optimum
2742bf14 - Add the ability to choose the ONNX runtime execution provider in `ORTModel` (#137)

Commit

3 years ago

Add the ability to choose the ONNX runtime execution provider in `ORTModel` (#137) * added option for onnxruntime execution provider * formatting * better description * changed ort provider to model arguments * added documentation * changed ort provider name * formatting * remove wrong files * trigger actions * added error catch in case the given arguments for cpu-gpu optimization are contradictory * remove unused files * correct wrong catch * styling Co-authored-by: Felix Marty <felix@huggingface.co>

References

#137 - Add the ability to choose the ONNX runtime execution provider in `ORTModel`

Author

fxmarty

Parents

d36d6ee4

Files7

examples/onnxruntime
- optimization
  - question-answering
    - run_qa.py
  - text-classification
    - run_glue.py
  - token-classification
    - run_ner.py
- quantization
  - question-answering
    - run_qa.py
  - text-classification
    - run_glue.py
  - token-classification
    - run_ner.py
optimum/onnxruntime
- model.py

optimum 2742bf14 - Add the ability to choose the ONNX runtime execution provider in `ORTModel` (#137)

optimum
2742bf14 - Add the ability to choose the ONNX runtime execution provider in `ORTModel` (#137)