未验证 提交 a1c281f0 编写于 作者: Y Yibing Liu 提交者: GitHub

Merge pull request #7603 from kuke/simplify_erase

Enhance GPU kernel of sequence erase op 
上级 41b83884 8809d43a
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/0.12.0 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 lite-v0.1
无相关合并请求
......@@ -86,4 +86,5 @@ REGISTER_OP_WITHOUT_GRADIENT(sequence_erase, ops::SequenceEraseOp,
ops::SequenceEraseOpMaker);
REGISTER_OP_CPU_KERNEL(
sequence_erase,
ops::SequenceEraseKernel<paddle::platform::CPUDeviceContext, int32_t>);
ops::SequenceEraseKernel<paddle::platform::CPUDeviceContext, int32_t>,
ops::SequenceEraseKernel<paddle::platform::CPUDeviceContext, int64_t>);
......@@ -23,27 +23,22 @@ using platform::PADDLE_CUDA_NUM_THREADS;
using LoDTensor = framework::LoDTensor;
template <typename T>
__global__ void LabelErasedIdx(const T* in_dat, const int in_len,
const T* tokens, const int tokens_len,
int* num_erased) {
__global__ void LabelErasedIdx(const T* in_dat, const int64_t in_len,
const int* tokens, const size_t tokens_len,
size_t* num_erased) {
int index = blockIdx.x * blockDim.x + threadIdx.x;
if (index < in_len) {
int erased = 0;
for (int i = 0; i < tokens_len; ++i) {
for (size_t i = 0; i < tokens_len; ++i) {
if (in_dat[index] == tokens[i]) {
erased = 1;
num_erased[index + 1] = 1;
break;
}
}
num_erased[index + 1] = erased;
if (index == 0) {
num_erased[0] = 0;
}
}
}
template <typename T>
__global__ void GetOutLod(const T* num_erased, const int* in_lod,
const int lod_len, int* out_lod0) {
__global__ void GetOutLod(const size_t* num_erased, const size_t* in_lod,
const size_t lod_len, size_t* out_lod0) {
int index = blockIdx.x * blockDim.x + threadIdx.x;
if (index < lod_len) {
out_lod0[index] = in_lod[index] - num_erased[in_lod[index]];
......@@ -51,11 +46,11 @@ __global__ void GetOutLod(const T* num_erased, const int* in_lod,
}
template <typename T>
__global__ void SetOutput(const T* in_dat, const int in_len,
const int* num_erased, T* out_dat) {
__global__ void SetOutput(const T* in_dat, const int64_t in_len,
const size_t* num_erased, T* out_dat) {
int index = blockIdx.x * blockDim.x + threadIdx.x;
if (index < in_len) {
if (in_dat[index] != in_dat[index + 1]) {
if (num_erased[index] == num_erased[index + 1]) {
out_dat[index - num_erased[index]] = in_dat[index];
}
}
......@@ -72,53 +67,44 @@ class SequenceEraseOpCUDAKernel : public framework::OpKernel<T> {
PADDLE_ENFORCE_EQ(lod.size(), 1UL, "Only support one level sequence now.");
PADDLE_ENFORCE_EQ(lod[0].back(), (size_t)in->numel(),
"The actual size mismatches with the LoD information.");
auto tokens = ctx.Attr<std::vector<T>>("tokens");
auto tokens_len = tokens.size();
auto tokens = ctx.Attr<std::vector<int>>("tokens");
auto in_len = in->numel();
auto in_dat = in->data<T>();
auto lod0 = lod[0];
thrust::host_vector<T> host_tokens(tokens_len);
for (size_t i = 0; i < tokens.size(); ++i) {
host_tokens[i] = tokens[i];
}
thrust::device_vector<T> dev_tokens = host_tokens;
thrust::device_vector<int> num_erased(in_len + 1);
T* dev_tokens_ptr = thrust::raw_pointer_cast(dev_tokens.data());
int* num_erased_ptr = thrust::raw_pointer_cast(num_erased.data());
// Copy tokens to GPU
thrust::device_vector<int> dev_tokens(tokens.begin(), tokens.end());
int* dev_tokens_ptr = thrust::raw_pointer_cast(dev_tokens.data());
// Count number of elements to be erased
thrust::device_vector<size_t> num_erased(in_len + 1, 0);
size_t* num_erased_ptr = thrust::raw_pointer_cast(num_erased.data());
auto stream = ctx.cuda_device_context().stream();
LabelErasedIdx<<<(in_len - 1) / PADDLE_CUDA_NUM_THREADS + 1,
PADDLE_CUDA_NUM_THREADS, 0, stream>>>(
in_dat, in_len, dev_tokens_ptr, tokens_len, num_erased_ptr);
in_dat, in_len, dev_tokens_ptr, tokens.size(), num_erased_ptr);
thrust::inclusive_scan(num_erased.begin() + 1, num_erased.end(),
num_erased.begin() + 1);
// Calc LoD
// Copy LoD to GPU
auto lod0 = lod[0];
auto lod_len = lod0.size();
thrust::host_vector<int> host_lod(lod_len);
for (size_t i = 0; i < lod_len; ++i) {
host_lod[i] = lod0[i];
}
thrust::device_vector<int> dev_in_lod = host_lod;
thrust::device_vector<int> dev_out_lod(lod_len);
int* dev_in_lod_ptr = thrust::raw_pointer_cast(dev_in_lod.data());
int* dev_out_lod_ptr = thrust::raw_pointer_cast(dev_out_lod.data());
thrust::device_vector<size_t> dev_in_lod = lod0;
size_t* dev_in_lod_ptr = thrust::raw_pointer_cast(dev_in_lod.data());
// Calc output LoD
thrust::device_vector<size_t> dev_out_lod(lod_len);
size_t* dev_out_lod_ptr = thrust::raw_pointer_cast(dev_out_lod.data());
GetOutLod<<<(lod_len - 1) / PADDLE_CUDA_NUM_THREADS + 1,
PADDLE_CUDA_NUM_THREADS, 0, stream>>>(
num_erased_ptr, dev_in_lod_ptr, lod_len, dev_out_lod_ptr);
thrust::host_vector<int> host_out_lod = dev_out_lod;
std::vector<int> out_lod0(lod_len, 0);
for (size_t i = 0; i < lod_len; i++) {
out_lod0[i] = host_out_lod[i];
}
// Set LoD for output
thrust::host_vector<size_t> out_lod0 = dev_out_lod;
framework::LoD out_lod;
out_lod.push_back(out_lod0);
out->set_lod(out_lod);
// Set output
out->Resize({out_lod0.back(), 1});
out->Resize({static_cast<int64_t>(out_lod0.back()), 1});
auto out_dat = out->mutable_data<T>(ctx.GetPlace());
SetOutput<<<(in_len - 1) / PADDLE_CUDA_NUM_THREADS + 1,
PADDLE_CUDA_NUM_THREADS, 0, stream>>>(in_dat, in_len,
......@@ -130,4 +116,5 @@ class SequenceEraseOpCUDAKernel : public framework::OpKernel<T> {
} // namespace paddle
REGISTER_OP_CUDA_KERNEL(sequence_erase,
paddle::operators::SequenceEraseOpCUDAKernel<int32_t>);
paddle::operators::SequenceEraseOpCUDAKernel<int32_t>,
paddle::operators::SequenceEraseOpCUDAKernel<int64_t>);
......@@ -29,7 +29,7 @@ def sequence_erase(in_seq, lod0, tokens):
return np.array(out_seq).astype("int32"), new_lod0
class TestSequenceEraseOp(OpTest):
class TestSequenceEraseOpInt32(OpTest):
def setUp(self):
self.op_type = "sequence_erase"
in_seq = np.random.randint(0, 10, (30, 1)).astype("int32")
......@@ -44,5 +44,35 @@ class TestSequenceEraseOp(OpTest):
self.check_output()
class TestSequenceEraseOpInt64(OpTest):
def setUp(self):
self.op_type = "sequence_erase"
in_seq = np.random.randint(0, 10, (30, 1)).astype("int64")
lod = [[0, 9, 13, 24, 30]]
tokens = [2, 3, 5]
out_seq, new_lod0 = sequence_erase(in_seq, lod[0], tokens)
self.attrs = {'tokens': tokens}
self.inputs = {'X': (in_seq, lod)}
self.outputs = {'Out': (out_seq, [new_lod0])}
def test_check_output(self):
self.check_output()
class TestSequenceEraseOpEmpty(OpTest):
def setUp(self):
self.op_type = "sequence_erase"
in_seq = np.random.randint(0, 10, (30, 1)).astype("int32")
lod = [[0, 9, 13, 24, 30]]
tokens = []
out_seq, new_lod0 = sequence_erase(in_seq, lod[0], tokens)
self.attrs = {'tokens': tokens}
self.inputs = {'X': (in_seq, lod)}
self.outputs = {'Out': (out_seq, [new_lod0])}
def test_check_output(self):
self.check_output()
if __name__ == '__main__':
unittest.main()
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