode) as fout: | |||||
| self.gen_cache_data_header(fout, fname2cache_data) | self.gen_cache_data_header(fout, fname2cache_data) | ||||
| logger.info('done') | logger.info('done') | ||||
| @@ -107,7 +113,15 @@ if __name__ == '__main__': | |||||
| default=True, | default=True, | ||||
| help="whether remove platform infomation in the cache (default: True)" | help="whether remove platform infomation in the cache (default: True)" | ||||
| ) | ) | ||||
| parser.add_argument( | |||||
| "-a", | |||||
| "--append-cache", | |||||
| action='store_true', | |||||
| default=True, | |||||
| help="whether append the cache (default: True)" | |||||
| ) | |||||
| parser.add_argument('cache', help='cache files to be embedded', nargs='+') | parser.add_argument('cache', help='cache files to be embedded', nargs='+') | ||||
| args = parser.parse_args() | args = parser.parse_args() | ||||
| cache_generator = CacheDataGenerator(args.cache, args.remove_plat_info) | |||||
| cache_generator = CacheDataGenerator(args.cache, args.remove_plat_info, | |||||
| args.append_cache) | |||||
| cache_generator.invoke(args.output) | cache_generator.invoke(args.output) | ||||
+ 110
- 0
src/gopt/test/layout_transform_pass.cpp
View File
| @@ -812,7 +812,9 @@ TEST(TestLayoutTransform, Resnet18_F16) { | |||||
| .add_pass<LayoutTransformPass>(std::move(ctx), std::move(solver)) | .add_pass<LayoutTransformPass>(std::move(ctx), std::move(solver)) | ||||
| .add_pass<ShuffleShuffleRemovePass>() | .add_pass<ShuffleShuffleRemovePass>() | ||||
| .add_pass(FuseNCHW4Int8Preprocess::make()) | .add_pass(FuseNCHW4Int8Preprocess::make()) | ||||
| #if CUDA_VERSION >= 10020 | |||||
| .add_pass<FoldingConvBiasDimshufflePass>() | .add_pass<FoldingConvBiasDimshufflePass>() | ||||
| #endif | |||||
| .add_pass<ParamFusePass>() | .add_pass<ParamFusePass>() | ||||
| .add_pass<ParamMergePass>() | .add_pass<ParamMergePass>() | ||||
| .apply({{output}}) | .apply({{output}}) | ||||
| @@ -1205,4 +1207,112 @@ TEST(TestLayoutTransform, MobileNetV2_NCHW44_DOT) { | |||||
| MGB_ASSERT_TENSOR_EQ(t1, t2); | MGB_ASSERT_TENSOR_EQ(t1, t2); | ||||
| } | } | ||||
| #if MGB_CUDA | |||||
| TEST(TestLayoutTransform, Concat) { | |||||
| REQUIRE_GPU(1); | |||||
| auto cn = CompNode::load("gpu0"); | |||||
| cn.activate(); | |||||
| REQUIRE_CUDA_COMPUTE_CAPABILITY(6, 1); | |||||
| constexpr size_t N = 16, C = 3, H = 736, W = 1280; | |||||
| HostTensorGenerator<dtype::Uint8> gen; | |||||
| auto graph = ComputingGraph::make(); | |||||
| auto h2d = opr::Host2DeviceCopy::make(*graph, gen({N, C, H, W}, cn)); | |||||
| auto data = opr::TypeCvt::make(h2d, dtype::Float32()); | |||||
| auto sub_128 = data + (-128); | |||||
| auto x = opr::TypeCvt::make(sub_128, dtype::QuantizedS8(1.f)); | |||||
| auto mkcvar = [&](const char* name, const TensorShape& shp, const DType& dtype) { | |||||
| return opr::TypeCvt::make( | |||||
| opr::SharedDeviceTensor::make(*graph, *gen(shp, cn)).rename(name), | |||||
| dtype); | |||||
| }; | |||||
| auto w = mkcvar("w", {2, 3, 3, 3}, dtype::QuantizedS8(1.f)); | |||||
| auto b = mkcvar("b", {1, 2, 1, 1}, dtype::QuantizedS32(1.f)); | |||||
| opr::ConvBias::Param param; | |||||
| param.format = opr::ConvBias::Param::Format::NCHW; | |||||
| param.nonlineMode = opr::ConvBias::Param::NonlineMode::RELU; | |||||
| param.stride_h = param.stride_w = 2; | |||||
| param.pad_h = param.pad_w = 1; | |||||
| auto conv_1 = opr::ConvBias::make( | |||||
| x, w, b, param, {}, OperatorNodeConfig(dtype::QuantizedS8(1.f))); | |||||
| auto conv_1_cat = opr::Concat::make({conv_1, -conv_1}, 1); | |||||
| auto w2 = mkcvar("w", {4, 4, 3, 3}, dtype::QuantizedS8(1.f)); | |||||
| auto b2 = mkcvar("b", {1, 4, 1, 1}, dtype::QuantizedS32(1.f)); | |||||
| auto conv_2 = opr::ConvBias::make( | |||||
| conv_1_cat, w2, b2, param, {}, OperatorNodeConfig(dtype::QuantizedS8(1.f))); | |||||
| auto conv_2_cat = opr::Concat::make({conv_2, -conv_2}, 1); | |||||
| auto w3 = mkcvar("w", {16, 8, 3, 3}, dtype::QuantizedS8(1.f)); | |||||
| auto b3 = mkcvar("b", {1, 16, 1, 1}, dtype::QuantizedS32(1.f)); | |||||
| auto y = opr::ConvBias::make( | |||||
| conv_2_cat, w3, b3, param, {}, OperatorNodeConfig(dtype::QuantizedS8(1.f))); | |||||
| using S = opr::mixin::AlgoChooserHelper::ExecutionPolicy::Strategy; | |||||
| S strategy = S::PROFILE; | |||||
| gopt::modify_opr_algo_strategy_inplace({y}, strategy); | |||||
| using OprFormatConfigID = LayoutTransformContext::OprFormatConfigID; | |||||
| using OprList = LayoutTransformContext::OprList; | |||||
| using Attribute = LayoutTransformContext::Attribute; | |||||
| using Target = LayoutTransformContext::Target; | |||||
| OprList opr_list = { | |||||
| opr::ConvBiasForward::typeinfo(), opr::ElemwiseMultiType::typeinfo(), | |||||
| opr::Elemwise::typeinfo(), opr::TypeCvt::typeinfo(), | |||||
| opr::Concat::typeinfo(), | |||||
| }; | |||||
| SmallVector<TensorFormats> available_tensor_formats = { | |||||
| TensorFormats::NCHW, TensorFormats::NCHWc4}; | |||||
| Attribute attribute = { | |||||
| OprFormatConfigID::NCHW, TensorFormats::NCHW, Target::UNSPEC}; | |||||
| auto ctx = std::make_unique<LayoutTransformContext>( | |||||
| std::move(opr_list), std::move(available_tensor_formats), attribute); | |||||
| ctx->add_opr_config( | |||||
| opr::ConvBiasForward::typeinfo(), | |||||
| {OprFormatConfigID::NCHW, OprFormatConfigID::NCHW4}); | |||||
| #if MGB_WITH_CACHED_TEST | |||||
| auto profiler = std::make_unique<ProfilerMock>( | |||||
| static_cast<const uint8_t*>(TestLayoutTransform_Concat.data()), | |||||
| TestLayoutTransform_Concat.size()); | |||||
| #else | |||||
| auto profiler = | |||||
| ProfilerBase::make_cached_profiler("TestLayoutTransform.Concat.cache"); | |||||
| #endif | |||||
| std::unique_ptr<SolverBase> solver{ | |||||
| new DynamicProgrammingSolver(std::move(profiler))}; | |||||
| auto new_out_vars = | |||||
| gopt::GraphOptimizer{} | |||||
| .add_pass<LayoutTransformPass>(std::move(ctx), std::move(solver)) | |||||
| .add_pass<ShuffleShuffleRemovePass>() | |||||
| .add_pass(FuseNCHW4Int8Preprocess::make()) | |||||
| #if CUDA_VERSION >= 10020 | |||||
| .add_pass<FoldingConvBiasDimshufflePass>() | |||||
| .add_pass<FoldingConvBiasTypecvtPass>() | |||||
| #endif | |||||
| .add_pass<ParamFusePass>() | |||||
| .add_pass<ParamMergePass>() | |||||
| .apply(SymbolVarArray{y}) | |||||
| .endpoint_vars(); | |||||
| const auto& v = new_out_vars[0]; | |||||
| using OutputSpecItem = cg::ComputingGraph::OutputSpecItem; | |||||
| std::vector<OutputSpecItem> outs; | |||||
| for (auto&& i : new_out_vars) { | |||||
| outs.emplace_back(OutputSpecItem{i, {}}); | |||||
| } | |||||
| GraphProfiler gprof{graph.get()}; | |||||
| auto func = graph->compile(outs); | |||||
| func->execute(); | |||||
| gprof.to_json_full(func.get())->writeto_fpath(output_file("conv_cat.json")); | |||||
| SmallVector<cg::OperatorNodeBase*> oprs; | |||||
| auto cb = [&oprs](cg::OperatorNodeBase* opr) { | |||||
| if (opr->same_type<opr::Concat>()) { | |||||
| oprs.push_back(opr); | |||||
| } | |||||
| }; | |||||
| cg::DepOprIter{cb}.add(v.node()->owner_opr()); | |||||
| ASSERT_EQ(oprs.size(), 4); | |||||
| ASSERT_EQ(oprs[0]->output(0)->shape().ndim, 4); | |||||
| ASSERT_EQ(oprs[2]->output(0)->shape().ndim, 5); | |||||
| } | |||||
| #endif | |||||
| // vim: syntax=cpp.doxygen foldmethod=marker foldmarker=f{{{,f}}} | // vim: syntax=cpp.doxygen foldmethod=marker foldmarker=f{{{,f}}} | ||||