提交 7443b2e4 编写于 作者: Q QI JUN 提交者: GitHub

Merge pull request #3596 from QiJune/implement_random_function

refine random related operators
上级 6eab5638 aff90d8e
develop 2.0.1-rocm-post Ligoml-patch-1 OliverLPH-patch-1 OliverLPH-patch-2 PaddlePM-patch-1 PaddlePM-patch-2 ZHUI-patch-1 add_default_att add_model_benchmark_ci add_some_yaml_config addfile all_new_design_exec ascendrc ascendrelease cherry_undefined_var compile_windows delete_2.0.1-rocm-post delete_add_default_att delete_all_new_design_exec delete_ascendrc delete_compile_windows delete_delete_addfile delete_disable_iterable_dataset_unittest delete_fix_dataloader_memory_leak delete_fix_imperative_dygraph_error delete_fix_retry_ci delete_fix_undefined_var delete_improve_sccache delete_incubate/lite delete_paddle_tiny_install delete_paralleltest delete_prv-disable-more-cache delete_revert-31068-fix_conv3d_windows delete_revert-31562-mean delete_revert-33630-bug-fix delete_revert-34159-add_npu_bce_logical_dev delete_revert-34910-spinlocks_for_allocator delete_revert-35069-revert-34910-spinlocks_for_allocator delete_revert-36057-dev/read_flags_in_ut dingjiaweiww-patch-1 disable_iterable_dataset_unittest dy2static enable_eager_model_test final_state_gen_python_c final_state_intermediate fix-numpy-issue fix_concat_slice fix_dataloader_memory_leak fix_imperative_dygraph_error fix_npu_ci fix_op_flops fix_retry_ci fix_rnn_docs fix_tensor_type fix_undefined_var fixiscan fixiscan1 fixiscan2 fixiscan3 github/fork/123malin/netifaces github/fork/123malin/tdm_abacus github/fork/AshburnLee/dev_unique github/fork/ForFishes/fix_memory_matmul github/fork/ForFishes/rm_fluid github/fork/LielinJiang/move-2.0-api github/fork/LielinJiang/visual-dl-cb github/fork/LiuChiachi/add-transformer-generate-square-subsequent-mask-api github/fork/LiuChiachi/fix-example-code-for-hapi-Model github/fork/LiuChiachi/remove-input-requirment-in-dygraph-Model github/fork/MrChengmo/fix_ps_profiler github/fork/MrChengmo/update_ps_heter github/fork/PWhiddy/patch-1 github/fork/Shixiaowei02/dev/save_load_upgrade github/fork/TCChenlong/fix_hapi github/fork/TCChenlong/fix_inden github/fork/Thunderbrook/xpu_slice github/fork/XieYunshen/disable_ut_test_parallel_executor_fetch_isolated_var github/fork/XieYunshen/disable_ut_test_parallel_executor_fetch_isolated_var_2 github/fork/XieYunshen/disable_ut_test_parallel_executor_fetch_isolated_var_3 github/fork/XieYunshen/timeout_20S_ut github/fork/ZeyuChen/remove-nltk github/fork/arlesniak/arlesniak/selective__mkldnn_flags github/fork/baiyfbupt/code_doc_mig github/fork/chalsliu/set_timeout github/fork/chen-zhiyu/develop github/fork/chenwhql/ci/try_to_find_test_buffer_shared_memory_reuse_pass_error github/fork/chenwhql/dygraph/remove_scale_loss_and_apply_collective_grads github/fork/chenwhql/saveload/add_get_inference_program github/fork/chenwhql/saveload/remove_save_load_config github/fork/cryoco/pass-compatibility-trt github/fork/danleifeng/isempty_api2.0 github/fork/frankwhzhang/api_transfer github/fork/hbwx24/error_msg/cuda_kernel_error_msg github/fork/heavengate/cherry_yolo_box github/fork/heavengate/update_yolo_box github/fork/iclementine/rnn_fix github/fork/iducn/testestse github/fork/jczaja/prv-25537-fix github/fork/jeff41404/release/1.8 github/fork/jiweibo/api_2.0 github/fork/jiweibo/fix_lite_resnet50_test github/fork/juncaipeng/fix_doc_1 github/fork/lfchener/sample_code github/fork/littletomatodonkey/fix_reg_doc github/fork/liym27/dy2stat_update_assign_to_rc20 github/fork/luotao1/profiler_ut github/fork/mapingshuo/add_wait github/fork/mapingshuo/doc_2.0 github/fork/mapingshuo/zero-0.5 github/fork/miraiwk/dev github/fork/pangyoki/add-Categorical-class-branch github/fork/pangyoki/add-multinomial-op-branch github/fork/pangyoki/fix-test_distritbution-CI github/fork/qjing666/doublegrad github/fork/qjing666/fix_hdfs_download github/fork/sandyhouse/add_gather_etc github/fork/sandyhouse/add_send_recv_alltoall_etc github/fork/sandyhouse/pipeline_exe_run github/fork/seiriosPlus/feature/large_scale_kv_save_delta github/fork/seiriosPlus/fix/paddle_errors_fix github/fork/seiriosPlus/fix/paddle_op_errors github/fork/shangzhizhou/fix_test_activation_op_random_bug github/fork/smallv0221/yxp0924 github/fork/smallv0221/yxp0925 github/fork/swtkiwi/del-matplotlib github/fork/tianshuo78520a/kunlun_test github/fork/tianshuo78520a/update_dockerfile github/fork/wanghaoshuang/bert_fuse github/fork/wanghaoshuang/label_smooth github/fork/wanghuancoder/develop_CUDASynchronize github/fork/wanghuancoder/develop_Layer_doc github/fork/wanghuancoder/develop_ParameterList_doc github/fork/wanghuancoder/develop_Sequential_doc github/fork/wanghuancoder/develop_bilinear_tensor_product github/fork/wanghuancoder/develop_coverage_build_sh github/fork/wanghuancoder/develop_in_dynamic_mode_doc github/fork/wanghuancoder/develop_unique_name_doc github/fork/wangxicoding/fleet_meta_combine github/fork/wawltor/error_message_fix_5 github/fork/willthefrog/remove_l2_norm github/fork/windstamp/momentum_op github/fork/windstamp/mv_op_5 github/fork/windstamp/normal_api github/fork/wojtuss/wojtuss/fusion_gru_quantization github/fork/wojtuss/wojtuss/quantization-with-shift github/fork/wzzju/fix_err_info github/fork/wzzju/pure_fp16 github/fork/xiemoyuan/op_error_message github/fork/xiemoyuan/optimize_error_message github/fork/yaoxuefeng6/fix_doc github/fork/yaoxuefeng6/mod_dataset_v2 github/fork/yongqiangma/lod github/fork/ysh329/fix-clip-by-norm-error github/fork/ysh329/fix-error-clip-by-value github/fork/yukavio/error_info github/fork/zhangting2020/conv_filter_grad github/fork/zhangting2020/is_compile_with_cuda github/fork/zhangting2020/place_doc github/fork/zhangting2020/program github/fork/zhhsplendid/fix_any github/fork/zhhsplendid/refine_api2 github/fork/zhhsplendid/refine_api2_test github/fork/zhhsplendid/refine_api_test_ptb_lm github/fork/zhhsplendid/refine_api_test_resnet github/fork/zhhsplendid/refine_api_test_simnet github/fork/zhiqiu/dev/refine_initializer github/fork/zhiqiu/dev/remove_inplace_argument github/fork/zlsh80826/nvinfer_plugin_var_len_cuda11 improve_sccache incubate/infrt incubate/lite inplace_addto make_flag_adding_easier master move_embedding_to_phi move_histogram_to_pten move_sgd_to_phi move_slice_to_pten move_temporal_shift_to_phi move_yolo_box_to_phi npu_fix_alloc numel paddle_tiny_install paralleltest preln_ernie prv-disable-more-cache prv-md-even-more prv-onednn-2.5 pten_tensor_refactor release/0.11.0 release/0.12.0 release/0.13.0 release/0.14.0 release/0.15.0 release/1.0.0 release/1.1 release/1.2 release/1.3 release/1.4 release/1.5 release/1.6 release/1.7 release/1.8 release/2.0 release/2.0-alpha release/2.0-beta release/2.0-rc release/2.0-rc1 release/2.1 release/2.2 release/2.3 release/2.3-fc-ernie-fix release/2.4 release/lite-0.1 revert-24981-add_device_attr_for_regulization revert-26856-strategy_example2 revert-27520-disable_pr revert-31068-fix_conv3d_windows revert-31562-mean revert-32290-develop-hardlabel revert-33037-forci revert-33475-fix_cifar_label_dimension revert-33630-bug-fix revert-34159-add_npu_bce_logical_dev revert-34406-add_copy_from_tensor revert-34910-spinlocks_for_allocator revert-35069-revert-34910-spinlocks_for_allocator revert-36057-dev/read_flags_in_ut revert-36201-refine_fast_threaded_ssa_graph_executor revert-36985-add_license revert-37318-refactor_dygraph_to_eager revert-37926-eager_coreops_500 revert-37956-revert-37727-pylayer_support_tuple revert-38100-mingdong revert-38301-allocation_rearrange_pr revert-38703-numpy_bf16_package_reupload revert-38732-remove_useless_header_in_elementwise_mul_grad revert-38959-Reduce_Grad revert-39143-adjust_empty revert-39227-move_trace_op_to_pten revert-39268-dev/remove_concat_fluid_kernel revert-40170-support_partial_grad revert-41056-revert-40727-move_some_activaion_to_phi revert-41065-revert-40993-mv_ele_floordiv_pow revert-41068-revert-40790-phi_new revert-41944-smaller_inference_api_test revert-42149-do-not-reset-default-stream-for-stream-safe-cuda-allocator revert-43155-fix_ut_tempfile revert-43882-revert-41944-smaller_inference_api_test revert-45808-phi/simplify_size_op revert-46827-deform_comment rocm_dev_0217 support_weight_transpose test_benchmark_ci test_feature_precision_test_c test_model_benchmark test_model_benchmark_ci zhiqiu-patch-1 v2.4.0-rc0 v2.3.2 v2.3.1 v2.3.0 v2.3.0-rc0 v2.2.2 v2.2.1 v2.2.0 v2.2.0-rc0 v2.2.0-bak0 v2.1.3 v2.1.2 v2.1.1 v2.1.0 v2.1.0-rc0 v2.0.2 v2.0.1 v2.0.0 v2.0.0-rc1 v2.0.0-rc0 v2.0.0-beta0 v2.0.0-alpha0 v1.8.5 v1.8.4 v1.8.3 v1.8.2 v1.8.1 v1.8.0 v1.7.2 v1.7.1 v1.7.0 v1.6.3 v1.6.2 v1.6.1 v1.6.0 v1.6.0-rc0 v1.5.2 v1.5.1 v1.5.0 v1.4.1 v1.4.0 v1.3.2 v1.3.1 v1.3.0 v1.2.1 v1.2.0 v1.1.0 v1.0.2 v1.0.1 v1.0.0 v1.0.0-rc0 v0.15.0 v0.15.0-rc0 v0.14.0 v0.13.0 v0.12.0 v0.11.1a2 v0.11.1a1 v0.11.0 lite-v0.1
无相关合并请求
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
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.
......@@ -19,25 +16,25 @@ namespace paddle {
namespace operators {
template <typename T>
class GaussianRandomKernel : public framework::OpKernel {
class CPUGaussianRandomKernel : public framework::OpKernel {
public:
void Compute(const framework::ExecutionContext& context) const override {
float mean = context.op_.GetAttr<float>("mean");
float std = context.op_.GetAttr<float>("std");
auto* tensor = context.Output<framework::Tensor>(0);
auto* tensor = context.Output<framework::Tensor>("Out");
T* data = tensor->mutable_data<T>(context.GetPlace());
// TODO(dzh): attribute does not support unsigned int.
// And we need a global random seed configuration.
int seed = context.op_.GetAttr<int>("seed");
unsigned int seed =
static_cast<unsigned int>(context.op_.GetAttr<int>("seed"));
std::minstd_rand engine;
if (seed == 0) {
seed = std::random_device()();
}
std::mt19937 g(seed);
std::normal_distribution<T> distribution(mean, std);
engine.seed(seed);
std::normal_distribution<T> dist(mean, std);
ssize_t size = framework::product(tensor->dims());
for (int i = 0; i < size; ++i) {
data[i] = distribution(g);
for (ssize_t i = 0; i < size; ++i) {
data[i] = dist(engine);
}
}
};
......@@ -48,7 +45,7 @@ class GaussianRandomOp : public framework::OperatorWithKernel {
protected:
void InferShape(const framework::InferShapeContext& context) const override {
auto* tensor = context.Output<framework::Tensor>(0);
auto* tensor = context.Output<framework::Tensor>("Out");
auto dims = GetAttr<std::vector<int>>("dims");
PADDLE_ENFORCE(dims.size() > 0UL,
"dims can be one int or array. dims must be set.");
......@@ -68,8 +65,8 @@ Use to initialize tensor with gaussian random generator.
)DOC");
AddAttr<std::vector<int>>("dims", "The dimension of random tensor.");
AddAttr<float>("mean", "mean value of random.").SetDefault(.0f);
AddAttr<float>("std", "minimum value of random value.").SetDefault(1.0f);
AddAttr<float>("mean", "mean of random tensor.").SetDefault(.0f);
AddAttr<float>("std", "std of random tensor.").SetDefault(1.0f);
AddAttr<int>("seed",
"Random seed of generator."
"0 means use system wide seed")
......@@ -83,4 +80,4 @@ Use to initialize tensor with gaussian random generator.
namespace ops = paddle::operators;
REGISTER_OP_WITHOUT_GRADIENT(gaussian_random, ops::GaussianRandomOp,
ops::GaussianRandomOpMaker);
REGISTER_OP_CPU_KERNEL(gaussian_random, ops::GaussianRandomKernel<float>);
REGISTER_OP_CPU_KERNEL(gaussian_random, ops::CPUGaussianRandomKernel<float>);
\ No newline at end of file
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
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. */
#include <memory>
#include <random>
#include "paddle/platform/dynload/curand.h"
#include "paddle/platform/gpu_info.h"
#include <thrust/device_ptr.h>
#include <thrust/iterator/counting_iterator.h>
#include <thrust/random.h>
#include <thrust/transform.h>
#include "paddle/framework/op_registry.h"
#include "paddle/framework/operator.h"
namespace paddle {
namespace operators {
template <typename T>
class GaussianRandomKernel : public framework::OpKernel {
struct GaussianGenerator {
T mean_, std_;
unsigned int seed_;
__host__ __device__ GaussianGenerator(T mean, T std, int seed)
: mean_(mean), std_(std), seed_(seed) {}
__host__ __device__ T operator()(const unsigned int n) const {
thrust::minstd_rand rng;
rng.seed(seed_);
thrust::normal_distribution<T> dist(mean_, std_);
rng.discard(n);
return dist(rng);
}
};
template <typename T>
class GPUGaussianRandomKernel : public framework::OpKernel {
public:
void Compute(const framework::ExecutionContext& context) const override {
float mean = context.op_.GetAttr<float>("mean");
float std = context.op_.GetAttr<float>("std");
auto* tensor = context.Output<framework::Tensor>(0);
auto* tensor = context.Output<framework::Tensor>("Out");
T* data = tensor->mutable_data<T>(context.GetPlace());
int seed = context.op_.GetAttr<int>("seed");
unsigned int seed =
static_cast<unsigned int>(context.op_.GetAttr<int>("seed"));
if (seed == 0) {
std::random_device rd;
seed = rd();
}
curandGenerator_t g;
PADDLE_ENFORCE(platform::dynload::curandCreateGenerator(
&g, CURAND_RNG_PSEUDO_DEFAULT));
PADDLE_ENFORCE(
platform::dynload::curandSetPseudoRandomGeneratorSeed(g, seed));
platform::dynload::curandGenerateNormal(
g, data, framework::product(tensor->dims()), mean, std);
T mean = static_cast<T>(context.op_.GetAttr<float>("mean"));
T std = static_cast<T>(context.op_.GetAttr<float>("std"));
thrust::counting_iterator<unsigned int> index_sequence_begin(0);
ssize_t N = framework::product(tensor->dims());
thrust::transform(index_sequence_begin, index_sequence_begin + N,
thrust::device_ptr<T>(data),
GaussianGenerator<T>(mean, std, seed));
}
};
} // namespace operators
} // namespace paddle
namespace ops = paddle::operators;
REGISTER_OP_GPU_KERNEL(gaussian_random, ops::GaussianRandomKernel<float>);
REGISTER_OP_GPU_KERNEL(gaussian_random,
paddle::operators::GPUGaussianRandomKernel<float>);
......@@ -13,7 +13,6 @@
limitations under the License. */
#include "paddle/operators/mul_op.h"
#include "paddle/operators/math/math_function.h"
namespace paddle {
namespace operators {
......
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
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.
......@@ -39,7 +36,8 @@ class CPUUniformRandomKernel : public framework::OpKernel {
std::uniform_real_distribution<T> dist(
static_cast<T>(context.op_.GetAttr<float>("min")),
static_cast<T>(context.op_.GetAttr<float>("max")));
for (ssize_t i = 0; i < framework::product(tensor->dims()); ++i) {
ssize_t size = framework::product(tensor->dims());
for (ssize_t i = 0; i < size; ++i) {
data[i] = dist(engine);
}
}
......@@ -66,7 +64,6 @@ class UniformRandomOpMaker : public framework::OpProtoAndCheckerMaker {
: framework::OpProtoAndCheckerMaker(proto, op_checker) {
AddOutput("Out", "The output tensor of uniform random op");
AddComment(R"DOC(Uniform random operator.
Used to initialize tensor with uniform random generator.
)DOC");
AddAttr<std::vector<int>>("dims", "the dimension of random tensor");
......@@ -84,4 +81,4 @@ Used to initialize tensor with uniform random generator.
REGISTER_OP_WITHOUT_GRADIENT(uniform_random, paddle::operators::UniformRandomOp,
paddle::operators::UniformRandomOpMaker);
REGISTER_OP_CPU_KERNEL(uniform_random,
paddle::operators::CPUUniformRandomKernel<float>);
paddle::operators::CPUUniformRandomKernel<float>);
\ No newline at end of file
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
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.
......@@ -68,4 +65,4 @@ class GPUUniformRandomKernel : public framework::OpKernel {
} // namespace paddle
REGISTER_OP_GPU_KERNEL(uniform_random,
paddle::operators::GPUUniformRandomKernel<float>);
paddle::operators::GPUUniformRandomKernel<float>);
\ No newline at end of file
......@@ -114,9 +114,6 @@ CUDADeviceContext::~CUDADeviceContext() {
PADDLE_ENFORCE(dynload::cudnnDestroy(cudnn_handle_));
}
if (curand_generator_) {
PADDLE_ENFORCE(dynload::curandDestroyGenerator(curand_generator_));
}
eigen_stream_.reset();
eigen_device_.reset();
PADDLE_ENFORCE(cudaStreamDestroy(stream_));
......@@ -152,19 +149,6 @@ cudnnHandle_t CUDADeviceContext::cudnn_handle() {
cudaStream_t CUDADeviceContext::stream() { return stream_; }
curandGenerator_t CUDADeviceContext::curand_generator() {
if (!curand_generator_) {
SetDeviceId(place_.device);
PADDLE_ENFORCE(dynload::curandCreateGenerator(&curand_generator_,
CURAND_RNG_PSEUDO_DEFAULT));
PADDLE_ENFORCE(
dynload::curandSetPseudoRandomGeneratorSeed(curand_generator_, seed_));
PADDLE_ENFORCE(dynload::curandSetStream(curand_generator_, stream_));
}
return curand_generator_;
}
#endif // PADDLE_ONLY_CPU
} // namespace platform
......
......@@ -17,7 +17,6 @@ limitations under the License. */
#ifndef PADDLE_ONLY_CPU
#include "paddle/platform/dynload/cublas.h"
#include "paddle/platform/dynload/cudnn.h"
#include "paddle/platform/dynload/curand.h"
#include "paddle/platform/gpu_info.h"
#define EIGEN_USE_GPU
#endif
......@@ -40,7 +39,7 @@ class DeviceContext {
class CPUDeviceContext : public DeviceContext {
public:
CPUDeviceContext();
explicit CPUDeviceContext(CPUPlace);
explicit CPUDeviceContext(CPUPlace place);
virtual ~CPUDeviceContext() {}
Eigen::DefaultDevice* eigen_device() const;
......@@ -56,7 +55,7 @@ class EigenCudaStreamDevice;
class CUDADeviceContext : public DeviceContext {
public:
explicit CUDADeviceContext(GPUPlace);
explicit CUDADeviceContext(GPUPlace place);
virtual ~CUDADeviceContext();
/*! \brief Wait for all operations completion in the stream. */
......@@ -75,9 +74,6 @@ class CUDADeviceContext : public DeviceContext {
/*! \brief Return cudnn handle in the device context. */
cudnnHandle_t cudnn_handle();
/*! \brief Return curand handle in the device context. */
curandGenerator_t curand_generator();
/*! \brief Return cuda stream in the device context. */
cudaStream_t stream();
// clang-format on
......@@ -85,18 +81,13 @@ class CUDADeviceContext : public DeviceContext {
private:
GPUPlace place_;
private:
std::unique_ptr<Eigen::GpuDevice> eigen_device_;
std::unique_ptr<EigenCudaStreamDevice> eigen_stream_;
private:
uint64_t seed_;
// clang-format off
cudaStream_t stream_{nullptr};
cudnnHandle_t cudnn_handle_{nullptr};
cublasHandle_t cublas_handle_{nullptr};
curandGenerator_t curand_generator_{nullptr};
// clang-format on
};
......
......@@ -43,8 +43,6 @@ TEST(Device, CUDADeviceContext) {
ASSERT_NE(nullptr, cudnn_handle);
cublasHandle_t cublas_handle = device_context->cublas_handle();
ASSERT_NE(nullptr, cublas_handle);
curandGenerator_t curand_handle = device_context->curand_generator();
ASSERT_NE(nullptr, curand_handle);
ASSERT_NE(nullptr, device_context->stream());
delete device_context;
}
......
......@@ -22,7 +22,7 @@ py_test(test_rowwise_add_op SRCS test_rowwise_add_op.py)
py_test(test_default_scope_funcs SRCS test_default_scope_funcs.py)
py_test(test_operator SRCS test_operator.py)
# py_test(test_gaussian_random_op SRCS test_gaussian_random_op.py)
py_test(test_gaussian_random_op SRCS test_gaussian_random_op.py)
py_test(test_uniform_random_op SRCS test_uniform_random_op.py)
py_test(test_recurrent_op SRCS test_recurrent_op.py)
py_test(test_sgd_op SRCS test_sgd_op.py)
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册
反馈
建议
客服 返回
顶部