* fix boost download url (#7843)
* Topo sort the model before saving (#7913)
* checkin toposort
* review comments
* revert and add TODO
* Add shape inference to custom symbolic functions (#7937)
**Description**: As title.
**Motivation and Context**
- PyTorch ONNX exporter heavily depends on ONNX shape inference to export accurate and efficient model. Custom symbolic function exports the op as contrib ops, thus exporter is unable to perform standard onnx shape inference. Models with dynamic shape inputs are affected.
* Fix missing files on linux (#8066)
* [Mobile package] Update required operator config with additional ops for wav2vec2. (#8079)
Add some additional ops to the mobile package that are needed for the wav2vec2 model.
* Add module attribute to ORTModule to support HuggingFace Trainer save_model (#8088)
* Fix input schema extrator for ORTModule (#8098)
* Fix 32bit Android java API crash (#8122)
* Fix 32bit Android java API crash
* fix code formating
* [Mobile package] Update required operator config with additional ops for newer version of Wav2Vec 2. (#8123)
This is an update to https://github.com/microsoft/onnxruntime/pull/8079
The sample application motivating the original update changed to use an updated version of the model. Now, fewer ops are required. This change removes the previously added ops which are no longer needed.
* Add int64 as a required type to ConstantOfShape as it's used by the pytorch converter for Pad. (#8128)
It's also used pointlessly for torch.tensor.repeat (although that usage should always be able to be constant folded).
* Update logic in props.xml to account for shared provider library changes (#8138)
* Ortmodule override torch.manual_seed() (#8131)
* Ortmodule override torch.manual_seed()
* Fix Python Cuda loading issues (#7939)
* Fix mac shared_provider warning (#8153)
Co-authored-by: Guoyu Wang <62914304+gwang-msft@users.noreply.github.com>
Co-authored-by: Ye Wang <52801275+wangyems@users.noreply.github.com>
Co-authored-by: Bowen Bao <bowbao@microsoft.com>
Co-authored-by: Ryan Hill <38674843+RyanUnderhill@users.noreply.github.com>
Co-authored-by: Edward Chen <18449977+edgchen1@users.noreply.github.com>
Co-authored-by: baijumeswani <bmeswani@microsoft.com>
Co-authored-by: Thiago Crepaldi <thiago.crepaldi@microsoft.com>
Co-authored-by: Scott McKay <skottmckay@gmail.com>
Co-authored-by: Hariharan Seshadri <shariharan91@gmail.com>
Co-authored-by: Sherlock <baihan.huang@gmail.com>
* Cache initializers and avoid device check ot end of forward (#7905)
* ATenOp Enhancement (#7725)
* config parser, default argument values
* ut
* win build
* maxpool2d
* fix win build
* fix build
* unfold atenop
* Update CMakeLists.txt for openvino EP (#7980)
* Add SoftmaxCrossEntropyLossInternal to Support Dynamic ignore_index Input (#7899)
* add SoftmaxCrossEntropyLossInternal
* bugfix and ut
* fix ut
* fix ut
* support torch1.8.1
* function body for nll_loss_internal
* Override ORTModule named_modules to support extra arg (#7954)
* add missing provider_options.h in packages (#7995)
* consolidate copy binary script for gpu/trt tarball package
* add provider_options.h
* add provider_options.h
* Add cuda provides files (#8002)
* Save module output for backward if needed (#8010)
* Save module output for backward if needed
* Make logic in InsertCastTransformer around forcing a node to fp32 more precise. (#8018)
* Address #7981
Reworked the logic around forcing a node to run on fp32 even if it was supported on fp16.
The github issue had multiple factors. In ORT 1.8 we remove Identity nodes that produce graph outputs as they're not needed. That resulted in a Loop node no longer having output nodes (it produces graph outputs instead), which meant the check in IsSingleInputNodeFloat16Node returned true as there was no longer a downstream Identity node processing fp16 data.
We shouldn't only force a node to fp32 in very specific circumstances, and the changes hopefully check for those more precisely.
* Fix Memory Leak from DlpackToOrtValue (#8029)
* Update DirectML EP changes from DmlDev as of 2021-06-07 (#7987)
* Merged PR 6093117: Fix test_DynamicQuantizedLinear_max_adjusted_expanded by allowing Identity operator to run on non-float inputs
Motivation:
As part of the OnnxConformance Backend tests, DynamicQuantizedLinear_max_adjusted_expanded is failing.
Root Cause:
- The test model has `Identity` operator as one of the node. The input of this node is of non-float data type.
- In DML, `Identity` operator is registered as operator which requires floating input.
- As per `DirectMLSchema.h`, support for non-float input has been added for `Identity` operator in DML but the same has not been reflected in the `OperatorRegistration.cpp`.
Changes:
- Removed all traces of the requiresFloatFormatsForGraph flag from it's definition and usage. This flag was only used for Identity and it's related operator.
- Added null check for the graphOutput nodeArg in GraphDescBuilder.cpp to stop the crash of the test.
Related work items: #33076298
* Merged PR 6103324: Remove usage of non-generic error code (FWP_E_NULL_POINTER)
Motivation:
Addressing Dwayne comment on the previous PR. [Ref: [6093117](https://dev.azure.com/microsoft/WindowsAI/_git/onnxruntime/pullrequest/6093117?discussionId=44292162&path=%2Fonnxruntime%2Fcore%2Fproviders%2Fdml%2FDmlExecutionProvider%2Fsrc%2FGraphPartitioner.cpp)]
Changes:
Inside the DML EP, we should not use some other platform specific error codes. Instead we should a appropriate generic error code.
Related work items: #33076298
Co-authored-by: Sumit Agarwal <sumitagarwal@microsoft.com>
* [js/react_native] Use a mobile ORT instead of a full ORT (#8042)
* Change full ort to mobile ort
* Update Android example to load mobile ort
* Change the format of test models to ort
* update ios to use mobile ort
* revise README
* use onnxruntime-mobile-c CocoaPods in a npm package
* fix PATH addition in windows
should set PATH, not add to the tail the copy of PATH
* Reduce Kernel Optimization (#8067)
* reduce optimization
* bug fix
* add a check
* add ut
* refactor
* add ut cases for keepdims=true
Co-authored-by: baijumeswani <bmeswani@microsoft.com>
Co-authored-by: Vincent Wang <wangwchpku@outlook.com>
Co-authored-by: Changming Sun <chasun@microsoft.com>
Co-authored-by: George Wu <jywu@microsoft.com>
Co-authored-by: Ryan Hill <38674843+RyanUnderhill@users.noreply.github.com>
Co-authored-by: Sherlock <baihan.huang@gmail.com>
Co-authored-by: Scott McKay <skottmckay@gmail.com>
Co-authored-by: sumitsays <sumitagarwal330@gmail.com>
Co-authored-by: Sumit Agarwal <sumitagarwal@microsoft.com>
Co-authored-by: Sunghoon <35605090+hanbitmyths@users.noreply.github.com>
Co-authored-by: iperov <lepersorium@gmail.com>
* Include ORT format model conversion scripts and infrastructure in ORT python package.
- tweak existing script setup so it can be easily run directly and from the ORT python package
Add config file and readme for Android minimal build package
Update ORT Mobile doco
Disable warning if 'all' optimizations are enabled but NCHWc transformer is excluded (device specific optimizations don't apply in this scenario so the warning is moot).
* Address PR comments
* working on re-organizing js code for ortweb
* remove dup files
* move folder
* fix common references
* fix common es5
* add webpack to common
* split interfact/impl
* use cjs for node
* add npmignore for common
* update sourcemap config for common
* update node
* adjust folder/path in CI and build
* update folder
* nit: readme
* add bundle for dev
* correct nodejs paths
* enable ORT_API_MANUAL_INIT
* set name for umd library
* correct name for commonjs export
* add priority into registerBackend()
* fix npm ci pwd
* update eslintrc
* revise code
* revert package-lock lockfileVersion 2->1
* update prebuild
* resolve comments
* update document
* revise eslint config
* update eslint for typescript rules
* revert changes by mistake in backend.ts
* add env
* resolve comments
* Updates to some operators to always support int32 and int64 based on testing of Android package build config with a minimal build.
If an operator can be used for shape manipulation (int64) it is frequently used for indices manipulation (int32), so we enable both types for that set of ops.
- e.g. BERT models take indices as input
- Scatter/Gather ops utilize indices
Misc. fix to python bindings to exclude call that fails in a minimal build.
Enable type reduction for Scatter/ScatterElements CPU kernels. Some refactoring to reduce binary size.
Add MLTypeCallDispatcher methods.
Minor cleanup for Pad CPU kernel.
* Allow specific optimizers to be disabled.
- replace unused ability to specify just the optimizers to run
- never used so not needed
Allow the disabled list to be specified via the python bindings
- expected usage is internal, so using kwargs for that so as not to pollute the documentation with stuff no user is likely to need
Update the ORT format model conversion script to disable NCHWc transformer when level is 'all'
- currently there aren't any known use cases where we'd want the NCHWc transformations to run as they create a device specific model and aren't used on ARM
- the ORT format model is not expected to be generated on the target device (e.g. generate on Windows/Linux/macOS to deploy to Android/iOS so there's a good chance we'd generate a useless/invalid model
- default to 'all' as ARM and MLAS prefer NHWC and the NHWC transformer runs at that level
* Add matching changes to optimizer generation in training code
Enable type reduction for Shrink, Sign, SplitToSequence CPU kernels.
Some other type reduction changes including refactoring to specify element types in a single place.
Add types support for DenseToSparse and SparseToDense conversions
Address the case of empty sparse values and indicies when the initializer does
not contain any NNZ.
Add sparsify script.
Update the kernel def hashing in ORT format models. The new hashing logic ignores the ordering of type constraint types.
This is a backward compatibility breaking change, but we don't guarantee backward compatibility yet.
* Add support for custom ops library to the ORT model conversion script
Simplify model conversion now that we read ops from the ORT format model.
Enable custom ops in the python bindings if custom ops are turned on in a minimal build.
* Add test of model conversion involving custom ops.
* Add type reduction support to Min, Max and Pow
Update the C++ type reduction infrastructure to allow specifying an opset for the supported types list, as those can change across opset versions.
Minor updates to the type usage tracking script
* Add 'all opsets' macros and constant
* Add ability to generate configuration that includes required types for individual operators, to allow build size reduction based on that.
- Add python bindings for ORT format models
- Add script to update bindings and help info
- Add parsing of ORT format models
- Add ability to enable type reduction to config generation
- Update build.py to only allow operator/type reduction via config
- simpler to require config to be generated first
- can't mix a type aware (ORT format model only) and non-type aware config as that may result in insufficient types being enabled
- Add script to create reduced build config
- Update CIs
* Expand the documentation on using compiling EPs with a minimal build to call out a 'simple' option that is easier to use. Provide more background on what happens to help users choose the best option for them.
Tweak conversion script to be noisier about attempted usage of 'all' optimization level.
Co-authored-by: manashgoswami <magoswam@microsoft.com>
Update ORT model conversion script
- add args for specifying optimization level and whether to use NNAPI
- add logic to create a list of required ops and ORT format model that can be used with NNAPI
Follow up to #5811 to automate cleanup of the build docker image cache.
Added a script and build definition to clean up docker images that haven't been accessed recently.
This PR adds infrastructure to automatically cache docker images used in CI builds in a container registry.
Currently, build images are pulled from a container registry for some builds and built every time for others. The container registry requires maintenance to keep the images up to date and building images every time wastes build agent resources.
With this change, a given build image can be looked up in a cache container registry and if present, pulled, and otherwise, built and pushed. The uniqueness of a build image is determined by a hash digest of the dockerfile, docker build context directory, and certain "docker build" options. This digest is part of the image tag in the cache container repository.
The cache container registry will need to be cleaned up periodically. This is not automated yet.
* Remove serialization of outer scope node arg info in ORT format model. We don't currently need it in a minimal build as only SessionState calls Graph::IsConstantInitializer and it doesn't search outer scope. If we do need it in the future the information can be calculated at runtime (small binary size cost to do so).
Motivation: ORT format model was 32% bigger for a BERT model with multiple levels of subgraph and a lot of nodes due to this. Size is about 5% larger of the original ONNX model with the change. ORT format has type/shape info for all nodes, and this model has 2000 nodes so this seems reasonable.
Added example code to dump ORT format model to json.
Fixed misc bug in python test script around handling float and non-float expected output.
* Add minimal build option to build.py
Group some of the build settings so binary size reduction options are all together
Make some cmake variable naming more consistent
Replace usage of std::hash with murmurhash3 for kernel. std::hash is implementation dependent so can't be used.
Add initial doco and ONNX to ORT model conversion script
Misc cleanups of minimal build breaks.
* Changes to enable saving and loading an ORT format model via the public APIs.
Cleanup session.py to try and make slightly more understandable. More refactoring is needed here.
Couple of bug fixes
* Fix bug in handling NodeArg serialization for optional inputs which has a name and no type info.
* Address PR comments
- tweak SessionOptions config to avoid double lookup
- merge duplicated functionality in python binding around registering an EP with optional options
Fix a couple of build issues.
* Update C API to be consistent with python API
- only load model in InferenceSession ctor if required
- support loading ORT model in minimal build
* Fix nodejs test.
We get an invalid path error from LoadInterOp first now
* Another attempt at fixing nodejs test.
Error message depends on whether ENABLE_LANGUAGE_INTEROP_OPS is defined. Make the output consistent.
The interop implementation looks suspicious given it appears to be internal code that is going via the public api. TBD if that should be fixed.
* Fix couple of build issues.
* Disable test temporarily so PR can be checked in.
Will fix in separate PR that adds final pieces for minimal build as the test is required there.
* Give up on nodejs test and make the match simpler.
Fix init call in TrainingSession python to not pass through sess. it wasn't being used in Session anyway so passing it through just adds confusion.
* Fix call to Session.__init__ in TrainingSession.
Session now initializes Session._sess to None to make it clearer where the 'ownership' of that member is, and that needs to happen before TrainingSession sets it.
* minor fix for test dir util
* add pause option for onnx_test_runner
* add flush std to show pause prompt text
Co-authored-by: gwang0000 <62914304+gwang0000@users.noreply.github.com>
