Open eaidova opened 1 week ago
as discussed in https://github.com/huggingface/optimum-intel/pull/781 openvino runtime implicitly converts int64 inputs into i32. For reducing overhead internal data conversion, convert models with int32 inputs.
There is no intention to merge, just checking for understanding impact
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.
What does this PR do?
as discussed in https://github.com/huggingface/optimum-intel/pull/781 openvino runtime implicitly converts int64 inputs into i32. For reducing overhead internal data conversion, convert models with int32 inputs.
There is no intention to merge, just checking for understanding impact
Before submitting