huggingface
Fix ORT CI
What does this PR do?
Fixes ort ci failures due to 1.18
Before submitting
- [ ] This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
- [ ] Did you make sure to update the documentation with your changes?
- [ ] Did you write any new necessary tests?
Who can review?
one test is still failing during generation with old model optimum/gpt2
https://github.com/huggingface/optimum/blob/main/tests/onnxruntime/test_modeling.py#L2277
run pytest -v -s tests/onnxruntime -k "test_inference_old_onnx_model_1" to see the error:
RuntimeError: Error in execution: Non-zero status code returned while running Where node. Name:'Where_393' Status Message: /onnxruntime_src/onnxruntime/core/providers/cpu/math/element_wise_ops.h:640 onnxruntime::Broadcaster::Broadcaster(gsl::span<const long int>, gsl::span<const long int>) largest <= 1 was false. Can broadcast 0 by 0 or 1. 5 is invalid.
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.
ort tests are passing locally, the issue with broadcasting was solved. but for some reason output ids are matching locally and not on the runner (two tests with old onnx model)
but for some reason output ids are matching locally and not on the runner (two tests with old onnx model)
Do you know where this could come from ?
