Currently, when using OrtEnableProfiling to enable profiling using the C API,
the profile output file is created but is always empty.
The reason is that InferenceSession::EndProfiling() needs to be called to
write the profiling data to the output file.
However there's currently no way to call this function via the C API.
This adds a call to EndProfiling() to the descructor of the session if
profiling is enabled in the session options.
Generalize node removal method in graph_utils. This is a higher-level method that keeps the graph consistent so that no Resolve is needed after the removal of a node.
The new method supports the removal of nodes with a single input (be it an incoming node or an initializer) and a single output (but allowing multiple output edges of that output). It also takes into account the case that one of the output edges is fed to a subgraph.
Also updated the rewrite rules to use this new, less restrictive method, and improved the rules' conditions. Introduced a GraphEdge struct to simplify various methods in graph_utils.
* Update NMS to compatible with both TF & Pytoch models
* update text
* set max_output_boxes_per_batch, iou_threshold, score_threshold as optional input to support dynamic value
* fix typo
* Set the last output selected_indices as optional output
* fix shape inference in case the input don't have shape
* Update schema to remove scores & boxes from output. support for class broadcast.
* change max_output_boxes_per_batch to max_output_boxes_per_class
* update schema to remove the class dimension from boxes
* Update BUILD.md
* Update README.md
* Update tensorrt_execution_provider.cc
remap node index to handle the case that nodes in graph may be deleted and node index is not continuous.
* Update onnxruntime_providers.cmake
Solve conflicts to onnx-tensorrt
* Update tensorrt_execution_provider.h
* Update tensorrt_execution_provider.cc
* Update tensorrt_execution_provider.h
* Update tensorrt_execution_provider.cc
* Update tensorrt_execution_provider.h
* Update tensorrt_execution_provider.cc
* Update tensorrt_execution_provider.cc
* Update tensorrt_execution_provider.cc
* Update build.py
* Update onnx
* Support updated function schema in ORT
* Update onnx related commit hash
* Check out an older commit in ONNX
* Add support for subgraph attribute
* Add comments
* added tools for doc gen, added doc
* doc updated
* some fixes
* hooked up with build.py
* hooked up with build.py and fail on nonupdated doc
* update
* fix graph transformers and refactor tests
* fix merge master
* Set default optimization level to Level1
* fix build warnings for Linux
* try root cause tensorrt test failures
* try root cause tensorrt test failure
* Test level2 transformers with all CI builds
* remove ConvActivation fusion transformer
* change default level back to level1
* remove providers from apply api
* more changes
* Convert unsqueeze elimination to rewrite rule
* Simplify the way we register predefined transformers and rules in the inference session (all details are now moved to the graph transformer utils)
* Some reorganization and renaming of methods in graph_utils
* Updates in graph transformers test
* Update in edge removal to not perform unnecessary check of node args that led to race conditions when updating the graph
* Improve documentation for rewrite rules
* Remove top-down rule-based transformer (given we currently have only one type of rule-based transformer)
* Refactor InferenceSession interface
* Make some member and func private
* more protected members
* more protected
* reorder class members
* reordering
* reordering
The InferenceSession was implemented in the pImpl idiom, which hides the actual implementation. There are requirements to expose the implementation to other new classes, so this change is to pave the way.
The main changes are: abandon the pImpl idiom of InferenceSession
