burn/burn-import/onnx-tests
Luni-4 445603401d
ci/Check dependencies (#895)
2023-11-19 10:35:03 -05:00
..
src Support no-std for models generated by burn-import (#822) 2023-09-21 11:43:07 -04:00
tests ci/Fix `cargo clippy` action (#942) 2023-11-16 19:35:38 -05:00
Cargo.toml ci/Check dependencies (#895) 2023-11-19 10:35:03 -05:00
README.md Add Dropout ONNX OP and other refactoring (#606) 2023-08-09 09:53:20 -04:00
build.rs Implement tensor.recip() function to calculate elementwise reciprocals (#953) 2023-11-15 09:17:32 -05:00

README.md

ONNX Tests

This crate contains ONNX models that are utilized in testing the conversion of ONNX to Burn source code through the burn-import crate. The tests are designed as end-to-end tests, ensuring that ONNX models are accurately converted into Burn source code. Of utmost importance is verifying that the converted Burn source code compiles without errors and produces the same output as the original ONNX model.

Here is the directory structure of this crate:

  • tests/<model>: This directory contains the ONNX model and the Python script to generate it.
  • tests/<model>/<model>.onnx: The ONNX model is generated by the script.
  • tests/<model>/<model>.py: This is the Python script responsible for generating the ONNX model using PyTorch.
  • tests/onnx_tests.rs: This is the main test file, where all the tests are contained.
  • build.rs: This build script generates the ONNX models and is executed by cargo test before running the actual tests.

Adding new tests

Here are the steps to add a new test:

  1. Add your Python script to the tests/<model> directory. Refer to existing scripts for examples.
  2. Run your Python script to generate the ONNX model and inspect the output of the model with the test data. Use the inputs and outputs in your test.
  3. Make sure the ONNX output contains the desired operators by verifying with the Netron app. Sometimes PyTorch will optimize the model and remove operators that are not necessary for the model to run. If this happens, you can disable optimization by setting torch.onnx.export(..., do_constant_folding=False).
  4. Add an entry to the build.rs file to account for the generation of the new ONNX model.
  5. Add an entry to include_models! in tests/onnx_tests.rs to include the new ONNX model in the tests.
  6. Include a test in tests/onnx_tests.rs to test the new ONNX model.
  7. Run cargo test to ensure your test passes.