Hi all,
I have used the docker based setup in linux pc and tried to compile the yolov5l model provided by TI model zoo. I am successfully able to compile the model and also I am able to run the model using TIDL artifacts inside docker based setup for am69a device using my linux pc.
Now I copied the artifact folder onto cloud and tried to run the model using these artifacts on the device am69a using cloud service.
This is the main part of code I am running-
import onnxruntime as rt
onnx_model_path = '/home/root/notebooks/custom_models/yolov5l6_640_ti_lite_47p1_65p6.onnx'
delegate_options = {}
so = rt.SessionOptions()
delegate_options['artifacts_folder'] = '/home/root/notebooks/custom-artifacts/yolov5l/'
delegate_options.update(optional_options)
EP_list = ['TIDLExecutionProvider','CPUExecutionProvider']
sess = rt.InferenceSession(onnx_model_path ,providers=EP_list, provider_options=[delegate_options, {}], sess_options=so)
input_details = sess.get_inputs()
output_details = sess.get_outputs()
I am getting these errors when I am trying to run the model using onnxrt env-
RuntimeErrorTraceback (most recent call last)
/usr/local/lib/python3.6/dist-packages/onnxruntime/capi/onnxruntime_inference_collection.py in __init__(self, path_or_bytes, sess_options, providers, provider_options)
282 try:
--> 283 self._create_inference_session(providers, provider_options)
284 except RuntimeError:
/usr/local/lib/python3.6/dist-packages/onnxruntime/capi/onnxruntime_inference_collection.py in _create_inference_session(self, providers, provider_options)
314 # initialize the C++ InferenceSession
--> 315 sess.initialize_session(providers, provider_options)
316
RuntimeError: std::exception
During handling of the above exception, another exception occurred:
AttributeErrorTraceback (most recent call last)
<ipython-input-9-4e829141b539> in <module>
1 EP_list = ['TIDLExecutionProvider','CPUExecutionProvider']
----> 2 sess = rt.InferenceSession(onnx_model_path ,providers=EP_list, provider_options=[delegate_options, {}], sess_options=so)
3
4 input_details = sess.get_inputs()
5 output_details = sess.get_outputs()
/usr/local/lib/python3.6/dist-packages/onnxruntime/capi/onnxruntime_inference_collection.py in __init__(self, path_or_bytes, sess_options, providers, provider_options)
284 except RuntimeError:
285 if self._enable_fallback:
--> 286 print("EP Error using {}".format(self._providers))
287 print("Falling back to {} and retrying.".format(self._fallback_providers))
288 self._create_inference_session(self._fallback_providers, None)
AttributeError: 'InferenceSession' object has no attribute '_providers'
The model I picked is from here https://github.com/TexasInstruments/edgeai-yolov5/blob/master/pretrained_models/models/detection/coco/edgeai-yolov5/yolov5l6_640_ti_lite_47p1_65p6.onnx.link
Thanks
Akhilesh
