torch-mlir/build_tools/torchscript_e2e_heavydep_tests
Yi Zhang fadd76e9b8 E2e for MiniLM-L6-H384-uncased-sst2
Replace the original BertSequenceClassification with this new one.
The ops needed to support are identical.
2021-10-05 12:45:19 -04:00
..
README.md Updates to Readme.md (#334) 2021-09-28 13:50:25 -07:00
basic_mt.py Dual license the torch-mlir project. 2021-10-01 10:46:08 -07:00
generate_serialized_tests.sh Add BertSequenceClassification model to e2e 2021-09-30 13:30:29 -04:00
main.py E2e for MiniLM-L6-H384-uncased-sst2 2021-10-05 12:45:19 -04:00
minilm_sequence_classification.py E2e for MiniLM-L6-H384-uncased-sst2 2021-10-05 12:45:19 -04:00

README.md

Additional end-to-end tests with heavy dependencies (heavydep tests)

Some end-to-end tests require additional dependencies which don't make sense to include as part of the default npcomp setup. Additionally, these dependencies often don't work with the same HEAD PyTorch version that npcomp builds against at the C++ level.

We have a self-contained script that generates all the needed artifacts from a self-contained virtual environment. It can be used like so:

# Build the virtual environment in the specified directory and generate the
# serialized test artifacts in the other specified directory.
# This command is safe to re-run if you have already built the virtual
# environment and just changed the tests.
build_tools/torchscript_e2e_heavydep_tests/generate_serialized_tests.sh \
  path/to/heavydep_venv \
  path/to/heavydep_serialized_tests

# Add the --serialized-test-dir flag to point at the directory containing the
# serialized tests. All other functionality is the same as the normal invocation
# of torchscript_e2e_test.sh, but the serialized tests will be available.
tools/torchscript_e2e_test.sh --serialized-test-dir=path/to/heavydep_serialized_tests

The tests use the same (pure-Python) test framework as the normal torchscript_e2e_test.sh, but the tests are added in build_tools/torchscript_e2e_heavydep_tests instead of frontends/pytorch/e2e_testing/torchscript.

We rely critically on serialized TorchScript compatibility across PyTorch versions to transport the tests + pure-Python compatibility of the torch API, which has worked well so far.