Sample TensorFlow model for testing and examples
Closed this issue · 2 comments
Since the correctness of this library depends a lot on how it interacts with engine files that are created from the C++ library it would be greatly helpful to create a TensorFlow model that we can use for testing and example code.
I'm thinking something along the lines of a common object detector using image net or something of the like. No matter what model is chosen it should be something that's widely known and accessible to others.
We should keep the .hd5 and .uff model files stored somewhere accessible (AWS?) and then work on a .onnx version when support for that comes online.
Taking a quick peek at the tutorial examples provided by TensorFlow, I think going with the CNN example is going to be the best.
It's a simple image classification rather than object detection but it'll get the point across and importantly it's not a complicated model so it doesn't distract from highlighting the workings of the library.
Ended up going with using the lenet5 model from the TensorRT samples.
lenet5.uff is the test input for the uff parser and the engine created from that parser is serialized to a file and then read back in and tested against the original file.
I'll still have to find a good way to include the .uff file in the git repository for future contributors to have to test with.