You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The new_operator.md documentation has a section "2. Generate the C++ template code of the Custom operator from the ONNX Model (optional)" which states that a template can be generated with python -m onnxruntime-extensions.cmd --cpp-gen <model_path> <repository_dir>
@thiagocrepaldi , this part of doc is kind of out-of-date. Given the last the custom op already supported a C++ function as an op kernel, The auto-gen C++ template code seems unnecessary now.
@wenbingl thanks, is there anyone who could help update the documentation with the correct steps? Maybe @natke? Hardware partners like AMD could leverage this information
By the way, i remember there was a slack channel for ORT partners/vendors. Would you know what is the slack workspace/channel for me to join? Thanks
The new_operator.md documentation has a section "2. Generate the C++ template code of the Custom operator from the ONNX Model (optional)" which states that a template can be generated with
python -m onnxruntime-extensions.cmd --cpp-gen <model_path> <repository_dir>
The problem is that onnxruntime_extensions/cmd.py doesn't have
--cpp-gen
optionWas it deleted intentionally? Is there an easy way to have a template?
The text was updated successfully, but these errors were encountered: