From 2012672c10c53369936f988aab47ab9001fabab0 Mon Sep 17 00:00:00 2001 From: Shuangchi He <34329208+Yulv-git@users.noreply.github.com> Date: Thu, 12 May 2022 23:25:03 +0800 Subject: [PATCH] Fix some typos in paddle/. (#42408) --- .../fluid/distributed/ps/service/ps_client.h | 2 +- paddle/fluid/framework/data_set.h | 2 +- .../details/sparse_all_reduce_op_handle.cc | 4 +- .../fluid/framework/heter_pipeline_trainer.cc | 2 +- .../fuse_optimizer_op_pass.cc | 2 +- .../framework/ir/fusion_group/operation.cc | 2 +- .../framework/ir/graph_pattern_detector.cc | 2 +- .../framework/ir/graph_pattern_detector.h | 2 +- .../framework/new_executor/interpretercore.cc | 2 +- .../new_executor/interpretercore_util.cc | 2 +- .../new_executor/workqueue/workqueue.h | 2 +- .../framework/paddle2cinn/build_cinn_pass.cc | 2 +- paddle/fluid/framework/prune.cc | 2 +- paddle/fluid/imperative/layer.cc | 2 +- paddle/fluid/imperative/reducer.cc | 4 +- .../fluid/inference/analysis/CMakeLists.txt | 2 +- .../analysis/ir_passes/subgraph_util.cc | 2 +- .../analysis/ir_passes/subgraph_util.h | 2 +- .../ir_passes/tensorrt_subgraph_pass.cc | 2 +- .../inference/api/paddle_analysis_config.h | 2 +- .../inference/tensorrt/convert/swish_op.cc | 2 +- .../analyzer_lexical_analysis_gru_tester.cc | 6 +-- paddle/fluid/operators/activation_op.h | 2 +- paddle/fluid/operators/affine_grid_op.cc | 2 +- paddle/fluid/operators/batch_norm_op.cc | 2 +- .../operators/collective/c_broadcast_op.cu.cc | 2 +- .../collective/c_broadcast_op_mlu.cc | 2 +- .../collective/c_broadcast_op_npu.cc | 2 +- paddle/fluid/operators/conv_op.h | 6 +-- paddle/fluid/operators/conv_op_mlu.cc | 2 +- paddle/fluid/operators/ctc_align_op.cu | 8 ++-- .../operators/deformable_psroi_pooling_op.cu | 2 +- .../operators/deformable_psroi_pooling_op.h | 2 +- .../operators/detection/matrix_nms_op.cc | 2 +- paddle/fluid/operators/dropout_op_npu.cc | 2 +- .../operators/elementwise/elementwise_mlu.h | 2 +- paddle/fluid/operators/fc_op.cc | 4 +- paddle/fluid/operators/fc_op.h | 2 +- paddle/fluid/operators/fill_constant_op.cc | 2 +- paddle/fluid/operators/fold_op.cc | 16 +++---- .../fluid/operators/fused/conv_fusion_op.cc | 4 +- .../fluid/operators/fused/conv_fusion_op.cu | 6 +-- .../fused/cudnn_bn_stats_finalize.cu.h | 2 +- .../operators/fused/cudnn_norm_conv.cu.h | 12 ++--- .../fused/cudnn_scale_bias_add_relu.cu.h | 6 +-- .../fused_fc_elementwise_layernorm_op.cc | 16 +++---- .../fluid/operators/fused/fusion_group_op.cc | 2 +- paddle/fluid/operators/inverse_op.cc | 6 +-- paddle/fluid/operators/lod_reset_op.h | 2 +- paddle/fluid/operators/math/cross_entropy.h | 2 +- .../operators/math/selected_rows_functor.cc | 40 ++++++++-------- .../operators/math/selected_rows_functor.cu | 46 +++++++++---------- paddle/fluid/operators/metrics/accuracy_op.cc | 2 +- paddle/fluid/operators/mlu/mlu_baseop.cc | 4 +- .../operators/positive_negative_pair_op.cc | 4 +- .../operators/prune_gate_by_capacity_op.cu | 2 +- .../pscore/heter_listen_and_serv_op.cc | 4 +- paddle/fluid/operators/rnn_op_xpu.cc | 2 +- paddle/fluid/operators/sample_logits_op.cc | 2 +- ...igmoid_cross_entropy_with_logits_op_npu.cc | 2 +- paddle/fluid/operators/sum_op.cu | 4 +- paddle/fluid/operators/tdm_child_op.h | 2 +- paddle/fluid/operators/warpctc_op.cc | 2 +- paddle/fluid/platform/CMakeLists.txt | 2 +- .../fluid/platform/device/npu/npu_op_runner.h | 2 +- paddle/fluid/platform/profiler/event_node.cc | 2 +- paddle/infrt/common/object.h | 2 +- paddle/phi/infermeta/binary.cc | 2 +- paddle/phi/infermeta/multiary.cc | 4 +- paddle/phi/infermeta/unary.cc | 20 ++++---- paddle/phi/kernels/cpu/conv_util.h | 2 +- paddle/phi/kernels/cpu/rnn_kernel.cc | 2 +- paddle/phi/kernels/funcs/broadcast_function.h | 8 ++-- paddle/phi/kernels/funcs/elementwise_base.h | 2 +- paddle/phi/kernels/gpu/bce_loss_kernel.cu | 2 +- .../kernels/impl/searchsorted_kernel_impl.h | 2 +- paddle/phi/ops/compat/scale_sig.cc | 2 +- python/paddle/README.rst | 2 +- .../auto_parallel/process_group.py | 2 +- .../fleet/base/distributed_strategy.py | 8 ++-- python/paddle/distributed/fleet/launch.py | 2 +- .../meta_optimizers/ascend/ascend_parser.py | 2 +- .../dygraph_sharding_optimizer.py | 4 +- python/paddle/distributed/launch/main.py | 2 +- .../distributed/passes/auto_parallel_fp16.py | 2 +- python/paddle/distributed/ps/utils/public.py | 2 +- ..._post_training_quantization_mobilenetv1.py | 2 +- python/paddle/fluid/dygraph/checkpoint.py | 2 +- .../dygraph_to_static/convert_operators.py | 2 +- python/paddle/fluid/executor.py | 2 +- .../fleet/parameter_server/ir/trainer_pass.py | 2 +- python/paddle/fluid/layer_helper_base.py | 2 +- python/paddle/fluid/layers/nn.py | 2 +- python/paddle/fluid/layers/rnn.py | 2 +- .../unittests/ir/inference/auto_scan_test.py | 2 +- .../test_shuffle_channel_detect_pass.py | 4 +- .../fluid/tests/unittests/ir/pass_test.py | 2 +- .../paddle/fluid/tests/unittests/op_test.py | 6 +-- .../tests/unittests/seresnext_test_base.py | 4 +- .../static_model_parallel_fused_attention.py | 4 +- ...static_model_parallel_fused_feedforward.py | 4 +- .../tests/unittests/test_fleet_rolemaker.py | 2 +- .../unittests/test_fleet_rolemaker_new.py | 2 +- .../tests/unittests/test_gradient_clip.py | 18 ++++---- .../fluid/tests/unittests/test_hsigmoid_op.py | 2 +- .../fluid/tests/unittests/test_optimizer.py | 2 +- ...st_parallel_executor_seresnext_base_cpu.py | 2 +- ...st_parallel_executor_seresnext_base_gpu.py | 2 +- .../fluid/tests/unittests/test_unpool_op.py | 4 +- python/paddle/framework/io.py | 2 +- python/paddle/hapi/model.py | 4 +- python/paddle/incubate/autotune.py | 2 +- .../distributed/models/moe/grad_clip.py | 2 +- .../incubate/nn/layer/fused_transformer.py | 20 ++++---- python/paddle/nn/functional/loss.py | 6 +-- python/paddle/nn/layer/activation.py | 2 +- python/paddle/nn/layer/transformer.py | 22 ++++----- python/paddle/profiler/profiler.py | 2 +- python/paddle/profiler/timer.py | 8 ++-- python/paddle/profiler/utils.py | 2 +- python/paddle/tensor/math.py | 4 +- python/paddle/tensor/to_string.py | 2 +- python/paddle/vision/models/mobilenetv3.py | 2 +- 123 files changed, 262 insertions(+), 262 deletions(-) diff --git a/paddle/fluid/distributed/ps/service/ps_client.h b/paddle/fluid/distributed/ps/service/ps_client.h index 0d3d23be4e8..926bb7e7c9f 100644 --- a/paddle/fluid/distributed/ps/service/ps_client.h +++ b/paddle/fluid/distributed/ps/service/ps_client.h @@ -109,7 +109,7 @@ class PSClient { size_t table_id) = 0; // 保留 // firstly push dense param for parameter server - // this is neccessary because dense weight initialized in trainer on cold + // this is necessary because dense weight initialized in trainer on cold // start virtual std::future PushDenseParam(const Region *regions, size_t region_num, diff --git a/paddle/fluid/framework/data_set.h b/paddle/fluid/framework/data_set.h index 1947c669e9b..3f10cd7765b 100644 --- a/paddle/fluid/framework/data_set.h +++ b/paddle/fluid/framework/data_set.h @@ -152,7 +152,7 @@ class Dataset { virtual void DestroyPreLoadReaders() = 0; // set preload thread num virtual void SetPreLoadThreadNum(int thread_num) = 0; - // seperate train thread and dataset thread + // separate train thread and dataset thread virtual void DynamicAdjustChannelNum(int channel_num, bool discard_remaining_ins = false) = 0; virtual void DynamicAdjustReadersNum(int thread_num) = 0; diff --git a/paddle/fluid/framework/details/sparse_all_reduce_op_handle.cc b/paddle/fluid/framework/details/sparse_all_reduce_op_handle.cc index d198eb14592..7e63c5ffb9a 100644 --- a/paddle/fluid/framework/details/sparse_all_reduce_op_handle.cc +++ b/paddle/fluid/framework/details/sparse_all_reduce_op_handle.cc @@ -75,7 +75,7 @@ void SparseAllReduceOpHandle::RunImplEncoded() { in_var_handles.size(), places_.size(), platform::errors::PreconditionNotMet( "The number of input variables should be equal to the number of " - "places, but got the number of input variables is %zu and the the " + "places, but got the number of input variables is %zu and the " "number of places is %zu.", in_var_handles.size(), places_.size())); PADDLE_ENFORCE_EQ( @@ -83,7 +83,7 @@ void SparseAllReduceOpHandle::RunImplEncoded() { platform::errors::PreconditionNotMet( "The number of input variables should be equal to the number of " "output variables, but got the number of input variables is %zu and " - "the the number of output variables is %zu.", + "the number of output variables is %zu.", in_var_handles.size(), out_var_handles.size())); std::vector ins; diff --git a/paddle/fluid/framework/heter_pipeline_trainer.cc b/paddle/fluid/framework/heter_pipeline_trainer.cc index 13eb78874c3..d0d3c2fea3b 100644 --- a/paddle/fluid/framework/heter_pipeline_trainer.cc +++ b/paddle/fluid/framework/heter_pipeline_trainer.cc @@ -282,7 +282,7 @@ void HeterPipelineTrainer::Run() { if (threads_.size() > 0) { threads_.clear(); } - VLOG(3) << "Epoch Trainging done"; + VLOG(3) << "Epoch Training done"; } void HeterPipelineTrainer::Finalize() { diff --git a/paddle/fluid/framework/ir/fuse_optimizer_ops_pass/fuse_optimizer_op_pass.cc b/paddle/fluid/framework/ir/fuse_optimizer_ops_pass/fuse_optimizer_op_pass.cc index 48df5869a7a..40e1de8a523 100644 --- a/paddle/fluid/framework/ir/fuse_optimizer_ops_pass/fuse_optimizer_op_pass.cc +++ b/paddle/fluid/framework/ir/fuse_optimizer_ops_pass/fuse_optimizer_op_pass.cc @@ -172,7 +172,7 @@ void FuseOptimizerOpPass::ApplyImpl(ir::Graph *graph) const { VLOG(6) << "The number of new gradients is " << new_grad_idx.size(); if (new_grad_idx.size() == 1) return; // NOTE(zcd): If the gradients of backward stage and optimization stage - // have diff, Only take care of the the gradient of optimization stage. + // have diff, Only take care of the gradient of optimization stage. GradientsFilter(new_grad_idx, &opt_nodes, &aux_var_map); } } diff --git a/paddle/fluid/framework/ir/fusion_group/operation.cc b/paddle/fluid/framework/ir/fusion_group/operation.cc index 921cf0904f6..2b7a3e1899c 100644 --- a/paddle/fluid/framework/ir/fusion_group/operation.cc +++ b/paddle/fluid/framework/ir/fusion_group/operation.cc @@ -127,7 +127,7 @@ void OperationMap::InsertUnaryElementwiseOperations() { // scale // out = (bias_after_scale) ? scale * X + bias : scale(X + bias) - // here we use '=' operator to seperate th default value + // here we use '=' operator to separate th default value // TODO(wangchaochaohu): Later we need to support Tensor input for scale and // bias. insert_handler( diff --git a/paddle/fluid/framework/ir/graph_pattern_detector.cc b/paddle/fluid/framework/ir/graph_pattern_detector.cc index fbd8fda131b..8c8d9fdddec 100644 --- a/paddle/fluid/framework/ir/graph_pattern_detector.cc +++ b/paddle/fluid/framework/ir/graph_pattern_detector.cc @@ -720,7 +720,7 @@ bool HasOutput(Node *op, const std::string &argument) { PADDLE_ENFORCE_EQ( op->IsOp(), true, platform::errors::InvalidArgument( - "First parameter of function HasOuput must be Node::Op")); + "First parameter of function HasOutput must be Node::Op")); auto const &names = op->Op()->OutputNames(); if (std::find(names.begin(), names.end(), argument) == names.end()) return false; diff --git a/paddle/fluid/framework/ir/graph_pattern_detector.h b/paddle/fluid/framework/ir/graph_pattern_detector.h index c9fea057d44..9e5a82fc445 100644 --- a/paddle/fluid/framework/ir/graph_pattern_detector.h +++ b/paddle/fluid/framework/ir/graph_pattern_detector.h @@ -1432,7 +1432,7 @@ struct PriorBox : public PatternBase { }; // Conv + ElementwiseAdd + an activation -// This pattern can futher fuse the conv related ops after the conv+bn fusion. +// This pattern can further fuse the conv related ops after the conv+bn fusion. struct ConvElementwiseaddAct : public PatternBase { ConvElementwiseaddAct(PDPattern* pattern, const std::string& name_scope) : PatternBase(pattern, name_scope, "conv_elementwiseadd_act") {} diff --git a/paddle/fluid/framework/new_executor/interpretercore.cc b/paddle/fluid/framework/new_executor/interpretercore.cc index 6735406aacd..da2fd0c8c61 100644 --- a/paddle/fluid/framework/new_executor/interpretercore.cc +++ b/paddle/fluid/framework/new_executor/interpretercore.cc @@ -277,7 +277,7 @@ void InterpreterCore::Convert( } for (size_t i = 0; i < vec_instruction_.size(); ++i) { - // checkout ouput + // checkout output for (auto& item : vec_instruction_[i].Outputs()) { for (auto var_id : item.second) { if (input_var2op_info_.at(var_id).size() == 0) { diff --git a/paddle/fluid/framework/new_executor/interpretercore_util.cc b/paddle/fluid/framework/new_executor/interpretercore_util.cc index d6de37a72c7..f601a4ad28b 100644 --- a/paddle/fluid/framework/new_executor/interpretercore_util.cc +++ b/paddle/fluid/framework/new_executor/interpretercore_util.cc @@ -666,7 +666,7 @@ std::map> get_downstream_map( VLOG(6) << "downstream count: " << downstream_map_count(); VLOG(6) << "downstream_map: " << std::endl << downstream_map_to_str(); - // step2: remove unneccessary downstream ops + // step2: remove unnecessary downstream ops // for example, a->b->c // a: b, c // b: c diff --git a/paddle/fluid/framework/new_executor/workqueue/workqueue.h b/paddle/fluid/framework/new_executor/workqueue/workqueue.h index e9c658e3b9d..2c2576528fe 100644 --- a/paddle/fluid/framework/new_executor/workqueue/workqueue.h +++ b/paddle/fluid/framework/new_executor/workqueue/workqueue.h @@ -89,7 +89,7 @@ struct WorkQueueOptions { // If you need to blocking the calling thread to wait "queue empty", set // track_task = true and set events_waiter. EventsWaiter::WaitEvent will // block the calling thread until any of events (including "queue empty") - // occured. + // occurred. bool track_task; // If you need to be noticed when a WorkQueue Destruct() , set detached = // false and set events_waiter. diff --git a/paddle/fluid/framework/paddle2cinn/build_cinn_pass.cc b/paddle/fluid/framework/paddle2cinn/build_cinn_pass.cc index 0de89aaad3b..3de702027bb 100644 --- a/paddle/fluid/framework/paddle2cinn/build_cinn_pass.cc +++ b/paddle/fluid/framework/paddle2cinn/build_cinn_pass.cc @@ -511,7 +511,7 @@ void AddCinnOpToGraph(const GraphNodeSet& cluster, ExtractOpRole(cluster)); cinn_op_desc.Flush(); auto* cinn_op_node = graph->CreateOpNode(&cinn_op_desc); - // Add new links from or to the the cinn launch op node + // Add new links from or to the cinn launch op node AddLinkToCinnOp(cluster_inputs, cluster_outputs, cinn_op_node); VLOG(4) << "Add op [" << kCinnLaunchOp << "] into graph."; diff --git a/paddle/fluid/framework/prune.cc b/paddle/fluid/framework/prune.cc index efbab83f7d0..4c95f01ae56 100644 --- a/paddle/fluid/framework/prune.cc +++ b/paddle/fluid/framework/prune.cc @@ -421,7 +421,7 @@ void PruneBackwardImpl(proto::BlockDesc* origin, proto::BlockDesc* pruned) { for (const auto& name : var_names) { if (var_map.count(name)) { // NOTE(zhiqiu): For operator in a conditional block, the related vars - // may not exist in current block, but in its futher block. + // may not exist in current block, but in its further block. *pruned_vars->Add() = var_map[name]; } } diff --git a/paddle/fluid/imperative/layer.cc b/paddle/fluid/imperative/layer.cc index e928cbb6548..76f64ab73a6 100644 --- a/paddle/fluid/imperative/layer.cc +++ b/paddle/fluid/imperative/layer.cc @@ -186,7 +186,7 @@ template static void SetForwardDataTypeOfGradVars(const NameVarMap& outs) { for (auto& var_pair : outs) { for (auto& var : var_pair.second) { - // NOTE(zhiqu): The ouput may be NULL because of pruning. + // NOTE(zhiqu): The output may be NULL because of pruning. if (var) { SetForwardDataTypeOfGradVar(var); } diff --git a/paddle/fluid/imperative/reducer.cc b/paddle/fluid/imperative/reducer.cc index 03fa46eab53..c7fd2215eb4 100644 --- a/paddle/fluid/imperative/reducer.cc +++ b/paddle/fluid/imperative/reducer.cc @@ -879,7 +879,7 @@ void Reducer::MarkVarReady(const size_t var_index, const bool is_used_var) { } // TODO(liuyuhui): If BKCL support non-blocking communication, it should be -// fixed as same as multi gpus card trainging. +// fixed as same as multi gpus card training. void Reducer::MarkGroupReady(size_t group_index) { PADDLE_ENFORCE_GE( group_index, next_group_, @@ -957,7 +957,7 @@ void Reducer::FusedAllReduceSchedule(const int run_order, Group &group, // default stream for communicating, so there exist some problems in // synchronization. And need to add a WaitComm there. // TODO(liuyuhui): If BKCL support non-blocking communication, it should be -// fixed as multi gpus card trainging. +// fixed as multi gpus card training. #ifdef PADDLE_WITH_XPU_BKCL if (platform::is_xpu_place(group.dense_tensors_[0].place())) { parallel_ctx_->WaitComm(run_order); diff --git a/paddle/fluid/inference/analysis/CMakeLists.txt b/paddle/fluid/inference/analysis/CMakeLists.txt index dab1b9f7b11..3d1a467565c 100644 --- a/paddle/fluid/inference/analysis/CMakeLists.txt +++ b/paddle/fluid/inference/analysis/CMakeLists.txt @@ -1,5 +1,5 @@ unset(analysis_deps CACHE) -set(analysis_deps # analysis_deps can be extended accross the project +set(analysis_deps # analysis_deps can be extended across the project framework_proto proto_desc graph pass paddle_inference_io executor pretty_log ir_pass_manager CACHE INTERNAL "") diff --git a/paddle/fluid/inference/analysis/ir_passes/subgraph_util.cc b/paddle/fluid/inference/analysis/ir_passes/subgraph_util.cc index 09494a36027..0c9f8d7e165 100644 --- a/paddle/fluid/inference/analysis/ir_passes/subgraph_util.cc +++ b/paddle/fluid/inference/analysis/ir_passes/subgraph_util.cc @@ -13,7 +13,7 @@ See the License for the specific language governing permissions and limitations under the License. */ /* - * This file defines the the class to partition a graph. + * This file defines the class to partition a graph. */ #include "paddle/fluid/inference/analysis/ir_passes/subgraph_util.h" diff --git a/paddle/fluid/inference/analysis/ir_passes/subgraph_util.h b/paddle/fluid/inference/analysis/ir_passes/subgraph_util.h index 621c631b853..21bfe758206 100644 --- a/paddle/fluid/inference/analysis/ir_passes/subgraph_util.h +++ b/paddle/fluid/inference/analysis/ir_passes/subgraph_util.h @@ -13,7 +13,7 @@ See the License for the specific language governing permissions and limitations under the License. */ /* - * This file defines the the class to partition a graph. + * This file defines the class to partition a graph. */ #pragma once diff --git a/paddle/fluid/inference/analysis/ir_passes/tensorrt_subgraph_pass.cc b/paddle/fluid/inference/analysis/ir_passes/tensorrt_subgraph_pass.cc index 059a9cb21e1..bc7dc9704ac 100644 --- a/paddle/fluid/inference/analysis/ir_passes/tensorrt_subgraph_pass.cc +++ b/paddle/fluid/inference/analysis/ir_passes/tensorrt_subgraph_pass.cc @@ -286,7 +286,7 @@ void TensorRtSubgraphPass::CreateTensorRTOp( // There are models with the same structure but the different parameters, // when running in the 'use_serialize' mode, there is a bug. // serialization is affected by max_batch_size, but calibration is not. - // So we use seperate engine keys in serialization and calibration. + // So we use separate engine keys in serialization and calibration. auto engine_key = GenerateEngineKey( input_names_with_id, output_names_with_id, std::to_string(0), std::to_string(max_batch_size), diff --git a/paddle/fluid/inference/api/paddle_analysis_config.h b/paddle/fluid/inference/api/paddle_analysis_config.h index 8edbc494ab8..af6cf88a322 100644 --- a/paddle/fluid/inference/api/paddle_analysis_config.h +++ b/paddle/fluid/inference/api/paddle_analysis_config.h @@ -410,7 +410,7 @@ struct PD_INFER_DECL AnalysisConfig { /// \return int The NPU device id. /// int npu_device_id() const { return npu_device_id_; } - /// \brief Get the the number of IPU device . + /// \brief Get the number of IPU device . /// /// \return int The number of IPU device. /// diff --git a/paddle/fluid/inference/tensorrt/convert/swish_op.cc b/paddle/fluid/inference/tensorrt/convert/swish_op.cc index 0df5c013d34..0b9a6917dd9 100644 --- a/paddle/fluid/inference/tensorrt/convert/swish_op.cc +++ b/paddle/fluid/inference/tensorrt/convert/swish_op.cc @@ -52,7 +52,7 @@ class SwishOpConverter : public OpConverter { PADDLE_ENFORCE_EQ( output_num, 1UL, platform::errors::InvalidArgument( - "The ouput Out's size must equal to 1 in TRT swish op. " + "The output Out's size must equal to 1 in TRT swish op. " "But received Out's size %u.", output_num)); // Get attrs diff --git a/paddle/fluid/inference/tests/api/analyzer_lexical_analysis_gru_tester.cc b/paddle/fluid/inference/tests/api/analyzer_lexical_analysis_gru_tester.cc index cca8ac2634c..141e60513eb 100644 --- a/paddle/fluid/inference/tests/api/analyzer_lexical_analysis_gru_tester.cc +++ b/paddle/fluid/inference/tests/api/analyzer_lexical_analysis_gru_tester.cc @@ -147,10 +147,10 @@ void SetInput(std::vector> *inputs, file.read(reinterpret_cast(&total_words_num), sizeof(int64_t)); LOG(INFO) << "Total words in file: " << total_words_num; size_t lods_beginning_offset = static_cast(file.tellg()); - auto words_begining_offset = + auto words_beginning_offset = lods_beginning_offset + sizeof(size_t) * total_sentences_num; auto targets_beginning_offset = - words_begining_offset + sizeof(int64_t) * total_words_num; + words_beginning_offset + sizeof(int64_t) * total_words_num; std::vector lod_full = ReadSentenceLod(file, lods_beginning_offset, total_sentences_num); @@ -158,7 +158,7 @@ void SetInput(std::vector> *inputs, size_t lods_sum = std::accumulate(lod_full.begin(), lod_full.end(), 0UL); EXPECT_EQ(lods_sum, static_cast(total_words_num)); - TensorReader words_reader(file, words_begining_offset, "words"); + TensorReader words_reader(file, words_beginning_offset, "words"); TensorReader targets_reader(file, targets_beginning_offset, "targets"); // If FLAGS_iterations is set to 0, run all batches diff --git a/paddle/fluid/operators/activation_op.h b/paddle/fluid/operators/activation_op.h index 5448ed2a4bd..8214b733f86 100644 --- a/paddle/fluid/operators/activation_op.h +++ b/paddle/fluid/operators/activation_op.h @@ -634,7 +634,7 @@ struct SquareGradGradFunctor : public BaseActivationFunctor { // TODO(dengkaipeng): double gradient calculation for Square/Sqrt need // DOut(dy) as input(not output), tensor extraction is different from -// others. Impliment extraction kernel seperately here. +// others. Impliment extraction kernel separately here. inline void ExtractDoubleGradTensorWithInputDOut( const framework::ExecutionContext& ctx, const framework::Tensor** X, const framework::Tensor** ddX, framework::Tensor** dX, diff --git a/paddle/fluid/operators/affine_grid_op.cc b/paddle/fluid/operators/affine_grid_op.cc index efaea94f26e..e311d21bb54 100644 --- a/paddle/fluid/operators/affine_grid_op.cc +++ b/paddle/fluid/operators/affine_grid_op.cc @@ -136,7 +136,7 @@ class AffineGridOpMaker : public framework::OpProtoAndCheckerMaker { .AsExtra(); AddAttr("align_corners", "(bool, default false) Whether to align the corners of input" - "and ouput.") + "and output.") .SetDefault(true); AddAttr>( "output_shape", diff --git a/paddle/fluid/operators/batch_norm_op.cc b/paddle/fluid/operators/batch_norm_op.cc index 36a0d53e052..2663a081011 100644 --- a/paddle/fluid/operators/batch_norm_op.cc +++ b/paddle/fluid/operators/batch_norm_op.cc @@ -64,7 +64,7 @@ void BatchNormOp::InferShape(framework::InferShapeContext *ctx) const { (x_dims[i] == -1) || (x_dims[i] > 0), true, platform::errors::InvalidArgument( "Each dimension of input tensor is expected to be -1 or a " - "positive number, but recieved %d. Input's shape is [%s].", + "positive number, but received %d. Input's shape is [%s].", x_dims[i], x_dims)); } diff --git a/paddle/fluid/operators/collective/c_broadcast_op.cu.cc b/paddle/fluid/operators/collective/c_broadcast_op.cu.cc index 4bed282ace8..eeae16a0d71 100644 --- a/paddle/fluid/operators/collective/c_broadcast_op.cu.cc +++ b/paddle/fluid/operators/collective/c_broadcast_op.cu.cc @@ -77,7 +77,7 @@ class CBroadcastOpCUDAKernel : public framework::OpKernel { PADDLE_ENFORCE_GPU_SUCCESS( platform::dynload::ncclBcast(out->mutable_data(place), numel, dtype, root, comm->comm(), stream)); - VLOG(3) << "rank " << comm->rank() << " invoke Bcast. recieved " + VLOG(3) << "rank " << comm->rank() << " invoke Bcast. received " << phi::product(out->dims()); } diff --git a/paddle/fluid/operators/collective/c_broadcast_op_mlu.cc b/paddle/fluid/operators/collective/c_broadcast_op_mlu.cc index d1e269fb5a4..8f07480aaab 100644 --- a/paddle/fluid/operators/collective/c_broadcast_op_mlu.cc +++ b/paddle/fluid/operators/collective/c_broadcast_op_mlu.cc @@ -62,7 +62,7 @@ class CBroadcastOPMLUKernel : public framework::OpKernel { } else { PADDLE_ENFORCE_MLU_SUCCESS(cnclBcast(out->mutable_data(place), numel, dtype, root, comm->comm(), stream)); - VLOG(3) << "rank " << comm->rank() << " invoke Bcast. recieved " + VLOG(3) << "rank " << comm->rank() << " invoke Bcast. received " << phi::product(out->dims()); } diff --git a/paddle/fluid/operators/collective/c_broadcast_op_npu.cc b/paddle/fluid/operators/collective/c_broadcast_op_npu.cc index 31961d8a246..a065e49ff72 100644 --- a/paddle/fluid/operators/collective/c_broadcast_op_npu.cc +++ b/paddle/fluid/operators/collective/c_broadcast_op_npu.cc @@ -59,7 +59,7 @@ class CBroadcastOpASCENDKernel : public framework::OpKernel { PADDLE_ENFORCE_NPU_SUCCESS(platform::dynload::HcclBroadcast( ptr, numel, dtype, (uint32_t)root, comm->comm(), stream)); - VLOG(3) << "rank " << comm->rank() << " invoke Bcast. recieved " + VLOG(3) << "rank " << comm->rank() << " invoke Bcast. received " << phi::product(out->dims()); dev_ctx->Wait(); diff --git a/paddle/fluid/operators/conv_op.h b/paddle/fluid/operators/conv_op.h index a5d888765bf..58f2eeee256 100644 --- a/paddle/fluid/operators/conv_op.h +++ b/paddle/fluid/operators/conv_op.h @@ -43,7 +43,7 @@ inline int ConvOutputSize(int input_size, int filter_size, int dilation, output_size, 0, platform::errors::InvalidArgument( "The output's size is expected to be greater than 0. " - "But recieved: output's size is %d. The output's size is computed by " + "But received: output's size is %d. The output's size is computed by " "((input_size + 2 * padding - (dilation * (filter_size - 1) + 1)) / " "stride + 1), where input_size is %d, padding is %d, " "filter_size is %d, dilation is %d, stride is %d.", @@ -60,7 +60,7 @@ inline int ConvOutputSize(int input_size, int filter_size, int dilation, output_size, 0, platform::errors::InvalidArgument( "The output's size is expected to be greater than 0. " - "But recieved: output's size is %d. The output's size is computed by " + "But received: output's size is %d. The output's size is computed by " "((input_size + padding_1 + padding_2 - (dilation * (filter_size - " "1) + 1)) / stride + 1), where input_size is %d, padding is " "(%d, %d), filter_size is %d, dilation is %d, stride is %d.", @@ -90,7 +90,7 @@ inline void UpdatePaddingAndDilation(std::vector* paddings, platform::errors::InvalidArgument( "Attribute padding's size should be the same or twice as the " "input's dimension. " - "But recieved: padding's size is %d, padding is [%s]; input's " + "But received: padding's size is %d, padding is [%s]; input's " "dimension is %d, input's shape is [%s].", paddings->size(), phi::make_ddim(*paddings), data_dims.size(), data_dims)); diff --git a/paddle/fluid/operators/conv_op_mlu.cc b/paddle/fluid/operators/conv_op_mlu.cc index 1ee772ec729..c1517dbe16f 100644 --- a/paddle/fluid/operators/conv_op_mlu.cc +++ b/paddle/fluid/operators/conv_op_mlu.cc @@ -98,7 +98,7 @@ class MLUConvOpKernel : public framework::OpKernel { output_desc.get(), GetBasePtr(&output_tensor)); if (!channel_last) { - // transpose ouput from NHWC to NCHW + // transpose output from NHWC to NCHW const std::vector perm_to_nchw = {0, 3, 1, 2}; TransposeFromMLUTensor(ctx, perm_to_nchw, &output_tensor, output, false /*need_reshape_or_alloc*/); diff --git a/paddle/fluid/operators/ctc_align_op.cu b/paddle/fluid/operators/ctc_align_op.cu index b1f2e61ef39..ba90c677570 100644 --- a/paddle/fluid/operators/ctc_align_op.cu +++ b/paddle/fluid/operators/ctc_align_op.cu @@ -26,19 +26,19 @@ __global__ void MergeAndDelCudaKernel(const int64_t num_token, const T* tokens, const size_t num_seq, size_t* lod0, const int blank, const int merge_repeated, size_t* out_lod0, T* output) { - int ouput_idx = 0; + int output_idx = 0; out_lod0[0] = 0; for (int i = 0; i < num_seq; ++i) { T pre_token = -1; for (int j = lod0[i]; j < lod0[i + 1]; ++j) { if (tokens[j] != blank && !(merge_repeated && tokens[j] == pre_token)) { - output[ouput_idx] = tokens[j]; - ++ouput_idx; + output[output_idx] = tokens[j]; + ++output_idx; } pre_token = tokens[j]; } - out_lod0[i + 1] = ouput_idx; + out_lod0[i + 1] = output_idx; } } diff --git a/paddle/fluid/operators/deformable_psroi_pooling_op.cu b/paddle/fluid/operators/deformable_psroi_pooling_op.cu index 448f67a4bad..873950b2d2f 100644 --- a/paddle/fluid/operators/deformable_psroi_pooling_op.cu +++ b/paddle/fluid/operators/deformable_psroi_pooling_op.cu @@ -200,7 +200,7 @@ class DeformablePSROIPoolCUDAKernel : public framework::OpKernel { num_rois, out->dims()[0], platform::errors::InvalidArgument( "The number of Input(ROIs) should be same with the number of " - "Ouput(Output), but received ROIs number is:%d, Output number " + "Output(Output), but received ROIs number is:%d, Output number " "is:%d.", num_rois, out->dims()[0])); const int count = num_rois * output_dim * pooled_height * pooled_width; diff --git a/paddle/fluid/operators/deformable_psroi_pooling_op.h b/paddle/fluid/operators/deformable_psroi_pooling_op.h index 51a0fe4172c..3deabce54ed 100644 --- a/paddle/fluid/operators/deformable_psroi_pooling_op.h +++ b/paddle/fluid/operators/deformable_psroi_pooling_op.h @@ -175,7 +175,7 @@ class DeformablePSROIPoolCPUKernel : public framework::OpKernel { num_rois, out->dims()[0], platform::errors::InvalidArgument( "The number of Input(ROIs) should be same with the number of " - "Ouput(Output), but received ROIs number is:%d, Output number " + "Output(Output), but received ROIs number is:%d, Output number " "is:%d.", num_rois, out->dims()[0])); framework::Tensor roi_batch_id_list; diff --git a/paddle/fluid/operators/detection/matrix_nms_op.cc b/paddle/fluid/operators/detection/matrix_nms_op.cc index 713c2dc7fe9..3353739b01b 100644 --- a/paddle/fluid/operators/detection/matrix_nms_op.cc +++ b/paddle/fluid/operators/detection/matrix_nms_op.cc @@ -385,7 +385,7 @@ independently for each class. The outputs is a 2-D LoDTenosr, for each image, the offsets in first dimension of LoDTensor are called LoD, the number of offset is N + 1, where N is the batch size. If LoD[i + 1] - LoD[i] == 0, means there is no detected bbox for this image. Now this operator has one more -ouput, which is RoisNum. The size of RoisNum is N, RoisNum[i] means the number of +output, which is RoisNum. The size of RoisNum is N, RoisNum[i] means the number of detected bbox for this image. For more information on Matrix NMS, please refer to: diff --git a/paddle/fluid/operators/dropout_op_npu.cc b/paddle/fluid/operators/dropout_op_npu.cc index 07b3b538116..104ab1b5046 100644 --- a/paddle/fluid/operators/dropout_op_npu.cc +++ b/paddle/fluid/operators/dropout_op_npu.cc @@ -54,7 +54,7 @@ class DropoutNPUKernel : public framework::OpKernel { return; } - // only achive the default `upscale_in_train` method + // only achieve the default `upscale_in_train` method if (!is_test) { Tensor tmp_x(x->dtype()); Tensor tmp_out(out->dtype()); diff --git a/paddle/fluid/operators/elementwise/elementwise_mlu.h b/paddle/fluid/operators/elementwise/elementwise_mlu.h index 156cea81c0f..ff1e12103be 100644 --- a/paddle/fluid/operators/elementwise/elementwise_mlu.h +++ b/paddle/fluid/operators/elementwise/elementwise_mlu.h @@ -165,7 +165,7 @@ template void MLUUnary(const framework::ExecutionContext& ctx, cnnlComputationPreference_t prefer, const cnnlTensorDescriptor_t input_desc, const void* input, - const cnnlTensorDescriptor_t ouput_desc, void* output); + const cnnlTensorDescriptor_t output_desc, void* output); template <> inline void MLUUnary(const framework::ExecutionContext& ctx, diff --git a/paddle/fluid/operators/fc_op.cc b/paddle/fluid/operators/fc_op.cc index 33518953004..6e646f0d4bf 100644 --- a/paddle/fluid/operators/fc_op.cc +++ b/paddle/fluid/operators/fc_op.cc @@ -75,7 +75,7 @@ class FCOp : public framework::OperatorWithKernel { platform::errors::InvalidArgument( "The attribute in_num_col_dims used to flatten Input to " "a 2-D tensor, is expected to be less than the number of " - "Input's dimensions. But recieved in_num_col_dims is %d, " + "Input's dimensions. But received in_num_col_dims is %d, " "the number of Input's dimensions is %d, Input's shape is %s.", in_num_col_dims, in_dims.size(), in_dims)); @@ -93,7 +93,7 @@ class FCOp : public framework::OperatorWithKernel { in_dims.size() >= 2 && in_dims.size() <= 4, true, platform::errors::Unimplemented( "The Input of fc is expected to be a 2-D, 3-D or 4-D tensor when " - "use_mkldnn is set. But recieved the number of Input's " + "use_mkldnn is set. But received the number of Input's " "dimensions is %d, Input's shape is %s.", in_dims.size(), in_dims)); } diff --git a/paddle/fluid/operators/fc_op.h b/paddle/fluid/operators/fc_op.h index 6d3b531ce0a..47c71286035 100644 --- a/paddle/fluid/operators/fc_op.h +++ b/paddle/fluid/operators/fc_op.h @@ -36,7 +36,7 @@ inline void FCOutputSize(const framework::DDim& in_dims, in_mat_dims[1], w_dims0, platform::errors::InvalidArgument( "The input's second dimension and weight's first dimension is " - "expected to be the same. But recieved input's second dimension is " + "expected to be the same. But received input's second dimension is " "%d, input's shape is %s; weight's first dimension is %d, weight's " "shape is %s.", in_mat_dims[1], in_mat_dims, w_dims0, diff --git a/paddle/fluid/operators/fill_constant_op.cc b/paddle/fluid/operators/fill_constant_op.cc index 2e924da283a..07593a70f05 100644 --- a/paddle/fluid/operators/fill_constant_op.cc +++ b/paddle/fluid/operators/fill_constant_op.cc @@ -32,7 +32,7 @@ class FillConstantOp : public framework::OperatorWithKernel { shape[i], 0, platform::errors::InvalidArgument( "Each value of attribute 'shape' is expected to be no less " - "than 0. But recieved: shape[%u] = %d; shape = [%s].", + "than 0. But received: shape[%u] = %d; shape = [%s].", i, shape[i], phi::make_ddim(shape))); } } diff --git a/paddle/fluid/operators/fold_op.cc b/paddle/fluid/operators/fold_op.cc index 92f59e118c3..9c9183c8faf 100644 --- a/paddle/fluid/operators/fold_op.cc +++ b/paddle/fluid/operators/fold_op.cc @@ -76,47 +76,47 @@ class FoldOp : public framework::OperatorWithKernel { PADDLE_ENFORCE_GT(kernel_height, 0, platform::errors::InvalidArgument( "The `kernel_sizes` should be greater than zero, " - "but recieved kernel_height: %d kernel_width: %d.", + "but received kernel_height: %d kernel_width: %d.", kernel_sizes[0], kernel_sizes[1])); PADDLE_ENFORCE_GT(kernel_width, 0, platform::errors::InvalidArgument( "The `kernel_sizes` should be greater than zero, " - "but recieved kernel_height: %d kernel_width: %d.", + "but received kernel_height: %d kernel_width: %d.", kernel_sizes[0], kernel_sizes[1])); // check strides PADDLE_ENFORCE_GT(stride_height, 0, platform::errors::InvalidArgument( "The `strides` should be greater than zero, " - "but recieved strides_height: %d strides_width: %d.", + "but received strides_height: %d strides_width: %d.", strides[0], strides[1])); PADDLE_ENFORCE_GT(stride_width, 0, platform::errors::InvalidArgument( "The `strides` should be greater than zero, " - "but recieved strides_height: %d strides_width: %d.", + "but received strides_height: %d strides_width: %d.", strides[0], strides[1])); // check dilations PADDLE_ENFORCE_GT(output_height, 1, platform::errors::InvalidArgument( "The `output_height` should be greater than one, " - "but recieved output_height: %d .", + "but received output_height: %d .", output_height)); PADDLE_ENFORCE_GT(output_width, 1, platform::errors::InvalidArgument( "The `output_width` should be greater than one, " - "but recieved output_width: %d .", + "but received output_width: %d .", output_width)); // check output size PADDLE_ENFORCE_GT( dilation_height, 0, platform::errors::InvalidArgument( "The `dilations` should be greater than zero, " - "but recieved dilations_height: %d dilations_width: %d.", + "but received dilations_height: %d dilations_width: %d.", dilations[0], dilations[1])); PADDLE_ENFORCE_GT( dilation_width, 0, platform::errors::InvalidArgument( "The `dilations` should be greater than zero, " - "but recieved dilations_height: %d dilations_width: %d.", + "but received dilations_height: %d dilations_width: %d.", dilations[0], dilations[1])); std::vector out_dims; diff --git a/paddle/fluid/operators/fused/conv_fusion_op.cc b/paddle/fluid/operators/fused/conv_fusion_op.cc index e60fc44e9a6..671e94061cb 100644 --- a/paddle/fluid/operators/fused/conv_fusion_op.cc +++ b/paddle/fluid/operators/fused/conv_fusion_op.cc @@ -80,7 +80,7 @@ class Conv2DFusionOp : public operators::ConvOp { data_format, "NHWC", platform::errors::PermissionDenied( "Operator(Conv2DFusion) only supports data format of " - "channel first (NCHW) now. But recieved: data_format = '%s'.", + "channel first (NCHW) now. But received: data_format = '%s'.", data_format)); std::vector output_shape = ComputeOutputShape(ctx); @@ -113,7 +113,7 @@ class Conv2DFusionOp : public operators::ConvOp { split_channels_sum, output_shape[1], platform::errors::InvalidArgument( "The sum of Attr(split_channels) is expected to be equal to the " - "total output channels. But recieved: the sum of " + "total output channels. But received: the sum of " "Attr(split_channels) = %d, the total output channels = %d.", split_channels_sum, output_shape[1])); diff --git a/paddle/fluid/operators/fused/conv_fusion_op.cu b/paddle/fluid/operators/fused/conv_fusion_op.cu index 5dbf4fb88b2..8191c85f2a1 100644 --- a/paddle/fluid/operators/fused/conv_fusion_op.cu +++ b/paddle/fluid/operators/fused/conv_fusion_op.cu @@ -130,7 +130,7 @@ class CUDNNConvFusionOpKernel : public framework::OpKernel { default: PADDLE_THROW(platform::errors::PermissionDenied( "Operator Conv2DFusion expects Input to be a 4-D or 5-D Tensor. " - "But recieved the actual dimension = %d, shape = [%s].", + "But received the actual dimension = %d, shape = [%s].", rank, transformed_input_channel.dims())); } @@ -355,7 +355,7 @@ class CUDNNConvFusionOpKernel : public framework::OpKernel { workspace_size_in_bytes, workspace_size_limit, platform::errors::InvalidArgument( "The actual workspace size to be allocated for cuDNN is expected " - "to be less than the limit. But recieved: the actual workspace " + "to be less than the limit. But received: the actual workspace " "size = %d, limit = %d.", workspace_size_in_bytes, workspace_size_limit)); @@ -414,7 +414,7 @@ class CUDNNConvFusionOpKernel : public framework::OpKernel { } else { // TODO(qingiqng): do copy when batch size large than 1 PADDLE_THROW(platform::errors::Unimplemented( - "Input with batch size greater than 1 is unsupported. The recieved " + "Input with batch size greater than 1 is unsupported. The received " "batch size is %d, Input's shape is [%s].", x_dims[0], phi::make_ddim(x_dims))); } diff --git a/paddle/fluid/operators/fused/cudnn_bn_stats_finalize.cu.h b/paddle/fluid/operators/fused/cudnn_bn_stats_finalize.cu.h index 74cc92eb8ab..4b3ed56890e 100644 --- a/paddle/fluid/operators/fused/cudnn_bn_stats_finalize.cu.h +++ b/paddle/fluid/operators/fused/cudnn_bn_stats_finalize.cu.h @@ -40,7 +40,7 @@ struct BNStatsFinalizeArgs { PADDLE_ENFORCE_EQ( param_shape.size(), 4U, platform::errors::InvalidArgument( - "The size of param_shape is expected to 4. But recieved " + "The size of param_shape is expected to 4. But received " "param_shape's size is %d, param_shape is [%s].", param_shape.size(), phi::make_ddim(param_shape))); diff --git a/paddle/fluid/operators/fused/cudnn_norm_conv.cu.h b/paddle/fluid/operators/fused/cudnn_norm_conv.cu.h index f63fe4b96cb..b32f2e40933 100644 --- a/paddle/fluid/operators/fused/cudnn_norm_conv.cu.h +++ b/paddle/fluid/operators/fused/cudnn_norm_conv.cu.h @@ -45,13 +45,13 @@ struct NormConvolutionArgs { PADDLE_ENFORCE_EQ( input_shape.size(), 4U, platform::errors::InvalidArgument( - "The size of input_shape is expected to 4. But recieved " + "The size of input_shape is expected to 4. But received " "input_shape's size is %d, input_shape is [%s].", input_shape.size(), phi::make_ddim(input_shape))); PADDLE_ENFORCE_EQ( filter_shape.size(), 4U, platform::errors::InvalidArgument( - "The size of filter_shape is expected to 4. But recieved " + "The size of filter_shape is expected to 4. But received " "filter_shape's size is %d, filter_shape is [%s].", filter_shape.size(), phi::make_ddim(filter_shape))); PADDLE_ENFORCE_EQ(filter_shape[1] == filter_shape[2] && @@ -59,20 +59,20 @@ struct NormConvolutionArgs { true, platform::errors::InvalidArgument( "The filter_shape is expected to store as nhwc, and " - "h = w = 1 or 3. But recieved filter_shape is [%s].", + "h = w = 1 or 3. But received filter_shape is [%s].", phi::make_ddim(filter_shape))); PADDLE_ENFORCE_EQ((filter_shape[0] % 32 == 0 && filter_shape[3] % 8 == 0), true, platform::errors::InvalidArgument( "The input channel is expected to be multiple of 8, " "and the output channel is expected to be multiple " - "of 32. But recieved input channel is %d, output " + "of 32. But received input channel is %d, output " "channel is %d.", filter_shape[3], filter_shape[0])); PADDLE_ENFORCE_EQ( output_shape.size(), 4U, platform::errors::InvalidArgument( - "The size of output_shape is expected to 4. But recieved " + "The size of output_shape is expected to 4. But received " "filter_shape's size is %d, filter_shape is [%s].", output_shape.size(), phi::make_ddim(output_shape))); is_support = IsSupport(ctx, filter_shape, stride, dilation, group); @@ -83,7 +83,7 @@ struct NormConvolutionArgs { "compatiblity greater than or equal to 70 and the kernel size " "must be equal to 1 or 3. When the kernel size is 1, " "the stride must be 1 if the compatiblity is equal to 70. " - "Besides, the dilation and group must be equal to 1. But recieved " + "Besides, the dilation and group must be equal to 1. But received " "compatiblity is %d, kernel size is %d, stride is %d, " "dilation is %d, group is %d", ctx.GetComputeCapability(), filter_shape[1], stride, dilation, diff --git a/paddle/fluid/operators/fused/cudnn_scale_bias_add_relu.cu.h b/paddle/fluid/operators/fused/cudnn_scale_bias_add_relu.cu.h index 9d3090a7179..c8588b0c02e 100644 --- a/paddle/fluid/operators/fused/cudnn_scale_bias_add_relu.cu.h +++ b/paddle/fluid/operators/fused/cudnn_scale_bias_add_relu.cu.h @@ -43,19 +43,19 @@ struct ScaleBiasAddReluArgs { PADDLE_ENFORCE_EQ( data_shape.size(), 4U, platform::errors::InvalidArgument( - "The size of data_shape is expected to 4. But recieved " + "The size of data_shape is expected to 4. But received " "data_shape's size is %d, data_shape is [%s].", data_shape.size(), phi::make_ddim(data_shape))); PADDLE_ENFORCE_EQ( param_shape.size(), 4U, platform::errors::InvalidArgument( - "The size of param_shape is expected to 4. But recieved " + "The size of param_shape is expected to 4. But received " "param_shape's size is %d, param_shape is [%s].", param_shape.size(), phi::make_ddim(param_shape))); PADDLE_ENFORCE_EQ( bitmask_shape.size(), 3U, platform::errors::InvalidArgument( - "The size of bitmask_shape is expected to 3. But recieved " + "The size of bitmask_shape is expected to 3. But received " "bitmask_shape's size is %d, bitmask_shape is [%s].", bitmask_shape.size(), phi::make_ddim(bitmask_shape))); diff --git a/paddle/fluid/operators/fused/fused_fc_elementwise_layernorm_op.cc b/paddle/fluid/operators/fused/fused_fc_elementwise_layernorm_op.cc index 27dae277516..1b5b074ef1c 100644 --- a/paddle/fluid/operators/fused/fused_fc_elementwise_layernorm_op.cc +++ b/paddle/fluid/operators/fused/fused_fc_elementwise_layernorm_op.cc @@ -76,7 +76,7 @@ class FusedFCElementwiseLayerNormOp : public framework::OperatorWithKernel { platform::errors::InvalidArgument( "The attribute x_num_col_dims used to flatten input X to " "a 2-D tensor, is expected to be less than the number of " - "input X's dimensions. But recieved x_num_col_dims is %d, " + "input X's dimensions. But received x_num_col_dims is %d, " "the number of input X's dimensions is %d, input X's shape is %s.", x_num_col_dims, x_dims.size(), x_dims)); @@ -85,7 +85,7 @@ class FusedFCElementwiseLayerNormOp : public framework::OperatorWithKernel { x_mat_dims[1], w_dims[0], platform::errors::InvalidArgument( "The input's second dimension and weight's first dimension is " - "expected to be the same. But recieved input's second dimension is " + "expected to be the same. But received input's second dimension is " "%d, input's shape is %s; weight's first dimension is %d, weight's " "shape is %s.", x_mat_dims[1], x_mat_dims, w_dims[0], w_dims)); @@ -100,7 +100,7 @@ class FusedFCElementwiseLayerNormOp : public framework::OperatorWithKernel { PADDLE_ENFORCE_EQ(phi::make_ddim(fc_out_dims), y_dims, platform::errors::InvalidArgument( "The output's shape of fc is expected to be equal to " - "that of input Y. But recieved output's shape of fc " + "that of input Y. But received output's shape of fc " "is %s, input Y's shape is %s.", phi::make_ddim(fc_out_dims), y_dims)); @@ -110,7 +110,7 @@ class FusedFCElementwiseLayerNormOp : public framework::OperatorWithKernel { platform::errors::InvalidArgument( "The attribute begin_norm_axis used to flatten input Y to a 2-D " "tensor, is expected to be less than the number of input Y's " - "dimensions. But recieved begin_norm_axis is %d, the number of " + "dimensions. But received begin_norm_axis is %d, the number of " "input Y's dimensions is %d, input Y's shape is %s.", begin_norm_axis, y_dims.size(), y_dims)); @@ -122,7 +122,7 @@ class FusedFCElementwiseLayerNormOp : public framework::OperatorWithKernel { PADDLE_ENFORCE_EQ(scale_dims.size(), 1, platform::errors::InvalidArgument( "The input Scale is expected to be an 1-D tensor. " - "But recieved the number of input Scale's " + "But received the number of input Scale's " "dimensions is %d, input Scale's shape is %s.", scale_dims.size(), scale_dims)); @@ -132,7 +132,7 @@ class FusedFCElementwiseLayerNormOp : public framework::OperatorWithKernel { platform::errors::InvalidArgument( "The first dimension of input Scale is expected to be equal to " "the second dimension of input Y after flattened. " - "But recieved the first dimension of input Scale is %d, input " + "But received the first dimension of input Scale is %d, input " "Scale's shape is %s; the second dimension of flattened input " "Y is %d, input Y's shape is %s, flattened axis is %d.", scale_dims[0], scale_dims, dim_1, y_dims, begin_norm_axis)); @@ -144,7 +144,7 @@ class FusedFCElementwiseLayerNormOp : public framework::OperatorWithKernel { bias1_dims.size(), 1, platform::errors::InvalidArgument( "The input Bias1 is expected to be an 1-D tensor. " - "But recieved the number of input Bias1's dimension is %d, " + "But received the number of input Bias1's dimension is %d, " "input Bias1's shape is %s.", bias1_dims.size(), bias1_dims)); @@ -154,7 +154,7 @@ class FusedFCElementwiseLayerNormOp : public framework::OperatorWithKernel { platform::errors::InvalidArgument( "The first dimension of input Bias1 is expected to be equal to " "the second dimension of input Y after flattened. " - "But recieved the first dimension of input Bias1 is %d, input " + "But received the first dimension of input Bias1 is %d, input " "Bias1's shape is %s; the second dimension of flatten input " "Y is %d, input Y's shape is %s, flattened axis is %d.", bias1_dims[0], bias1_dims, dim_1, y_dims, begin_norm_axis)); diff --git a/paddle/fluid/operators/fused/fusion_group_op.cc b/paddle/fluid/operators/fused/fusion_group_op.cc index 738e0690815..1ebbdf792df 100644 --- a/paddle/fluid/operators/fused/fusion_group_op.cc +++ b/paddle/fluid/operators/fused/fusion_group_op.cc @@ -52,7 +52,7 @@ class FusionGroupOp : public framework::OperatorWithKernel { x_dims[0], x_dims[i], platform::errors::InvalidArgument( "All the inputs' dims is expected to be the same. " - "But recieved [%s] (name: %s) vs [%s] (name: %s).", + "But received [%s] (name: %s) vs [%s] (name: %s).", x_dims[0], input_names[0], x_dims[i], input_names[i])); } std::vector out_dims; diff --git a/paddle/fluid/operators/inverse_op.cc b/paddle/fluid/operators/inverse_op.cc index 8c1fd34ae87..f5b817a0e11 100644 --- a/paddle/fluid/operators/inverse_op.cc +++ b/paddle/fluid/operators/inverse_op.cc @@ -33,21 +33,21 @@ class InverseOp : public framework::OperatorWithKernel { input_rank, 2, platform::errors::InvalidArgument( "The dimension of Input(Input) is expected to be no less than 2. " - "But recieved: Input(Input)'s dimension = %d, shape = [%s].", + "But received: Input(Input)'s dimension = %d, shape = [%s].", input_rank, input_dims)); for (int64_t i = 0; i < input_rank; ++i) { PADDLE_ENFORCE_EQ( (input_dims[i] == -1) || (input_dims[i] > 0), true, platform::errors::InvalidArgument( "Each dimension of input tensor is expected to be -1 or a " - "positive number, but recieved %d. Input's shape is [%s].", + "positive number, but received %d. Input's shape is [%s].", input_dims[i], input_dims)); } if (input_dims[input_rank - 2] > 0 && input_dims[input_rank - 1] > 0) { PADDLE_ENFORCE_EQ(input_dims[input_rank - 2], input_dims[input_rank - 1], platform::errors::InvalidArgument( "The last two dimensions are expected to be equal. " - "But recieved: %d and %d; " + "But received: %d and %d; " "Input(Input)'s shape = [%s].", input_dims[input_rank - 2], input_dims[input_rank - 1], input_dims)); diff --git a/paddle/fluid/operators/lod_reset_op.h b/paddle/fluid/operators/lod_reset_op.h index 86327a4f2c1..642c8bcd9ae 100644 --- a/paddle/fluid/operators/lod_reset_op.h +++ b/paddle/fluid/operators/lod_reset_op.h @@ -77,7 +77,7 @@ class LoDResetKernel : public framework::OpKernel { platform::errors::InvalidArgument( "The last value of 'Target LoD''s last level LoD should be equal " "to the first dimension of Input(X). But received the 'Target LoD' " - "is %s, Input(X)'s shape is is %s.", + "is %s, Input(X)'s shape is %s.", phi::make_ddim(level0), in->dims())); for (size_t i = 0; i < level0.size() - 1; ++i) { PADDLE_ENFORCE_GE(level0[i + 1], level0[i], diff --git a/paddle/fluid/operators/math/cross_entropy.h b/paddle/fluid/operators/math/cross_entropy.h index e339be06d69..da7340e4eb0 100644 --- a/paddle/fluid/operators/math/cross_entropy.h +++ b/paddle/fluid/operators/math/cross_entropy.h @@ -38,7 +38,7 @@ struct TolerableValue { // NOTE(dzh): float16 value clip behave different. // 1. Our ValueClipping has a hardcore threshold 1e20 // for float number. 1e20 will resulting in overflow in float16. -// 2. float16 should expose the the real number overflow to python. +// 2. float16 should expose the real number overflow to python. // because mixed-training depends the inf/nan value to determine // if the scale value will be adjusted. // Also. In standard implementation of cross entropy, other diff --git a/paddle/fluid/operators/math/selected_rows_functor.cc b/paddle/fluid/operators/math/selected_rows_functor.cc index f77287826ff..a880afb0e9b 100644 --- a/paddle/fluid/operators/math/selected_rows_functor.cc +++ b/paddle/fluid/operators/math/selected_rows_functor.cc @@ -32,7 +32,7 @@ struct SelectedRowsAdd { PADDLE_ENFORCE_EQ( in1_height, input2.height(), platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, input2.height())); output->set_height(in1_height); @@ -56,27 +56,27 @@ struct SelectedRowsAdd { in1_row_numel, in2_value.numel() / in2_rows.size(), platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, in2_value.numel() / in2_rows.size())); PADDLE_ENFORCE_EQ( in1_row_numel, out_value->numel() / out_rows.size(), platform::errors::InvalidArgument( "The input and oupput width must be equal." - "But recieved input width = [%d], output width = [%d]", + "But received input width = [%d], output width = [%d]", in1_row_numel, out_value->numel() / out_rows.size())); auto in1_place = input1.place(); PADDLE_ENFORCE_EQ(platform::is_cpu_place(in1_place), true, platform::errors::InvalidArgument( - "The running enviroment is not on the CPU place.")); + "The running environment is not on the CPU place.")); auto in2_place = input2.place(); PADDLE_ENFORCE_EQ(platform::is_cpu_place(in2_place), true, platform::errors::InvalidArgument( - "The running enviroment is not on the CPU place.")); + "The running environment is not on the CPU place.")); auto out_place = context.GetPlace(); PADDLE_ENFORCE_EQ(platform::is_cpu_place(out_place), true, platform::errors::InvalidArgument( - "The running enviroment is not on the CPU place.")); + "The running environment is not on the CPU place.")); auto* out_data = out_value->data(); auto* in1_data = in1_value.data(); @@ -103,14 +103,14 @@ struct SelectedRowsAddTensor { PADDLE_ENFORCE_EQ( in1_height, in2_dims[0], platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, in2_dims[0])); PADDLE_ENFORCE_EQ( in1_height, out_dims[0], platform::errors::InvalidArgument( "The input and output height must be equal." - "But recieved input height = [%d], output height = [%d]", + "But received input height = [%d], output height = [%d]", in1_height, out_dims[0])); auto& in1_value = input1.value(); @@ -121,13 +121,13 @@ struct SelectedRowsAddTensor { in1_row_numel, input2.numel() / in1_height, platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, input2.numel() / in1_height)); PADDLE_ENFORCE_EQ( in1_row_numel, output->numel() / in1_height, platform::errors::InvalidArgument( "The input and output width must be equal." - "But recieved input width = [%d], output width = [%d]", + "But received input width = [%d], output width = [%d]", in1_row_numel, output->numel() / in1_height)); phi::funcs::SetConstant functor; @@ -161,7 +161,7 @@ struct SelectedRowsAddTo { PADDLE_ENFORCE_EQ( in1_height, input2->height(), platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, input2->height())); @@ -178,11 +178,11 @@ struct SelectedRowsAddTo { auto in1_place = input1.place(); PADDLE_ENFORCE_EQ(platform::is_cpu_place(in1_place), true, platform::errors::InvalidArgument( - "The running enviroment is not on the CPU place.")); + "The running environment is not on the CPU place.")); auto in2_place = input2->place(); PADDLE_ENFORCE_EQ(platform::is_cpu_place(in2_place), true, platform::errors::InvalidArgument( - "The running enviroment is not on the CPU place.")); + "The running environment is not on the CPU place.")); auto* in1_data = in1_value.data(); auto* in2_data = in2_value->data(); @@ -211,7 +211,7 @@ struct SelectedRowsSumTo { PADDLE_ENFORCE_EQ(in1_height, input2->height(), platform::errors::InvalidArgument( "The two inputs height must be equal." - "But recieved first input height = [%d], second " + "But received first input height = [%d], second " "input height = [%d]", in1_height, input2->height())); } @@ -253,7 +253,7 @@ struct SelectedRowsAddToTensor { PADDLE_ENFORCE_EQ( in1_height, in2_dims[0], platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, in2_dims[0])); @@ -265,7 +265,7 @@ struct SelectedRowsAddToTensor { in1_row_numel, input2->numel() / in1_height, platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, input2->numel() / in1_height)); auto* in1_data = in1_value.data(); @@ -293,7 +293,7 @@ struct SelectedRowsAddToTensor { PADDLE_ENFORCE_EQ( in1_height, in2_dims[0], platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, in2_dims[0])); @@ -305,7 +305,7 @@ struct SelectedRowsAddToTensor { in1_row_numel, input2->numel() / in1_height, platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, input2->numel() / in1_height)); auto* in1_data = in1_value.data(); @@ -842,7 +842,7 @@ struct UpdateToTensor { PADDLE_ENFORCE_EQ( in1_height, in2_dims[0], platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, in2_dims[0])); @@ -854,7 +854,7 @@ struct UpdateToTensor { in1_row_numel, input2->numel() / in1_height, platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, input2->numel() / in1_height)); auto* in1_data = in1_value.data(); diff --git a/paddle/fluid/operators/math/selected_rows_functor.cu b/paddle/fluid/operators/math/selected_rows_functor.cu index 542d4c97843..db5c66d3197 100644 --- a/paddle/fluid/operators/math/selected_rows_functor.cu +++ b/paddle/fluid/operators/math/selected_rows_functor.cu @@ -33,7 +33,7 @@ struct SelectedRowsAdd { PADDLE_ENFORCE_EQ( in1_height, input2.height(), platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, input2.height())); output->set_height(in1_height); @@ -57,13 +57,13 @@ struct SelectedRowsAdd { in1_row_numel, in2_value.numel() / in2_rows.size(), platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, in2_value.numel() / in2_rows.size())); PADDLE_ENFORCE_EQ( in1_row_numel, out_value->numel() / out_rows.size(), platform::errors::InvalidArgument( "The input and oupput width must be equal." - "But recieved input width = [%d], output width = [%d]", + "But received input width = [%d], output width = [%d]", in1_row_numel, out_value->numel() / out_rows.size())); auto* out_data = out_value->data(); @@ -72,15 +72,15 @@ struct SelectedRowsAdd { auto in1_place = input1.place(); PADDLE_ENFORCE_EQ(platform::is_gpu_place(in1_place), true, platform::errors::InvalidArgument( - "The running enviroment is not on the GPU place.")); + "The running environment is not on the GPU place.")); auto in2_place = input2.place(); PADDLE_ENFORCE_EQ(platform::is_gpu_place(in2_place), true, platform::errors::InvalidArgument( - "The running enviroment is not on the GPU place.")); + "The running environment is not on the GPU place.")); auto out_place = context.GetPlace(); PADDLE_ENFORCE_EQ(platform::is_gpu_place(out_place), true, platform::errors::InvalidArgument( - "The running enviroment is not on the GPU place.")); + "The running environment is not on the GPU place.")); memory::Copy(out_place, out_data, in1_place, in1_data, in1_value.numel() * sizeof(T), context.stream()); @@ -126,13 +126,13 @@ struct SelectedRowsAddTensor { in1_height, in2_dims[0], platform::errors::InvalidArgument( "The two inputs height must be equal." - "But recieved first input height = [%d], first input height = [%d]", + "But received first input height = [%d], first input height = [%d]", in1_height, in2_dims[0])); PADDLE_ENFORCE_EQ( in1_height, out_dims[0], platform::errors::InvalidArgument( "The input and output height must be equal." - "But recieved input height = [%d], output height = [%d]", + "But received input height = [%d], output height = [%d]", in1_height, out_dims[0])); auto& in1_value = input1.value(); @@ -143,13 +143,13 @@ struct SelectedRowsAddTensor { in1_row_numel, input2.numel() / in1_height, platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, input2.numel() / in1_height)); PADDLE_ENFORCE_EQ( in1_row_numel, output->numel() / in1_height, platform::errors::InvalidArgument( "The input and output width must be equal." - "But recieved input width = [%d], output width = [%d]", + "But received input width = [%d], output width = [%d]", in1_row_numel, output->numel() / in1_height)); auto* in1_data = in1_value.data(); @@ -186,13 +186,13 @@ struct SelectedRowsAddTensor { in1_height, in2_dims[0], platform::errors::InvalidArgument( "The two inputs height must be equal." - "But recieved first input height = [%d], first input height = [%d]", + "But received first input height = [%d], first input height = [%d]", in1_height, in2_dims[0])); PADDLE_ENFORCE_EQ( in1_height, out_dims[0], platform::errors::InvalidArgument( "The input and output height must be equal." - "But recieved input height = [%d], output height = [%d]", + "But received input height = [%d], output height = [%d]", in1_height, out_dims[0])); auto& in1_value = input1.value(); @@ -203,13 +203,13 @@ struct SelectedRowsAddTensor { in1_row_numel, input2.numel() / in1_height, platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, input2.numel() / in1_height)); PADDLE_ENFORCE_EQ( in1_row_numel, output->numel() / in1_height, platform::errors::InvalidArgument( "The input and output width must be equal." - "But recieved input width = [%d], output width = [%d]", + "But received input width = [%d], output width = [%d]", in1_row_numel, output->numel() / in1_height)); auto* in1_data = in1_value.data(); @@ -254,7 +254,7 @@ struct SelectedRowsAddTo { PADDLE_ENFORCE_EQ( in1_height, input2->height(), platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, input2->height())); @@ -273,11 +273,11 @@ struct SelectedRowsAddTo { auto in1_place = input1.place(); PADDLE_ENFORCE_EQ(platform::is_gpu_place(in1_place), true, platform::errors::InvalidArgument( - "The running enviroment is not on the GPU place.")); + "The running environment is not on the GPU place.")); auto in2_place = input2->place(); PADDLE_ENFORCE_EQ(platform::is_gpu_place(in1_place), true, platform::errors::InvalidArgument( - "The running enviroment is not on the GPU place.")); + "The running environment is not on the GPU place.")); auto* in1_data = in1_value.data(); auto* in2_data = in2_value->data(); @@ -322,7 +322,7 @@ struct SelectedRowsAddToTensor { PADDLE_ENFORCE_EQ( in1_height, in2_dims[0], platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, in2_dims[0])); @@ -334,7 +334,7 @@ struct SelectedRowsAddToTensor { in1_row_numel, input2->numel() / in1_height, platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, input2->numel() / in1_height)); auto* in1_data = in1_value.data(); @@ -359,7 +359,7 @@ struct SelectedRowsAddToTensor { PADDLE_ENFORCE_EQ( in1_height, in2_dims[0], platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, in2_dims[0])); @@ -371,7 +371,7 @@ struct SelectedRowsAddToTensor { in1_row_numel, input2->numel() / in1_height, platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, input2->numel() / in1_height)); auto* in1_data = in1_value.data(); @@ -675,7 +675,7 @@ struct UpdateToTensor { PADDLE_ENFORCE_EQ( in1_height, in2_dims[0], platform::errors::InvalidArgument("The two inputs height must be equal." - "But recieved first input height = " + "But received first input height = " "[%d], second input height = [%d]", in1_height, in2_dims[0])); @@ -687,7 +687,7 @@ struct UpdateToTensor { in1_row_numel, input2->numel() / in1_height, platform::errors::InvalidArgument( "The two inputs width must be equal." - "But recieved first input width = [%d], second input width = [%d]", + "But received first input width = [%d], second input width = [%d]", in1_row_numel, input2->numel() / in1_height)); auto* in1_data = in1_value.template data(); diff --git a/paddle/fluid/operators/metrics/accuracy_op.cc b/paddle/fluid/operators/metrics/accuracy_op.cc index 32ef0521198..ed58c90e170 100644 --- a/paddle/fluid/operators/metrics/accuracy_op.cc +++ b/paddle/fluid/operators/metrics/accuracy_op.cc @@ -36,7 +36,7 @@ class AccuracyOpMaker : public framework::OpProtoAndCheckerMaker { void Make() override { // TODO(typhoonzero): support both inference value and indices. AddInput("Out", "The network output of topk (inferences)"); - AddInput("Indices", "The the network output of topk (indices)"); + AddInput("Indices", "The network output of topk (indices)"); AddInput("Label", "Label of the training data"); // TODO(typhoonzero): AddInput("Weight", ... AddOutput("Accuracy", "The accuracy of current batch"); diff --git a/paddle/fluid/operators/mlu/mlu_baseop.cc b/paddle/fluid/operators/mlu/mlu_baseop.cc index 6b801924446..867c5f212ba 100644 --- a/paddle/fluid/operators/mlu/mlu_baseop.cc +++ b/paddle/fluid/operators/mlu/mlu_baseop.cc @@ -698,14 +698,14 @@ MLUCnnlTrigonDesc::~MLUCnnlTrigonDesc() { const bool exclusive, const bool reverse, const cnnlTensorDescriptor_t input_desc, const void* input, - const cnnlTensorDescriptor_t ouput_desc, + const cnnlTensorDescriptor_t output_desc, void* output) { cnnlHandle_t handle = GetHandleFromCTX(ctx); // NAN propagation mode: Only support CNNL_NOT_PROPAGATE_NAN now. cnnlNanPropagation_t mode = CNNL_NOT_PROPAGATE_NAN; PADDLE_ENFORCE_MLU_SUCCESS(cnnlCumsum(handle, input_desc, input, axis, - exclusive, reverse, mode, ouput_desc, + exclusive, reverse, mode, output_desc, output)); } diff --git a/paddle/fluid/operators/positive_negative_pair_op.cc b/paddle/fluid/operators/positive_negative_pair_op.cc index a9646b2e8ac..cbe58644f53 100644 --- a/paddle/fluid/operators/positive_negative_pair_op.cc +++ b/paddle/fluid/operators/positive_negative_pair_op.cc @@ -123,7 +123,7 @@ class PositiveNegativePairOp : public framework::OperatorWithKernel { column, depth, platform::errors::OutOfRange( "Attr(column) should be less than depth(the second " - "dimension of Input(Score)). Recieved Attr(column): %d, while " + "dimension of Input(Score)). Received Attr(column): %d, while " "depth is %d.", column, depth)); PADDLE_ENFORCE_GE( @@ -131,7 +131,7 @@ class PositiveNegativePairOp : public framework::OperatorWithKernel { platform::errors::OutOfRange( "Attr(column) should be greater than equal to negative " "depth, i.e. the second dimension of Input(Score). " - "Recieved Attr(column): %d, while negative depth is %d.", + "Received Attr(column): %d, while negative depth is %d.", column, -depth)); } diff --git a/paddle/fluid/operators/prune_gate_by_capacity_op.cu b/paddle/fluid/operators/prune_gate_by_capacity_op.cu index 7228bdbf380..6a2ed6592e7 100644 --- a/paddle/fluid/operators/prune_gate_by_capacity_op.cu +++ b/paddle/fluid/operators/prune_gate_by_capacity_op.cu @@ -98,7 +98,7 @@ static void VisitDataType(paddle::experimental::DataType type, visitor.template apply(); } else { PADDLE_THROW(platform::errors::InvalidArgument( - "The recieved values gate_id type %s can not meet input requirements. " + "The received values gate_id type %s can not meet input requirements. " "Because the given gate_id data type of operators must be " "int64. Please input appropriate gate_id again! ", "framework::DataTypeToString(type)")); diff --git a/paddle/fluid/operators/pscore/heter_listen_and_serv_op.cc b/paddle/fluid/operators/pscore/heter_listen_and_serv_op.cc index 2df0d7526a3..457e37744d3 100644 --- a/paddle/fluid/operators/pscore/heter_listen_and_serv_op.cc +++ b/paddle/fluid/operators/pscore/heter_listen_and_serv_op.cc @@ -63,7 +63,7 @@ void HeterListenAndServOp::RunAsyncLoop(framework::ProgramDesc *program) const { PADDLE_ENFORCE_EQ(pieces.size(), 2, platform::errors::PreconditionNotMet( "Invalid format of message_and_id argument. " - "Expected \"message:block_id\". Recieved %s", + "Expected \"message:block_id\". Received %s", grad_and_id.c_str())); PADDLE_ENFORCE_EQ(out_map->count(pieces[0]), 0, platform::errors::AlreadyExists( @@ -82,7 +82,7 @@ void HeterListenAndServOp::RunAsyncLoop(framework::ProgramDesc *program) const { PADDLE_ENFORCE_GE(num_blocks, 1, platform::errors::PreconditionNotMet( "Invalid number of blocks in server program. Expected " - "equal or greater than 1. Recieved %zu", + "equal or greater than 1. Received %zu", num_blocks)); std::vector block_list; for (size_t blkid = 1; blkid < num_blocks; ++blkid) { diff --git a/paddle/fluid/operators/rnn_op_xpu.cc b/paddle/fluid/operators/rnn_op_xpu.cc index 220d91bf4fa..941e463f63c 100644 --- a/paddle/fluid/operators/rnn_op_xpu.cc +++ b/paddle/fluid/operators/rnn_op_xpu.cc @@ -65,7 +65,7 @@ class RnnXPUKernel : public framework::OpKernel { auto* output = ctx.Output("Out"); auto* dropout_mask = ctx.Output("DropoutState"); auto* reserve_data = ctx.Output("Reserve"); - // Attrbutes + // Attributes const int& num_layers = ctx.Attr("num_layers"); const bool& is_bidirec = ctx.Attr("is_bidirec"); const int& hidden_size = ctx.Attr("hidden_size"); diff --git a/paddle/fluid/operators/sample_logits_op.cc b/paddle/fluid/operators/sample_logits_op.cc index 420c4c5f257..e02c7ade9a1 100644 --- a/paddle/fluid/operators/sample_logits_op.cc +++ b/paddle/fluid/operators/sample_logits_op.cc @@ -58,7 +58,7 @@ class SampleLogitsOpMaker : public framework::OpProtoAndCheckerMaker { AddOutput( "Probabilities", "(Tensor, default: Tensor), A 2-D tensor with shape [N, NT + S]." - "The probabilites of sampled positive and negtive labels.") + "The probabilities of sampled positive and negtive labels.") .AsIntermediate(); AddOutput("LogitsDim", "Store dim information of Logits for gradient op") .AsIntermediate(); diff --git a/paddle/fluid/operators/sigmoid_cross_entropy_with_logits_op_npu.cc b/paddle/fluid/operators/sigmoid_cross_entropy_with_logits_op_npu.cc index f186f95a2b9..ed173bb3ebf 100644 --- a/paddle/fluid/operators/sigmoid_cross_entropy_with_logits_op_npu.cc +++ b/paddle/fluid/operators/sigmoid_cross_entropy_with_logits_op_npu.cc @@ -22,7 +22,7 @@ using Tensor = framework::Tensor; const int kIgnoreIndex = -100; void CheckAttrs(const framework::ExecutionContext& ctx) { - // Add this check is is due to Ascend SigmoidCrossEntropyWithLogits + // Add this check is due to Ascend SigmoidCrossEntropyWithLogits // and SigmoidCrossEntropyWithLogitsGrad does't supoort // attr normalize and ignore_index bool normalize = ctx.Attr("normalize"); diff --git a/paddle/fluid/operators/sum_op.cu b/paddle/fluid/operators/sum_op.cu index 33590c1d7cc..8c6c083cde8 100644 --- a/paddle/fluid/operators/sum_op.cu +++ b/paddle/fluid/operators/sum_op.cu @@ -156,7 +156,7 @@ void SumToLoDTensor(const framework::ExecutionContext &context) { } } - // compute select rows seperately. + // compute select rows separately. if (!selectrow_index.empty()) { std::vector sr_in_out_data; size_t rows = 0; @@ -241,7 +241,7 @@ class SumKernel LodTensorArrayCompute(context); } else { PADDLE_THROW(platform::errors::InvalidArgument( - "Expected type of Ouput(out) must be Tensor, SelectedRows or " + "Expected type of Output(out) must be Tensor, SelectedRows or " "LodTensorArray. But got " "unsupport type: %s.", framework::ToTypeName(out_var->Type()))); diff --git a/paddle/fluid/operators/tdm_child_op.h b/paddle/fluid/operators/tdm_child_op.h index 963dfd3bf77..e437975320c 100644 --- a/paddle/fluid/operators/tdm_child_op.h +++ b/paddle/fluid/operators/tdm_child_op.h @@ -149,7 +149,7 @@ class TDMChildKernel : public framework::OpKernel { output_type == framework::proto::VarType::INT64; PADDLE_ENFORCE_EQ(out_type_match, true, platform::errors::InvalidArgument( - "Ouput(Child) & Output(LeafMask) holds the wrong " + "Output(Child) & Output(LeafMask) holds the wrong " "type, it holds %s, but " "desires to be %s or %s", paddle::framework::DataTypeToString(output_type), diff --git a/paddle/fluid/operators/warpctc_op.cc b/paddle/fluid/operators/warpctc_op.cc index 5cd9feee828..1583e5d84b2 100644 --- a/paddle/fluid/operators/warpctc_op.cc +++ b/paddle/fluid/operators/warpctc_op.cc @@ -95,7 +95,7 @@ An operator integrating the open-source https://arxiv.org/pdf/1512.02595v1.pdf), to compute Connectionist Temporal Classification (CTC) loss. It can be aliased as softmax with ctc, since a native softmax activation is -interated to the warp-ctc library, to to normalize values for each row of the +interated to the warp-ctc library, to normalize values for each row of the input tensor. More detail of CTC loss can be found by referring to diff --git a/paddle/fluid/platform/CMakeLists.txt b/paddle/fluid/platform/CMakeLists.txt index 356b5ab2cd2..24d39c25cf3 100644 --- a/paddle/fluid/platform/CMakeLists.txt +++ b/paddle/fluid/platform/CMakeLists.txt @@ -116,7 +116,7 @@ endif() cc_library(cudnn_workspace_helper SRCS cudnn_workspace_helper.cc DEPS boost) -# seperate init from device_context to avoid cycle dependencies +# separate init from device_context to avoid cycle dependencies cc_library(init SRCS init.cc DEPS device_context custom_kernel context_pool) # memcpy depends on device_context, here add deps individually for diff --git a/paddle/fluid/platform/device/npu/npu_op_runner.h b/paddle/fluid/platform/device/npu/npu_op_runner.h index 2409c14b760..739a3ef41e4 100644 --- a/paddle/fluid/platform/device/npu/npu_op_runner.h +++ b/paddle/fluid/platform/device/npu/npu_op_runner.h @@ -70,7 +70,7 @@ class NpuOpRunner { NpuOpRunner &AddInput(const Tensor &tensor); // NOTE(zhiqiu): CANN-5.0.2 support input tensors on host. - // Specifically, the tensor of shape, tensor of dims, etc, which are are small + // Specifically, the tensor of shape, tensor of dims, etc, which are small // vector/list. NpuOpRunner &AddInput(const Tensor &tensor, aclMemType mem_type); diff --git a/paddle/fluid/platform/profiler/event_node.cc b/paddle/fluid/platform/profiler/event_node.cc index 6c8be1811d7..b909fb5f25a 100644 --- a/paddle/fluid/platform/profiler/event_node.cc +++ b/paddle/fluid/platform/profiler/event_node.cc @@ -51,7 +51,7 @@ void NodeTrees::BuildTrees( const std::vector& host_event_nodes, std::vector& runtime_event_nodes, const std::vector& device_event_nodes) { - // seperate Host Event Nodes into different threads + // separate Host Event Nodes into different threads std::map> thread2host_event_nodes; // used to store HostTraceEventNodes per thread std::map> diff --git a/paddle/infrt/common/object.h b/paddle/infrt/common/object.h index ab2d00cce98..797595cc7c5 100644 --- a/paddle/infrt/common/object.h +++ b/paddle/infrt/common/object.h @@ -25,7 +25,7 @@ template class Shared; /** * Object is the basic element in the INFRT, with `Shared` wrapper, the object - * can be shared accross the system. + * can be shared across the system. */ struct Object { //! Get the type representation of this object. diff --git a/paddle/phi/infermeta/binary.cc b/paddle/phi/infermeta/binary.cc index 2139605fb20..837a43905e7 100644 --- a/paddle/phi/infermeta/binary.cc +++ b/paddle/phi/infermeta/binary.cc @@ -1534,7 +1534,7 @@ void MvInferMeta(const MetaTensor& x, const MetaTensor& vec, MetaTensor* out) { phi::errors::InvalidArgument( "X's second dimension is expected to be equal to " "Vec's first dimension" - "but recieved X'shape = [%s], Vec's shape = [%s]", + "but received X'shape = [%s], Vec's shape = [%s]", dim_x, dim_vec)); diff --git a/paddle/phi/infermeta/multiary.cc b/paddle/phi/infermeta/multiary.cc index e793eb8e668..48c40673ab8 100644 --- a/paddle/phi/infermeta/multiary.cc +++ b/paddle/phi/infermeta/multiary.cc @@ -458,7 +458,7 @@ void BatchNormInferMeta(const MetaTensor& x, true, phi::errors::InvalidArgument( "Each dimension of input tensor is expected to be -1 or a " - "positive number, but recieved %d. Input's shape is [%s].", + "positive number, but received %d. Input's shape is [%s].", x_dims[i], x_dims)); } @@ -755,7 +755,7 @@ inline int ConvOutputSize( 0, phi::errors::InvalidArgument( "The output's size is expected to be greater than 0. " - "But recieved: output's size is %d. The output's size is computed by " + "But received: output's size is %d. The output's size is computed by " "((input_size + 2 * padding - (dilation * (filter_size - 1) + 1)) / " "stride + 1), where input_size is %d, padding is %d, " "filter_size is %d, dilation is %d, stride is %d.", diff --git a/paddle/phi/infermeta/unary.cc b/paddle/phi/infermeta/unary.cc index 6d37a31f545..6c2956417a3 100644 --- a/paddle/phi/infermeta/unary.cc +++ b/paddle/phi/infermeta/unary.cc @@ -2746,7 +2746,7 @@ void UnfoldInferMeta(const MetaTensor& x, phi::errors::InvalidArgument( "The dims of X should be larger than that of kernel_sizes " "by a number of 2, due to the batch size and input channel dim. " - "But recieved dims(X:%u) - dims(kernel_sizes:%u) != 2", + "But received dims(X:%u) - dims(kernel_sizes:%u) != 2", in_dims.size(), kernel_sizes.size())); PADDLE_ENFORCE_EQ( @@ -2754,7 +2754,7 @@ void UnfoldInferMeta(const MetaTensor& x, kernel_sizes.size(), phi::errors::InvalidArgument( "The dims of strides should be the same with that of kernel_sizes. " - "But recieved dims(strides: %u) != dims(kernel_sizes: %u).", + "But received dims(strides: %u) != dims(kernel_sizes: %u).", strides.size(), kernel_sizes.size())); PADDLE_ENFORCE_EQ( @@ -2762,7 +2762,7 @@ void UnfoldInferMeta(const MetaTensor& x, 2 * strides.size(), phi::errors::InvalidArgument( "The dims of paddings should be 2 times of that of strides. " - "But recieved dims(paddings: %u) != 2*dims(strides: %u).", + "But received dims(paddings: %u) != 2*dims(strides: %u).", paddings.size(), strides.size())); PADDLE_ENFORCE_EQ( @@ -2770,7 +2770,7 @@ void UnfoldInferMeta(const MetaTensor& x, dilations.size(), phi::errors::InvalidArgument( "The dims of strides should be the same with that of dilations. " - "But recieved dims(strides: %u) != dims(dilations: %u).", + "But received dims(strides: %u) != dims(dilations: %u).", strides.size(), dilations.size())); @@ -2779,14 +2779,14 @@ void UnfoldInferMeta(const MetaTensor& x, 0, phi::errors::InvalidArgument( "The `kernel_sizes` should be greater than zero, " - "but recieved kernel_height: %d kernel_width: %d.", + "but received kernel_height: %d kernel_width: %d.", kernel_sizes[0], kernel_sizes[1])); PADDLE_ENFORCE_GT(kernel_sizes[1], 0, phi::errors::InvalidArgument( "The `kernel_sizes` should be greater than zero, " - "but recieved kernel_height: %d kernel_width: %d.", + "but received kernel_height: %d kernel_width: %d.", kernel_sizes[0], kernel_sizes[1])); // check strides @@ -2794,14 +2794,14 @@ void UnfoldInferMeta(const MetaTensor& x, 0, phi::errors::InvalidArgument( "The `strides` should be greater than zero, " - "but recieved strides_height: %d strides_width: %d.", + "but received strides_height: %d strides_width: %d.", strides[0], strides[1])); PADDLE_ENFORCE_GT(strides[1], 0, phi::errors::InvalidArgument( "The `strides` should be greater than zero, " - "but recieved strides_height: %d strides_width: %d.", + "but received strides_height: %d strides_width: %d.", strides[0], strides[1])); // check dilations @@ -2810,7 +2810,7 @@ void UnfoldInferMeta(const MetaTensor& x, 0, phi::errors::InvalidArgument( "The `dilations` should be greater than zero, " - "but recieved dilations_height: %d dilations_width: %d.", + "but received dilations_height: %d dilations_width: %d.", dilations[0], dilations[1])); PADDLE_ENFORCE_GT( @@ -2818,7 +2818,7 @@ void UnfoldInferMeta(const MetaTensor& x, 0, phi::errors::InvalidArgument( "The `dilations` should be greater than zero, " - "but recieved dilations_height: %d dilations_width: %d.", + "but received dilations_height: %d dilations_width: %d.", dilations[0], dilations[1])); diff --git a/paddle/phi/kernels/cpu/conv_util.h b/paddle/phi/kernels/cpu/conv_util.h index d26d89086b2..159a5cfbeb6 100644 --- a/paddle/phi/kernels/cpu/conv_util.h +++ b/paddle/phi/kernels/cpu/conv_util.h @@ -38,7 +38,7 @@ inline void UpdatePaddingAndDilation(std::vector* paddings, phi::errors::InvalidArgument( "Attribute padding's size should be the same or twice as the " "input's dimension. " - "But recieved: padding's size is %d, padding is [%s]; input's " + "But received: padding's size is %d, padding is [%s]; input's " "dimension is %d, input's shape is [%s].", paddings->size(), make_ddim(*paddings), diff --git a/paddle/phi/kernels/cpu/rnn_kernel.cc b/paddle/phi/kernels/cpu/rnn_kernel.cc index cae97eb0764..ae2c7a72635 100644 --- a/paddle/phi/kernels/cpu/rnn_kernel.cc +++ b/paddle/phi/kernels/cpu/rnn_kernel.cc @@ -808,7 +808,7 @@ struct BidirLayer : public Layer { mode, is_test); - // concat the the output result + // concat the output result funcs::ConcatFunctor concat_functor; concat_functor(dev_ctx, output_vec, static_cast(2), output); } diff --git a/paddle/phi/kernels/funcs/broadcast_function.h b/paddle/phi/kernels/funcs/broadcast_function.h index 38cd41d3b61..17735c05ada 100644 --- a/paddle/phi/kernels/funcs/broadcast_function.h +++ b/paddle/phi/kernels/funcs/broadcast_function.h @@ -53,7 +53,7 @@ struct DimensionsTransform { PADDLE_THROW(phi::errors::InvalidArgument( "The %d-th dimension of input tensor is expected to be equal " "with the %d-th dimension of output tensor %d or 1, but " - "recieved %d.", + "received %d.", in_idx + 1, axis + 1, out_dims[axis], @@ -70,7 +70,7 @@ struct DimensionsTransform { PADDLE_THROW(phi::errors::InvalidArgument( "The %d-th dimension of input tensor is expected to be equal " "with the %d-th dimension of output tensor %d or 1, but " - "recieved %d.", + "received %d.", in_idx + 1, in_idx + 1, out_dims[in_idx], @@ -552,7 +552,7 @@ void BroadcastKernelForDifferentDimSize( default: { PADDLE_THROW(phi::errors::InvalidArgument( "The maximum dimension of input tensor is expected to be less than " - "%d, but recieved %d.", + "%d, but received %d.", merge_dims.dim_size, phi::DDim::kMaxRank)); } @@ -578,7 +578,7 @@ void BroadcastKernelForDifferentVecSize( kArity, phi::errors::InvalidArgument( "The number of inputs is expected to be equal to the " - "arity of functor. But recieved: the number of inputs " + "arity of functor. But received: the number of inputs " "is %d, the arity of functor is %d.", ins.size(), kArity)); diff --git a/paddle/phi/kernels/funcs/elementwise_base.h b/paddle/phi/kernels/funcs/elementwise_base.h index 4ee46facc79..1093bdfa726 100644 --- a/paddle/phi/kernels/funcs/elementwise_base.h +++ b/paddle/phi/kernels/funcs/elementwise_base.h @@ -849,7 +849,7 @@ void ElementwiseKernel(const KPDevice &ctx, kArity, phi::errors::InvalidArgument( "The number of inputs is expected to be equal to the " - "arity of functor. But recieved: the number of inputs " + "arity of functor. But received: the number of inputs " "is %d, the arity of functor is %d.", ins.size(), kArity)); diff --git a/paddle/phi/kernels/gpu/bce_loss_kernel.cu b/paddle/phi/kernels/gpu/bce_loss_kernel.cu index adbcd3b2b62..b190bce4742 100644 --- a/paddle/phi/kernels/gpu/bce_loss_kernel.cu +++ b/paddle/phi/kernels/gpu/bce_loss_kernel.cu @@ -38,7 +38,7 @@ struct BCELossFunctor { HOSTDEVICE inline T operator()(const T x, const T label) const { PADDLE_ENFORCE( (x >= static_cast(0)) && (x <= one), - "Input is expected to be within the interval [0, 1], but recieved %f.", + "Input is expected to be within the interval [0, 1], but received %f.", x); T term1 = max(phi::kps::details::Log(x), neg_100); T term2 = max(phi::kps::details::Log(one - x), neg_100); diff --git a/paddle/phi/kernels/impl/searchsorted_kernel_impl.h b/paddle/phi/kernels/impl/searchsorted_kernel_impl.h index 82bd9fba2a6..e3cd6f5828d 100644 --- a/paddle/phi/kernels/impl/searchsorted_kernel_impl.h +++ b/paddle/phi/kernels/impl/searchsorted_kernel_impl.h @@ -158,7 +158,7 @@ static void VisitDataType(DataType type, Visitor visitor) { visitor.template apply(); } else { PADDLE_THROW(errors::InvalidArgument( - "The recieved values data type %s can not meet input requirements. " + "The received values data type %s can not meet input requirements. " "Because the given values data type of searchsorted operators must be " "float32, float64, int32 or int64. Please input appropriate " "sorted_sequence again! ", diff --git a/paddle/phi/ops/compat/scale_sig.cc b/paddle/phi/ops/compat/scale_sig.cc index 95deb007d99..8061a1fbd61 100644 --- a/paddle/phi/ops/compat/scale_sig.cc +++ b/paddle/phi/ops/compat/scale_sig.cc @@ -30,7 +30,7 @@ namespace phi { * The infrt declare like: * * def PDKEL_Reshape_to_CPU : Pat< - * (PD_ReshapeOp $x, $shape_tensor， $shape_attr), // OpMaker arguements + * (PD_ReshapeOp $x, $shape_tensor， $shape_attr), // OpMaker arguments * (PDKEL_ReshapeKernelAttr $x, fn($shape_attr)>; // Kernel arguments * def PDKEL_Reshape_to_CPU : Pat< * (PD_ReshapeOp $x, $shape_tensor， $shape_attr), diff --git a/python/paddle/README.rst b/python/paddle/README.rst index e779f1264c4..2d48ee4b26c 100644 --- a/python/paddle/README.rst +++ b/python/paddle/README.rst @@ -88,7 +88,7 @@ If you want to install paddlepaddle-gpu with cuda version of 9.0 ,10.0 ,10.1 ,or After the installation is complete, you can use `python` or `python3` to enter the Python interpreter and then use `import paddle.fluid` and `fluid.install_check.run_check()` -If `Your Paddle Fluid is installed succesfully!` appears, to verify that the installation was successful. +If `Your Paddle Fluid is installed successfully!` appears, to verify that the installation was successful. diff --git a/python/paddle/distributed/auto_parallel/process_group.py b/python/paddle/distributed/auto_parallel/process_group.py index 471448b031d..d1b6e57ddc1 100644 --- a/python/paddle/distributed/auto_parallel/process_group.py +++ b/python/paddle/distributed/auto_parallel/process_group.py @@ -156,6 +156,6 @@ class ProcessGroup: # Note that Process group 0 is reserved for representing all ranks. -# At the begining, group 0 is empty and new ranks will be added automatically. +# At the beginning, group 0 is empty and new ranks will be added automatically. _g_process_group_map = {} _g_process_group_map[0] = ProcessGroup(0, []) diff --git a/python/paddle/distributed/fleet/base/distributed_strategy.py b/python/paddle/distributed/fleet/base/distributed_strategy.py index 9d20e432d89..986d8e401e8 100644 --- a/python/paddle/distributed/fleet/base/distributed_strategy.py +++ b/python/paddle/distributed/fleet/base/distributed_strategy.py @@ -1168,9 +1168,9 @@ class DistributedStrategy(object): dp_degree(int, optional): specific the number of data parallelism group; when dp_degree >= 2, it will introduce dp_degree ways data parallelism as the outer parallelsim for the inner parallelsim. User is responsible to ensure global_world_size = mp_degree * sharding_degree * pp_degree * dp_degree. Default is 1. - mp_degree(int, optional): [Hybrid parallelism ONLY] specific the the number of gpus within each megatron parallelism group; and megatron parallelism will turn be off if mp_degree=1. Default is 1. + mp_degree(int, optional): [Hybrid parallelism ONLY] specific the number of gpus within each megatron parallelism group; and megatron parallelism will turn be off if mp_degree=1. Default is 1. - pp_degree(int, optional): [Hybrid parallelism ONLY] specific the the number of gpus within each pipeline parallelism group; and pipeline parallelism will turn be off if pp_degree=1. Default is 1. + pp_degree(int, optional): [Hybrid parallelism ONLY] specific the number of gpus within each pipeline parallelism group; and pipeline parallelism will turn be off if pp_degree=1. Default is 1. pp_allreduce_in_optimize(bool, optional): [Hybrid parallelism ONLY] move the allreduce operations from backward stage to update(optimize) stage when pipeline parallelsim is on. This configuration will affect the communication speed of Hybrid parallelism training depeneded on network topology. this strategy is experimental by now.. Default is False. @@ -1485,7 +1485,7 @@ class DistributedStrategy(object): **Notes**: k_steps(int) The local steps for training before parameter synchronization. Default 1. - begin_step(int) The step of begining training by localsgd. Default 1. + begin_step(int) The step of beginning training by localsgd. Default 1. Examples: @@ -1544,7 +1544,7 @@ class DistributedStrategy(object): init_k_steps(int) The initial steps for training before adaptive localsgd. Then, the adaptive localsgd method will modify init_k_steps automatically. Default 1. - begin_step(int) The step of begining training by adaptive localsgd. Default 1. + begin_step(int) The step of beginning training by adaptive localsgd. Default 1. Examples: diff --git a/python/paddle/distributed/fleet/launch.py b/python/paddle/distributed/fleet/launch.py index c5a9df50589..343cca7f4f0 100644 --- a/python/paddle/distributed/fleet/launch.py +++ b/python/paddle/distributed/fleet/launch.py @@ -556,7 +556,7 @@ def launch(): - ``--selected_mlus``: mlus aliases, recommend to use ``--mlus``. - - ``training_script``: The full path to the single GPU training program/script to be launched in parallel, followed by all the arguments for the training script. e.g., ``traing.py`` + - ``training_script``: The full path to the single GPU training program/script to be launched in parallel, followed by all the arguments for the training script. e.g., ``training.py`` - ``training_script_args``: The args of training_script. e.g., ``--lr=0.1`` diff --git a/python/paddle/distributed/fleet/meta_optimizers/ascend/ascend_parser.py b/python/paddle/distributed/fleet/meta_optimizers/ascend/ascend_parser.py index 8f1a4de86de..3a52041dc7e 100644 --- a/python/paddle/distributed/fleet/meta_optimizers/ascend/ascend_parser.py +++ b/python/paddle/distributed/fleet/meta_optimizers/ascend/ascend_parser.py @@ -1372,7 +1372,7 @@ class UniformRandomParser(AscendParserBase): max_v = self.op.attr("max") seed = self.op.attr("seed") dtype = self.op.attr("dtype") - assert max_v > min_v, "assert max_v > min_v, but recieved " + \ + assert max_v > min_v, "assert max_v > min_v, but received " + \ "as max_v={}, min_v={} ".format(max_v, min_v) tensor1 = self._create_ge_tensor([len(shape)], 2, shape) diff --git a/python/paddle/distributed/fleet/meta_optimizers/dygraph_optimizer/dygraph_sharding_optimizer.py b/python/paddle/distributed/fleet/meta_optimizers/dygraph_optimizer/dygraph_sharding_optimizer.py index b7edf583002..d487f35324d 100755 --- a/python/paddle/distributed/fleet/meta_optimizers/dygraph_optimizer/dygraph_sharding_optimizer.py +++ b/python/paddle/distributed/fleet/meta_optimizers/dygraph_optimizer/dygraph_sharding_optimizer.py @@ -170,7 +170,7 @@ class DygraphShardingOptimizer(object): result = self._inner_optimizer.minimize(loss, startup_program, parameters, no_grad_set) - # sync parameters accross sharding ranks + # sync parameters across sharding ranks self._sharding_sync_parameters() return result @@ -181,7 +181,7 @@ class DygraphShardingOptimizer(object): # actually updating self._inner_optimizer.step() - # sync parameters accross sharding ranks + # sync parameters across sharding ranks self._sharding_sync_parameters() # TODO is it a good way to make _grad_clip a property diff --git a/python/paddle/distributed/launch/main.py b/python/paddle/distributed/launch/main.py index 400a4472602..b2c87e737c8 100644 --- a/python/paddle/distributed/launch/main.py +++ b/python/paddle/distributed/launch/main.py @@ -54,7 +54,7 @@ def launch(): - ``--devices``: The selected accelerate devices on nodes, can be gpu/xpu/npu/mlu etc.. e.g., ``--devices=0,1,2,3`` will launch four training processes each bound to one device. - - ``training_script``: The full path to the single GPU training program/script to be launched in parallel, followed by all the arguments for the training script. e.g., ``traing.py`` + - ``training_script``: The full path to the single GPU training program/script to be launched in parallel, followed by all the arguments for the training script. e.g., ``training.py`` - ``training_script_args``: The args of training_script. e.g., ``--lr=0.1`` diff --git a/python/paddle/distributed/passes/auto_parallel_fp16.py b/python/paddle/distributed/passes/auto_parallel_fp16.py index 69c3eef7e37..9dda310e5c0 100644 --- a/python/paddle/distributed/passes/auto_parallel_fp16.py +++ b/python/paddle/distributed/passes/auto_parallel_fp16.py @@ -306,7 +306,7 @@ class FP16State(object): in_var_dist_attr = consume_op_attr.get_input_dist_attr( in_var.name) assert in_var_dist_attr is not None - # truely insert cast op + # truly insert cast op if cast_var is None or cast_var.dtype != dst_dtype: # NOTE we make the cast op and var's dist attr as the op that consume the # cast var instead of the op which generates the var diff --git a/python/paddle/distributed/ps/utils/public.py b/python/paddle/distributed/ps/utils/public.py index e7edc6fd859..7acfd6cfe19 100755 --- a/python/paddle/distributed/ps/utils/public.py +++ b/python/paddle/distributed/ps/utils/public.py @@ -748,7 +748,7 @@ def find_heter_ops(program, default_device="cpu"): def union_forward_gradient_op(program_block_ops_list): """ before analyzing the input & output of each block in program_block_list, we should - union the forward op and corresponding gradient op to elimincate the uneccessary variable + union the forward op and corresponding gradient op to elimincate the unnecessary variable transmit """ """ diff --git a/python/paddle/fluid/contrib/slim/tests/test_post_training_quantization_mobilenetv1.py b/python/paddle/fluid/contrib/slim/tests/test_post_training_quantization_mobilenetv1.py index 629529ff1b9..56d77f77b50 100644 --- a/python/paddle/fluid/contrib/slim/tests/test_post_training_quantization_mobilenetv1.py +++ b/python/paddle/fluid/contrib/slim/tests/test_post_training_quantization_mobilenetv1.py @@ -405,7 +405,7 @@ class TestPostTrainingAbsMaxForMobilenetv1(TestPostTrainingQuantization): is_full_quantize = False is_use_cache_file = False is_optimize_model = False - # The accuracy diff of post-traing quantization (abs_max) maybe bigger + # The accuracy diff of post-training quantization (abs_max) maybe bigger diff_threshold = 0.05 self.run_test(model, algo, round_type, data_urls, data_md5s, quantizable_op_type, is_full_quantize, is_use_cache_file, diff --git a/python/paddle/fluid/dygraph/checkpoint.py b/python/paddle/fluid/dygraph/checkpoint.py index 1ae57bcb303..ba5c709b1d8 100644 --- a/python/paddle/fluid/dygraph/checkpoint.py +++ b/python/paddle/fluid/dygraph/checkpoint.py @@ -257,7 +257,7 @@ def load_dygraph(model_path, **configs): para_dict = structured_para_dict else: # load state dict by `io.save_params/persistables` save format - # TODO(chenweihang): [ Now only supports loading parameters seperately ] + # TODO(chenweihang): [ Now only supports loading parameters separately ] # If users save all parameters as one file, the [ variable.name -> variable ] # mapping info will lost, so users need to give variable list, but users build # variable list in dygraph mode is difficult, we recommend users to use diff --git a/python/paddle/fluid/dygraph/dygraph_to_static/convert_operators.py b/python/paddle/fluid/dygraph/dygraph_to_static/convert_operators.py index 3a7b012b02b..576baf6cc29 100644 --- a/python/paddle/fluid/dygraph/dygraph_to_static/convert_operators.py +++ b/python/paddle/fluid/dygraph/dygraph_to_static/convert_operators.py @@ -167,7 +167,7 @@ def convert_logical_not(x): A function representation of a Python ``not`` statement. Args: - x(bool|Tensor): Operand of of ``not`` operator. + x(bool|Tensor): Operand of ``not`` operator. Returns: A python bool variable or a bool Tensor. diff --git a/python/paddle/fluid/executor.py b/python/paddle/fluid/executor.py index c6ff3a583d6..164545d0a05 100644 --- a/python/paddle/fluid/executor.py +++ b/python/paddle/fluid/executor.py @@ -873,7 +873,7 @@ class Executor(object): _fetch_list.append(item) else: raise TypeError( - "The item in fetch_list should be str, variable or optimize_op, but recieved %s.", + "The item in fetch_list should be str, variable or optimize_op, but received %s.", type(item)) for index, item in enumerate(fetch_list): diff --git a/python/paddle/fluid/incubate/fleet/parameter_server/ir/trainer_pass.py b/python/paddle/fluid/incubate/fleet/parameter_server/ir/trainer_pass.py index 46f26e8e52c..2c09abac9e7 100644 --- a/python/paddle/fluid/incubate/fleet/parameter_server/ir/trainer_pass.py +++ b/python/paddle/fluid/incubate/fleet/parameter_server/ir/trainer_pass.py @@ -1407,7 +1407,7 @@ def get_communicate_var_info(program, def union_forward_gradient_op(program_block_ops_list): """ before analyzing the input & output of each block in program_block_list, we should - union the forward op and corresponding gradient op to elimincate the uneccessary variable + union the forward op and corresponding gradient op to elimincate the unnecessary variable transmit """ """ diff --git a/python/paddle/fluid/layer_helper_base.py b/python/paddle/fluid/layer_helper_base.py index ce6fe6918b5..47f0c02d287 100644 --- a/python/paddle/fluid/layer_helper_base.py +++ b/python/paddle/fluid/layer_helper_base.py @@ -234,7 +234,7 @@ class LayerHelperBase(object): x=g, y=norm) # The shapes of g and norm are the same. # Currently, elementwise_mul only support broadcast when the shape # of y is a subset of the shape of x. Thus, we reshape y to squeeze - # to achive the subset. + # to achieve the subset. w = elementwise_mul( x=v, y=scale if dim is None else reshape( diff --git a/python/paddle/fluid/layers/nn.py b/python/paddle/fluid/layers/nn.py index 200e8feec1e..8be719758ef 100755 --- a/python/paddle/fluid/layers/nn.py +++ b/python/paddle/fluid/layers/nn.py @@ -13744,7 +13744,7 @@ def get_tensor_from_selected_rows(x, name=None): x.height = 20 x.value = [[1, 1] [2, 2] [2, 2] [3, 3] [6, 6]] - Ouput is LoDTensor: + Output is LoDTensor: out.shape = [5, 2] out.data = [[1, 1], [2, 2], diff --git a/python/paddle/fluid/layers/rnn.py b/python/paddle/fluid/layers/rnn.py index 707a1dc2cbc..b04cf90e1d8 100644 --- a/python/paddle/fluid/layers/rnn.py +++ b/python/paddle/fluid/layers/rnn.py @@ -673,7 +673,7 @@ def birnn(cell_fw, birnn creates a bidirectional recurrent neural network specified by RNNCell `cell_fw` and `cell_bw`, which performs :code:`cell.call()` (for dygraph mode :code:`cell.forward`) repeatedly until reaches to - the maximum length of `inputs` and then concat the ouputs for both RNNs + the maximum length of `inputs` and then concat the outputs for both RNNs along the last axis. Arguments: diff --git a/python/paddle/fluid/tests/unittests/ir/inference/auto_scan_test.py b/python/paddle/fluid/tests/unittests/ir/inference/auto_scan_test.py index bb8c6e73fde..161c785ef85 100755 --- a/python/paddle/fluid/tests/unittests/ir/inference/auto_scan_test.py +++ b/python/paddle/fluid/tests/unittests/ir/inference/auto_scan_test.py @@ -388,7 +388,7 @@ class PassAutoScanTest(AutoScanTest): used_time = time.time() - start_time if max_duration > 0 and used_time > max_duration: logging.error( - "The duration exceeds {} seconds, if this is neccessary, try to set a larger number for parameter `max_duration`.". + "The duration exceeds {} seconds, if this is necessary, try to set a larger number for parameter `max_duration`.". format(max_duration)) assert False diff --git a/python/paddle/fluid/tests/unittests/ir/inference/test_shuffle_channel_detect_pass.py b/python/paddle/fluid/tests/unittests/ir/inference/test_shuffle_channel_detect_pass.py index a864e2fe5a1..1781eb50483 100644 --- a/python/paddle/fluid/tests/unittests/ir/inference/test_shuffle_channel_detect_pass.py +++ b/python/paddle/fluid/tests/unittests/ir/inference/test_shuffle_channel_detect_pass.py @@ -62,13 +62,13 @@ class TestShuffleChannelDetectPass(PassAutoScanTest): "transpose2", inputs={"X": ["reshape2_output1"], }, outputs={ - "Out": ["transpose2_ouput"], + "Out": ["transpose2_output"], "XShape": ["transpose2_xshape"] }, axis=axis_v) reshape2_op2 = OpConfig( "reshape2", - inputs={"X": ["transpose2_ouput"], }, + inputs={"X": ["transpose2_output"], }, outputs={ "Out": ["reshape2_output2"], "XShape": ["reshape2_xshape2"] diff --git a/python/paddle/fluid/tests/unittests/ir/pass_test.py b/python/paddle/fluid/tests/unittests/ir/pass_test.py index aae1cc65c92..e92821387ae 100644 --- a/python/paddle/fluid/tests/unittests/ir/pass_test.py +++ b/python/paddle/fluid/tests/unittests/ir/pass_test.py @@ -167,7 +167,7 @@ class PassTest(unittest.TestCase): def _check_fused_ops(self, program): ''' - Check the number of specified fused op is equal to the the expected + Check the number of specified fused op is equal to the expected number. ''' if self.fused_op_type is None or self.num_fused_ops < 0: diff --git a/python/paddle/fluid/tests/unittests/op_test.py b/python/paddle/fluid/tests/unittests/op_test.py index 13c72bedefa..f7a3dfa1102 100644 --- a/python/paddle/fluid/tests/unittests/op_test.py +++ b/python/paddle/fluid/tests/unittests/op_test.py @@ -872,7 +872,7 @@ class OpTest(unittest.TestCase): eager_tensor_outputs = egr_oups if egr_oups else self.append_input_output_for_dygraph( op_proto, self.outputs, False, False, block) - # prepare attrbutes + # prepare attributes attrs_outputs = {} if hasattr(self, "attrs"): for attrs_name in self.attrs: @@ -906,7 +906,7 @@ class OpTest(unittest.TestCase): outputs = self.append_input_output_for_dygraph( op_proto, self.outputs, False, False, block) - # prepare attrbutes + # prepare attributes attrs_outputs = {} if hasattr(self, "attrs"): for attrs_name in self.attrs: @@ -2016,7 +2016,7 @@ class OpTest(unittest.TestCase): outputs = self.append_input_output_for_dygraph( op_proto, self.outputs, False, False, block) - # prepare attrbutes + # prepare attributes attrs_outputs = {} if hasattr(self, "attrs"): for attrs_name in self.attrs: diff --git a/python/paddle/fluid/tests/unittests/seresnext_test_base.py b/python/paddle/fluid/tests/unittests/seresnext_test_base.py index cc40b89b585..bf33adcf486 100644 --- a/python/paddle/fluid/tests/unittests/seresnext_test_base.py +++ b/python/paddle/fluid/tests/unittests/seresnext_test_base.py @@ -25,7 +25,7 @@ class TestResnetBase(TestParallelExecutorBase): check_func, use_device, delta2=1e-5, - compare_seperately=True): + compare_separately=True): if use_device == DeviceType.CUDA and not core.is_compiled_with_cuda(): return @@ -45,7 +45,7 @@ class TestResnetBase(TestParallelExecutorBase): batch_size=seresnext_net.batch_size(use_device), use_device=use_device) - if compare_seperately: + if compare_separately: for loss in zip(func_1_first_loss, func_2_first_loss): self.assertAlmostEquals(loss[0], loss[1], delta=1e-5) for loss in zip(func_1_last_loss, func_2_last_loss): diff --git a/python/paddle/fluid/tests/unittests/static_model_parallel_fused_attention.py b/python/paddle/fluid/tests/unittests/static_model_parallel_fused_attention.py index b57f2677623..4dc3fe6eab6 100644 --- a/python/paddle/fluid/tests/unittests/static_model_parallel_fused_attention.py +++ b/python/paddle/fluid/tests/unittests/static_model_parallel_fused_attention.py @@ -69,9 +69,9 @@ class ParallelFusedMultiHeadAttention(Layer): super(ParallelFusedMultiHeadAttention, self).__init__() assert embed_dim > 0, ("Expected embed_dim to be greater than 0, " - "but recieved {}".format(embed_dim)) + "but received {}".format(embed_dim)) assert num_heads > 0, ("Expected nhead to be greater than 0, " - "but recieved {}".format(num_heads)) + "but received {}".format(num_heads)) self.normalize_before = normalize_before self._dtype = self._helper.get_default_dtype() diff --git a/python/paddle/fluid/tests/unittests/static_model_parallel_fused_feedforward.py b/python/paddle/fluid/tests/unittests/static_model_parallel_fused_feedforward.py index 5f467da6a64..ad570fc0acf 100644 --- a/python/paddle/fluid/tests/unittests/static_model_parallel_fused_feedforward.py +++ b/python/paddle/fluid/tests/unittests/static_model_parallel_fused_feedforward.py @@ -172,10 +172,10 @@ class ParallelFusedFeedForward(Layer): name=None): super(ParallelFusedFeedForward, self).__init__() assert d_model > 0, ( - "Expected d_model to be greater than 0, but recieved {}".format( + "Expected d_model to be greater than 0, but received {}".format( d_model)) assert dim_feedforward > 0, ( - "Expected dim_feedforward to be greater than 0, but recieved {}". + "Expected dim_feedforward to be greater than 0, but received {}". format(dim_feedforward)) self._dtype = self._helper.get_default_dtype() diff --git a/python/paddle/fluid/tests/unittests/test_fleet_rolemaker.py b/python/paddle/fluid/tests/unittests/test_fleet_rolemaker.py index 4655b628dab..f382d61c637 100644 --- a/python/paddle/fluid/tests/unittests/test_fleet_rolemaker.py +++ b/python/paddle/fluid/tests/unittests/test_fleet_rolemaker.py @@ -52,7 +52,7 @@ class TestCloudRoleMaker(unittest.TestCase): self.assertTrue(ro.is_server()) self.assertEqual(ro.worker_num(), 2) - def test_traing_role(self): + def test_training_role(self): """Test training role.""" os.environ["TRAINING_ROLE"] = "TEST" ro = role_maker.PaddleCloudRoleMaker(is_collective=False) diff --git a/python/paddle/fluid/tests/unittests/test_fleet_rolemaker_new.py b/python/paddle/fluid/tests/unittests/test_fleet_rolemaker_new.py index 5e8be9a8522..86ee0db30ef 100644 --- a/python/paddle/fluid/tests/unittests/test_fleet_rolemaker_new.py +++ b/python/paddle/fluid/tests/unittests/test_fleet_rolemaker_new.py @@ -116,7 +116,7 @@ class TestCloudRoleMaker(unittest.TestCase): self.assertEqual(ro._all_gather(1, "worker"), 1) self.assertEqual(ro._all_reduce(1, "sum", "worker"), 1) - def test_traing_role(self): + def test_training_role(self): """Test training role.""" os.environ["TRAINING_ROLE"] = "TEST" diff --git a/python/paddle/fluid/tests/unittests/test_gradient_clip.py b/python/paddle/fluid/tests/unittests/test_gradient_clip.py index 7984ca55716..20a55af15c4 100644 --- a/python/paddle/fluid/tests/unittests/test_gradient_clip.py +++ b/python/paddle/fluid/tests/unittests/test_gradient_clip.py @@ -162,7 +162,7 @@ class TestGradientClipByGlobalNorm(TestGradientClip): "gradient clip by global norm has wrong results!, \nu={}\nv={}\ndiff={}". format(u, v, u - v)) - # test whether the ouput is right when use 'set_gradient_clip' + # test whether the output is right when use 'set_gradient_clip' def test_old_gradient_clip(self): def func(params_grads): clip = fluid.clip.GradientClipByGlobalNorm(clip_norm=self.clip_norm) @@ -172,7 +172,7 @@ class TestGradientClipByGlobalNorm(TestGradientClip): self.clip_gradient = func self.check_gradient_clip(fluid.CPUPlace()) - # test whether the ouput is right when use grad_clip + # test whether the output is right when use grad_clip def test_new_gradient_clip(self): def func(params_grads): clip = fluid.clip.GradientClipByGlobalNorm(clip_norm=self.clip_norm) @@ -181,7 +181,7 @@ class TestGradientClipByGlobalNorm(TestGradientClip): self.clip_gradient = func self.check_gradient_clip(fluid.CPUPlace()) - # test whether the ouput is right when use grad_clip under float64 + # test whether the output is right when use grad_clip under float64 def test_new_gradient_clip_fp64(self): def func(params_grads): clip = fluid.clip.GradientClipByGlobalNorm(clip_norm=self.clip_norm) @@ -267,7 +267,7 @@ class TestGradientClipByNorm(TestGradientClip): a=u, b=v, rtol=1e-5, atol=1e-8), "gradient clip by norm has wrong results!") - # test whether the ouput is right when use grad_clip + # test whether the output is right when use grad_clip def test_gradient_clip(self): def func(params_grads): clip = fluid.clip.GradientClipByNorm(clip_norm=self.clip_norm) @@ -311,7 +311,7 @@ class TestGradientClipByValue(TestGradientClip): a=u, b=v, rtol=1e-6, atol=1e-8), "gradient clip by value has wrong results!") - # test whether the ouput is right when use grad_clip + # test whether the output is right when use grad_clip def test_gradient_clip(self): def func(params_grads): clip = fluid.clip.GradientClipByValue(max=self.max, min=self.min) @@ -397,7 +397,7 @@ class TestDygraphGradientClipByGlobalNorm(TestDygraphGradientClip): self.assertTrue( np.isclose( a=a, b=b, rtol=1e-6, atol=1e-8), - "gradient clip by global norm has wrong results, expetcd:%f, but recieved:%f" + "gradient clip by global norm has wrong results, expetcd:%f, but received:%f" % (a, b)) @@ -426,7 +426,7 @@ class TestDygraphGradientClipByNorm(TestDygraphGradientClip): self.assertTrue( np.isclose( a=a, b=b, rtol=1e-6, atol=1e-8), - "gradient clip by norm has wrong results, expetcd:%f, but recieved:%f" + "gradient clip by norm has wrong results, expetcd:%f, but received:%f" % (a, b)) @@ -517,7 +517,7 @@ class TestDygraphGradientClipFP16(unittest.TestCase): self.assertTrue( np.isclose( a=a, b=b, rtol=1e-3, atol=1e-8), - "gradient clip by global norm has wrong results, expetcd:%f, but recieved:%f" + "gradient clip by global norm has wrong results, expetcd:%f, but received:%f" % (a, b)) @@ -563,7 +563,7 @@ class TestDygraphGradientClipFP64(unittest.TestCase): self.assertTrue( np.isclose( a=a, b=b, rtol=1e-6, atol=1e-8), - "gradient clip by global norm has wrong results, expetcd:%f, but recieved:%f" + "gradient clip by global norm has wrong results, expetcd:%f, but received:%f" % (a, b)) diff --git a/python/paddle/fluid/tests/unittests/test_hsigmoid_op.py b/python/paddle/fluid/tests/unittests/test_hsigmoid_op.py index 965ae65614a..51ff8ec943d 100644 --- a/python/paddle/fluid/tests/unittests/test_hsigmoid_op.py +++ b/python/paddle/fluid/tests/unittests/test_hsigmoid_op.py @@ -198,7 +198,7 @@ class TestHSigmoidOp(OpTest): @skip_check_grad_ci( - reason="For 'TestHSigmoidOpSparse', check_grad is is separately calculated by 'TestHSigmoidOpWithSparseGrad'." + reason="For 'TestHSigmoidOpSparse', check_grad is separately calculated by 'TestHSigmoidOpWithSparseGrad'." ) class TestHSigmoidOpSparse(OpTest): def setUp(self): diff --git a/python/paddle/fluid/tests/unittests/test_optimizer.py b/python/paddle/fluid/tests/unittests/test_optimizer.py index ba1e9be815d..a0c5ce77f1d 100644 --- a/python/paddle/fluid/tests/unittests/test_optimizer.py +++ b/python/paddle/fluid/tests/unittests/test_optimizer.py @@ -1123,7 +1123,7 @@ class TestOptimizerDtype(unittest.TestCase): class TestMasterWeightSaveForFP16(unittest.TestCase): ''' - For Amp-O2, some optimizer(Momentum, Adam ...) will create master weights for parameters to to improve the accuracy. + For Amp-O2, some optimizer(Momentum, Adam ...) will create master weights for parameters to improve the accuracy. Master weights will be saved by optimizer::state_dict. ''' diff --git a/python/paddle/fluid/tests/unittests/test_parallel_executor_seresnext_base_cpu.py b/python/paddle/fluid/tests/unittests/test_parallel_executor_seresnext_base_cpu.py index 20a5fcb7af3..9b48a87bff7 100644 --- a/python/paddle/fluid/tests/unittests/test_parallel_executor_seresnext_base_cpu.py +++ b/python/paddle/fluid/tests/unittests/test_parallel_executor_seresnext_base_cpu.py @@ -32,7 +32,7 @@ class TestResnetCPU(TestResnetBase): self._compare_result_with_origin_model( check_func, use_device=DeviceType.CPU, - compare_seperately=False, + compare_separately=False, delta2=1e-3) diff --git a/python/paddle/fluid/tests/unittests/test_parallel_executor_seresnext_base_gpu.py b/python/paddle/fluid/tests/unittests/test_parallel_executor_seresnext_base_gpu.py index 9d1364cc592..ff529ce94bd 100644 --- a/python/paddle/fluid/tests/unittests/test_parallel_executor_seresnext_base_gpu.py +++ b/python/paddle/fluid/tests/unittests/test_parallel_executor_seresnext_base_gpu.py @@ -30,7 +30,7 @@ class TestResnetGPU(TestResnetBase): optimizer=seresnext_net.optimizer, use_parallel_executor=False) self._compare_result_with_origin_model( - check_func, use_device=DeviceType.CUDA, compare_seperately=False) + check_func, use_device=DeviceType.CUDA, compare_separately=False) if __name__ == '__main__': diff --git a/python/paddle/fluid/tests/unittests/test_unpool_op.py b/python/paddle/fluid/tests/unittests/test_unpool_op.py index f6dc3fba6a2..95ad254a6df 100644 --- a/python/paddle/fluid/tests/unittests/test_unpool_op.py +++ b/python/paddle/fluid/tests/unittests/test_unpool_op.py @@ -116,7 +116,7 @@ class TestUnpoolOpcase1(TestUnpoolOp): self.output_size = None -class TestUnpoolOpOuputsize(TestUnpoolOp): +class TestUnpoolOpOutputsize(TestUnpoolOp): def init_test_case(self): self.unpool2d_forward_naive = unpool2dmax_forward_naive self.unpooling_type = "max" @@ -127,7 +127,7 @@ class TestUnpoolOpOuputsize(TestUnpoolOp): self.output_size = [9, 9] -class TestUnpoolOpOuput(TestUnpoolOp): +class TestUnpoolOpOutput(TestUnpoolOp): def init_test_case(self): self.unpool2d_forward_naive = unpool2dmax_forward_naive self.unpooling_type = "max" diff --git a/python/paddle/framework/io.py b/python/paddle/framework/io.py index c1891d24b88..8e8dd785511 100644 --- a/python/paddle/framework/io.py +++ b/python/paddle/framework/io.py @@ -1039,7 +1039,7 @@ def _legacy_load(path, **configs): config) else: # load state dict by `io.save_params/persistables` save format - # TODO(chenweihang): [ Now only supports loading parameters seperately ] + # TODO(chenweihang): [ Now only supports loading parameters separately ] # If users save all parameters as one file, the [ variable.name -> variable ] # mapping info will lost, so users need to give variable list, but users build # variable list in dygraph mode is difficult, we recommend users to use diff --git a/python/paddle/hapi/model.py b/python/paddle/hapi/model.py index c17a56fc28d..4b349a19577 100644 --- a/python/paddle/hapi/model.py +++ b/python/paddle/hapi/model.py @@ -915,7 +915,7 @@ class Model(object): When training on GPU, auto mixed precision (AMP O1) and pure float16 (AMP O2) training are both supported in static mode and dynamic mode. - In static graph mode, before traing with pure float16 (AMP O2), + In static graph mode, before training with pure float16 (AMP O2), `multi_precision` could be set to True when creating optimizer, which can avoid poor accuracy or slow convergence in a way, and inputs of dtype float should be cast to float16 by users. `paddle.static.amp.fp16_guard` API @@ -2075,7 +2075,7 @@ class Model(object): # [input1, input2, ..., label1, lable2, ...] # 3. custumed iterator yield concated inputs and labels: # [input1, input2, ..., label1, lable2, ...] - # 4. custumed iterator yield seperated inputs and labels: + # 4. custumed iterator yield separated inputs and labels: # ([input1, input2, ...], [label1, lable2, ...]) # To handle all of these, flatten (nested) list to list. data = flatten(data) diff --git a/python/paddle/incubate/autotune.py b/python/paddle/incubate/autotune.py index e98a23bc52d..7ac555e2520 100644 --- a/python/paddle/incubate/autotune.py +++ b/python/paddle/incubate/autotune.py @@ -49,7 +49,7 @@ def set_config(config=None): dictionary, the key is the tuning type, and the value is a dictionary of the corresponding tuning parameters. If it is a string, the path of a json file will be specified and the tuning configuration will be set - by the the json file. Default: None, auto-tuning for kernel, layout and + by the json file. Default: None, auto-tuning for kernel, layout and dataloader will be enabled. Examples: diff --git a/python/paddle/incubate/distributed/models/moe/grad_clip.py b/python/paddle/incubate/distributed/models/moe/grad_clip.py index b620253b9f2..cf56f74d1f1 100644 --- a/python/paddle/incubate/distributed/models/moe/grad_clip.py +++ b/python/paddle/incubate/distributed/models/moe/grad_clip.py @@ -158,7 +158,7 @@ class ClipGradForMOEByGlobalNorm(ClipGradBase): normal_params_grads = [] moe_params_grads = [] - # seperate moe params from normal params + # separate moe params from normal params if self.moe_group is not None and self.moe_group.nranks > 1: for p, g in params_grads: if self.is_expert_param_func(p): diff --git a/python/paddle/incubate/nn/layer/fused_transformer.py b/python/paddle/incubate/nn/layer/fused_transformer.py index d76b990958c..072c7d9fcca 100644 --- a/python/paddle/incubate/nn/layer/fused_transformer.py +++ b/python/paddle/incubate/nn/layer/fused_transformer.py @@ -101,9 +101,9 @@ class FusedMultiHeadAttention(Layer): super(FusedMultiHeadAttention, self).__init__() assert embed_dim > 0, ("Expected embed_dim to be greater than 0, " - "but recieved {}".format(embed_dim)) + "but received {}".format(embed_dim)) assert num_heads > 0, ("Expected nhead to be greater than 0, " - "but recieved {}".format(num_heads)) + "but received {}".format(num_heads)) self.normalize_before = normalize_before self._dtype = self._helper.get_default_dtype() @@ -278,10 +278,10 @@ class FusedFeedForward(Layer): super(FusedFeedForward, self).__init__() assert d_model > 0, ( - "Expected d_model to be greater than 0, but recieved {}".format( + "Expected d_model to be greater than 0, but received {}".format( d_model)) assert dim_feedforward > 0, ( - "Expected dim_feedforward to be greater than 0, but recieved {}". + "Expected dim_feedforward to be greater than 0, but received {}". format(dim_feedforward)) self._dtype = self._helper.get_default_dtype() @@ -434,12 +434,12 @@ class FusedTransformerEncoderLayer(Layer): super(FusedTransformerEncoderLayer, self).__init__() assert d_model > 0, ("Expected d_model to be greater than 0, " - "but recieved {}".format(d_model)) + "but received {}".format(d_model)) assert nhead > 0, ("Expected nhead to be greater than 0, " - "but recieved {}".format(nhead)) + "but received {}".format(nhead)) assert dim_feedforward > 0, ( "Expected dim_feedforward to be greater than 0, " - "but recieved {}".format(dim_feedforward)) + "but received {}".format(dim_feedforward)) attn_dropout_rate = dropout_rate if attn_dropout_rate is None else attn_dropout_rate act_dropout_rate = dropout_rate if act_dropout_rate is None else act_dropout_rate self.normalize_before = normalize_before @@ -808,11 +808,11 @@ class FusedMultiTransformer(Layer): super(FusedMultiTransformer, self).__init__() assert embed_dim > 0, ("Expected embed_dim to be greater than 0, " - "but recieved {}".format(embed_dim)) + "but received {}".format(embed_dim)) assert num_heads > 0, ("Expected nhead to be greater than 0, " - "but recieved {}".format(num_heads)) + "but received {}".format(num_heads)) assert dim_feedforward > 0, ( - "Expected dim_feedforward to be greater than 0, but recieved {}". + "Expected dim_feedforward to be greater than 0, but received {}". format(dim_feedforward)) self.normalize_before = normalize_before diff --git a/python/paddle/nn/functional/loss.py b/python/paddle/nn/functional/loss.py index ca3ac177282..0d5ad46a4fb 100755 --- a/python/paddle/nn/functional/loss.py +++ b/python/paddle/nn/functional/loss.py @@ -542,7 +542,7 @@ def margin_ranking_loss(input, name=None): r""" - This op the calcluate the the margin rank loss between the input, other and label, use the math function as follows. + This op the calcluate the margin rank loss between the input, other and label, use the math function as follows. .. math:: margin\_rank\_loss = max(0, -label * (input - other) + margin) @@ -879,7 +879,7 @@ def kl_div(input, label, reduction='mean', name=None): While :attr:`reduction` is :attr:`none`, output loss is in the same shape as input, loss in each point is calculated - seperately and no reduction is applied. + separately and no reduction is applied. While :attr:`reduction` is :attr:`mean`, output loss is in shape of [1] and loss value is the mean value of all losses. @@ -2006,7 +2006,7 @@ def sigmoid_focal_loss(logit, Available dtype is float32, float64. normalizer (Tensor, optional): The number normalizes the focal loss. It has to be a 1-D Tensor whose shape is `[1, ]`. The data type is float32, float64. - For object detection task, it is the the number of positive samples. + For object detection task, it is the number of positive samples. If set to None, the focal loss will not be normalized. Default is None. alpha(int|float, optional): Hyper-parameter to balance the positive and negative example, it should be between 0 and 1. Default value is set to 0.25. diff --git a/python/paddle/nn/layer/activation.py b/python/paddle/nn/layer/activation.py index cd82fe12fff..7fd109843be 100644 --- a/python/paddle/nn/layer/activation.py +++ b/python/paddle/nn/layer/activation.py @@ -367,7 +367,7 @@ class PReLU(Layer): Parameters: num_parameters (int, optional): Number of `weight` to learn. The supported values are: 1 - a single parameter `alpha` is used for all input channels; - Number of channels - a seperate `alpha` is used for each input channel. + Number of channels - a separate `alpha` is used for each input channel. Default is 1. init (float, optional): Init value of learnable `weight`. Default is 0.25. weight_attr(ParamAttr, optional): The parameter attribute for the learnable `weight`. diff --git a/python/paddle/nn/layer/transformer.py b/python/paddle/nn/layer/transformer.py index b0b6e62a602..340372f9b6a 100644 --- a/python/paddle/nn/layer/transformer.py +++ b/python/paddle/nn/layer/transformer.py @@ -163,9 +163,9 @@ class MultiHeadAttention(Layer): super(MultiHeadAttention, self).__init__() assert embed_dim > 0, ("Expected embed_dim to be greater than 0, " - "but recieved {}".format(embed_dim)) + "but received {}".format(embed_dim)) assert num_heads > 0, ("Expected num_heads to be greater than 0, " - "but recieved {}".format(num_heads)) + "but received {}".format(num_heads)) self.embed_dim = embed_dim self.kdim = kdim if kdim is not None else embed_dim @@ -508,12 +508,12 @@ class TransformerEncoderLayer(Layer): super(TransformerEncoderLayer, self).__init__() assert d_model > 0, ("Expected d_model to be greater than 0, " - "but recieved {}".format(d_model)) + "but received {}".format(d_model)) assert nhead > 0, ("Expected nhead to be greater than 0, " - "but recieved {}".format(nhead)) + "but received {}".format(nhead)) assert dim_feedforward > 0, ( "Expected dim_feedforward to be greater than 0, " - "but recieved {}".format(dim_feedforward)) + "but received {}".format(dim_feedforward)) attn_dropout = dropout if attn_dropout is None else attn_dropout act_dropout = dropout if act_dropout is None else act_dropout @@ -813,12 +813,12 @@ class TransformerDecoderLayer(Layer): super(TransformerDecoderLayer, self).__init__() assert d_model > 0, ("Expected d_model to be greater than 0, " - "but recieved {}".format(d_model)) + "but received {}".format(d_model)) assert nhead > 0, ("Expected nhead to be greater than 0, " - "but recieved {}".format(nhead)) + "but received {}".format(nhead)) assert dim_feedforward > 0, ( "Expected dim_feedforward to be greater than 0, " - "but recieved {}".format(dim_feedforward)) + "but received {}".format(dim_feedforward)) attn_dropout = dropout if attn_dropout is None else attn_dropout act_dropout = dropout if act_dropout is None else act_dropout @@ -1220,12 +1220,12 @@ class Transformer(Layer): super(Transformer, self).__init__() assert d_model > 0, ("Expected d_model to be greater than 0, " - "but recieved {}".format(d_model)) + "but received {}".format(d_model)) assert nhead > 0, ("Expected nhead to be greater than 0, " - "but recieved {}".format(nhead)) + "but received {}".format(nhead)) assert dim_feedforward > 0, ( "Expected dim_feedforward to be greater than 0, " - "but recieved {}".format(dim_feedforward)) + "but received {}".format(dim_feedforward)) if isinstance(bias_attr, (list, tuple)): if len(bias_attr) == 1: diff --git a/python/paddle/profiler/profiler.py b/python/paddle/profiler/profiler.py index 77adbaff348..9df595bc3ae 100644 --- a/python/paddle/profiler/profiler.py +++ b/python/paddle/profiler/profiler.py @@ -150,7 +150,7 @@ def make_scheduler(*, def _default_state_scheduler(step: int): r""" - A default state scheduler, keep recording from the begining of the profiler until ending. + A default state scheduler, keep recording from the beginning of the profiler until ending. """ return ProfilerState.RECORD diff --git a/python/paddle/profiler/timer.py b/python/paddle/profiler/timer.py index 1fb06ddc55e..815775ebc6a 100644 --- a/python/paddle/profiler/timer.py +++ b/python/paddle/profiler/timer.py @@ -193,7 +193,7 @@ class TimerHook(Hook): def before_reader(self, benchmark): """ Initialize the start time of the dataloader. This function will be - called at the begining of `next` method in `_DataLoaderIterMultiProcess` or + called at the beginning of `next` method in `_DataLoaderIterMultiProcess` or `_DataLoaderIterSingleProcess`. """ @@ -220,8 +220,8 @@ class TimerHook(Hook): Record the cost for the current step. It will contain the cost of the loading data if there is a dataloader. Similar to `after_reader`, it will also update the maximum, minimum and the total time from the step 11 to the current step - as well as the the maximum and minimum speed of the model. This function will - be called in in `Profiler.step()`. + as well as the maximum and minimum speed of the model. This function will + be called in `Profiler.step()`. """ @@ -401,7 +401,7 @@ class Benchmark(object): # enter a new task but not calling beign() to record it. # we pause the timer until the end of new task, so that # the cost of new task is not added to the current event. - # eg. start evaluation in the traing task + # eg. start evaluation in the training task self.current_event.need_record = False else: # when the new task exits, continue timing for the current event. diff --git a/python/paddle/profiler/utils.py b/python/paddle/profiler/utils.py index fd75ab9550d..5e95c83129f 100644 --- a/python/paddle/profiler/utils.py +++ b/python/paddle/profiler/utils.py @@ -78,7 +78,7 @@ class RecordEvent(ContextDecorator): def begin(self): r""" - Record the time of begining. + Record the time of beginning. Examples: diff --git a/python/paddle/tensor/math.py b/python/paddle/tensor/math.py index 5ee372f7b95..7cfae842337 100644 --- a/python/paddle/tensor/math.py +++ b/python/paddle/tensor/math.py @@ -3346,7 +3346,7 @@ def increment(x, value=1.0, name=None): def all(x, axis=None, keepdim=False, name=None): """ - Computes the the ``logical and`` of tensor elements over the given dimension. + Computes the ``logical and`` of tensor elements over the given dimension. Args: x (Tensor): An N-D Tensor, the input data type should be `bool`. @@ -3442,7 +3442,7 @@ def all(x, axis=None, keepdim=False, name=None): def any(x, axis=None, keepdim=False, name=None): """ - Computes the the ``logical or`` of tensor elements over the given dimension. + Computes the ``logical or`` of tensor elements over the given dimension. Args: x (Tensor): An N-D Tensor, the input data type should be `bool`. diff --git a/python/paddle/tensor/to_string.py b/python/paddle/tensor/to_string.py index 71c97d4cac9..42d3bf9fca3 100644 --- a/python/paddle/tensor/to_string.py +++ b/python/paddle/tensor/to_string.py @@ -42,7 +42,7 @@ def set_printoptions(precision=None, Args: precision (int, optional): Number of digits of the floating number, default 8. threshold (int, optional): Total number of elements printed, default 1000. - edgeitems (int, optional): Number of elements in summary at the begining and ending of each dimension, default 3. + edgeitems (int, optional): Number of elements in summary at the beginning and ending of each dimension, default 3. sci_mode (bool, optional): Format the floating number with scientific notation or not, default False. linewidth (int, optional): Number of characters each line, default 80. diff --git a/python/paddle/vision/models/mobilenetv3.py b/python/paddle/vision/models/mobilenetv3.py index da7ae010c58..70aa1b833d6 100644 --- a/python/paddle/vision/models/mobilenetv3.py +++ b/python/paddle/vision/models/mobilenetv3.py @@ -39,7 +39,7 @@ model_urls = { class SqueezeExcitation(nn.Layer): """ This block implements the Squeeze-and-Excitation block from https://arxiv.org/abs/1709.01507 (see Fig. 1). - Parameters ``activation``, and ``scale_activation`` correspond to ``delta`` and ``sigma`` in in eq. 3. + Parameters ``activation``, and ``scale_activation`` correspond to ``delta`` and ``sigma`` in eq. 3. This code is based on the torchvision code with modifications. You can also see at https://github.com/pytorch/vision/blob/main/torchvision/ops/misc.py#L127 Args: -- GitLab