[Performance] yolov5-segmentation.Onnx Static Quantization about onnxruntime HOT 5 OPEN

I tried to replicate your issue but it works for me with onnxruntime from main branch. I following the tutorial here to get the model you are using: https://github.com/ultralytics/yolov5?tab=readme-ov-file.

from onnxruntime.

Could you recall the precise steps you took to quantize the YOLO library , Plz !

from onnxruntime.

I use the command python export.py --weights yolov5s.pt --include torchscript onnx from the repo I linked above then I applied the same command line you used. I did it with the current development version of onnxruntime.

from onnxruntime.

Initially, I appreciate your response. However, when I tried to use yolov5-seg.onnx, I encountered an error.

After exporting yolov5n-seg.pt to the ONNX format :

!python export.py --weights yolov5n-seg.pt --include torchscript onnx

I attempted to preprocess it for static quantization.
Unfortunately, I encountered an error during this process.
Traceback (most recent call last): File "<frozen runpy>", line 198, in _run_module_as_main File "<frozen runpy>", line 88, in _run_code File "G:\anaconda3\Lib\site-packages\onnxruntime\quantization\preprocess.py", line 127, in <module> quant_pre_process( File "G:\anaconda3\Lib\site-packages\onnxruntime\quantization\shape_inference.py", line 71, in quant_pre_process model = SymbolicShapeInference.infer_shapes( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "G:\anaconda3\Lib\site-packages\onnxruntime\tools\symbolic_shape_infer.py", line 2855, in infer_shapes all_shapes_inferred = symbolic_shape_inference._infer_impl() ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "G:\anaconda3\Lib\site-packages\onnxruntime\tools\symbolic_shape_infer.py", line 2644, in _infer_impl out_rank = len(get_shape_from_type_proto(vi.type)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "G:\anaconda3\Lib\site-packages\onnxruntime\tools\symbolic_shape_infer.py", line 37, in get_shape_from_type_proto assert not is_sequence(type_proto) ^^^^^^^^^^^^^^^^^^^^^^^ File "G:\anaconda3\Lib\site-packages\onnxruntime\tools\symbolic_shape_infer.py", line 32, in is_sequence assert cls_type in ["tensor_type", "sequence_type"] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ AssertionError

As I have limited experience with ONNX and quantization, and I'm still in the early stages of learning, I lack the expertise to resolve these errors. Your help would be greatly appreciated.

from onnxruntime.

This issue has been automatically marked as stale due to inactivity and will be closed in 30 days if no further activity occurs. If further support is needed, please provide an update and/or more details.

from onnxruntime.