1 /* Copyright 2020 The TensorFlow Authors. All Rights Reserved. 2 3 Licensed under the Apache License, Version 2.0 (the "License"); 4 you may not use this file except in compliance with the License. 5 You may obtain a copy of the License at 6 7 http://www.apache.org/licenses/LICENSE-2.0 8 9 Unless required by applicable law or agreed to in writing, software 10 distributed under the License is distributed on an "AS IS" BASIS, 11 WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 12 See the License for the specific language governing permissions and 13 limitations under the License. 14 ==============================================================================*/ 15 #ifndef TENSORFLOW_CORE_PROFILER_UTILS_COST_UTILS_H_ 16 #define TENSORFLOW_CORE_PROFILER_UTILS_COST_UTILS_H_ 17 18 #include <string> 19 20 #include "absl/container/flat_hash_set.h" 21 #include "tensorflow/core/grappler/costs/cost_estimator.h" 22 #include "tensorflow/core/grappler/costs/op_level_cost_estimator.h" 23 #include "tensorflow/core/platform/macros.h" 24 #include "tensorflow/core/platform/types.h" 25 #include "tensorflow/core/profiler/utils/xplane_visitor.h" 26 27 namespace tensorflow { 28 namespace profiler { 29 30 // This is a wrapper of tensorflow::grappler::OpLevelCostEstimator and use 31 // tracing time information to estimate the roof line stats for each traced 32 // tensorflow op. 33 class TfOpRoofLineCostEstimator 34 : public tensorflow::grappler::OpLevelCostEstimator { 35 public: 36 TfOpRoofLineCostEstimator() = default; 37 ~TfOpRoofLineCostEstimator() override; 38 39 grappler::DeviceInfo GetDeviceInfo( 40 const DeviceProperties& device) const override; 41 42 struct OpRoofLineStats { 43 uint64 flops = 0LL; 44 uint64 bytes_accessed = 0LL; 45 bool inaccurate = false; 46 }; 47 OpRoofLineStats Predict(const XEventVisitor& event); 48 49 private: 50 absl::flat_hash_set<std::string> 51 unsupported_ops_; // summary for unsupported ops. 52 53 TF_DISALLOW_COPY_AND_ASSIGN(TfOpRoofLineCostEstimator); 54 }; 55 56 } // namespace profiler 57 } // namespace tensorflow 58 59 #endif // TENSORFLOW_CORE_PROFILER_UTILS_COST_UTILS_H_ 60