Hangchen YU

## Motivation and Context The issue was reported by https://github.com/utcs-scea/ava/issues/124, while the error may appear at other APIs. ## How has this been tested? Benchmark: https://github.com/tensorflow/benchmarks/tree/master/scripts/tf_cnn_benchmarks Run worker: `./install/bin/legacy_manager --worker_path...

**Is your feature request related to a problem? Please describe.** Merge the changes in #175, #176 and #177 to ONNXRT and TF specs (probably also PT_DUMP spec).

## Description Just for test.

## Description The previous dump spec failed to dump fatbinaries with no kernel or no useful kernel. This diff handles this corner case. ## Types of changes - [x] Bug...

**Is your feature request related to a problem? Please describe.** This will enable us to remove libboost entirely from ava. **Describe the solution you'd like** Probably replace the asyncio with...

Preserve ordering of ava_async APIs between multiple threads

4

In the current prototype, `ava_async` means the API returns right after it's sent to the API server, when the API server may not receive or execute the API yet. In...

See [PR 131](https://github.com/utcs-scea/ava/pull/131/commits/4af7cd4d6f673bee80924a9ad68d769feee830ce#diff-c18e117174ed7bdd360872a5b7fc724fd6b59ef3236b30650acb4fab5f057962R59). ```cpp typedef struct { union Algorithm { cudnnConvolutionFwdAlgo_t convFwdAlgo; cudnnConvolutionBwdFilterAlgo_t convBwdFilterAlgo; cudnnConvolutionBwdDataAlgo_t convBwdDataAlgo; cudnnRNNAlgo_t RNNAlgo; cudnnCTCLossAlgo_t CTCLossAlgo; } algo; } cudnnAlgorithm_t; ``` The generated code is like:...

Update serializer to use C++17 standard

1

The imported [serializer](https://github.com/eyalz800/serializer) uses C++14 standard. I'd be good to upgrade it to comply C++17. This can also improve the performance of registering polymorphic types which can be slower in...