Stella Laurenzo 0c73c535d6 Capture backward conv and copy_ kernels. ... * This is sufficient to capture the forward and backward pass and gradients of a convolutional model with an nllloss. * As with the forward conv, the backward conv is a special case wrapped in an enigma on the PyTorch side. There aren't many like it, so special casing is just what we do. * When I traced this, I found that the copy_ op is not yet boxing compatible so I had to map it manually. If there are many more like this, I'll probably do something a bit more clever to reduce duplication. * This exposes new signature patterns that will need to be handled by the ATen lowering. Will take care of that next: It will be nice to have an e2e of a non-trivial case with full gradients. * Fixes #97.		2020-10-30 22:59:26 -07:00
..
test_conv_nllloss_grads.py	Capture backward conv and copy_ kernels.	2020-10-30 22:59:26 -07:00
test_export_ResA.py	NFC: Use new print with large_elements_limit in tests.	2020-10-22 13:04:24 -07:00
test_export_add3.py	Expose signature metadata to ops and implement ATenRecognizeKernelsPass pass.	2020-10-26 20:31:45 -07:00
test_export_batchnorm.py	More progress on PyTorch acap device capture.	2020-10-15 21:43:21 -07:00
test_export_conv2d_fwd.py	Expose signature metadata to ops and implement ATenRecognizeKernelsPass pass.	2020-10-26 20:31:45 -07:00
test_export_multi_out.py	Port prior acap export tests to new dispatcher based versions.	2020-10-13 16:37:46 -07:00