tree: 6d91ee2d620af4ebeda81a5e76891c4631e75d3b [path history] [tgz]
  1. BUILD
  2. check-no-tf.mlir
  3. legalize-tf.mlir
  4. lower-global-tensors.mlir
  5. lower-global-tensors_complex.mlir
  6. lower-global-tensors_invalid.mlir
  7. propagate-resource-casts.mlir
  8. README.md
  9. saved_model_adopt_exports.py
integrations/tensorflow/compiler/test/README.md

Running tests manually

$ bazel test :saved_model_adopt_exports

This will capture the output and pass it through FileCheck and report pass/fail, along with a hopefully informative description of what failed.

Debugging failures

During development, it can be useful to just see the raw output directly.

To see the raw output of the MLIR import and conversion process:

$ bazel run :saved_model_adopt_exports -- --disable_filecheck

Look for the RUN_TEST: <test_name> and FINISH_TEST: <test_name> lines to narrow in on the test that interests you.