You can find more detailed server logs here: http://perklabseg.asuscomm.com:5000/logs/
[2021-10-28 04:20:36] [INFO] (aiaa.www.api.api_v1) - ===============================================================================
[2021-10-28 04:20:36] [INFO] (aiaa.www.api.api_v1) - API_v1:: /v1/inference (model: clara_pt_prostate_mri_segmentation)
[2021-10-28 04:20:36] [INFO] (aiaa.www.api.api_v1) - ===============================================================================
[2021-10-28 04:20:37] [INFO] (aiaa.inference.inference_utils) - PRE - Run Transforms
[2021-10-28 04:20:37] [INFO] (aiaa.inference.inference_utils) - PRE - Input Keys: dict_keys(['image', 'image_path', 'params', 'transform_meta'])
[2021-10-28 04:20:37] [INFO] (aiaa.inference.inference_utils) - PRE - Transform (LoadImaged): Time: 0.0883; image: (512, 512, 27)
[2021-10-28 04:20:37] [INFO] (aiaa.inference.inference_utils) - PRE - Transform (AddChanneld): Time: 0.0000; image: (1, 512, 512, 27)
[2021-10-28 04:20:37] [INFO] (aiaa.inference.inference_utils) - PRE - Transform (Spacingd): Time: 0.1367; image: (1, 161, 161, 79)
[2021-10-28 04:20:37] [INFO] (aiaa.inference.inference_utils) - PRE - Transform (ScaleIntensityRanged): Time: 0.0055; image: (1, 161, 161, 79)
[2021-10-28 04:20:37] [INFO] (aiaa.inference.triton_inference) - Input Shape: (1, 1, 161, 161, 79), Type: ndarray
[2021-10-28 04:20:37] [ERROR] (aiaa.www.api.api_v1) - unexpected shape for input 'INPUT__0' for model 'clara_pt_prostate_mri_segmentation'. Expected [-1,2,192,192,64], got [1,1,192,192,64]
[2021-10-28 04:20:37] Traceback (most recent call last):
[2021-10-28 04:20:37] File "/opt/conda/lib/python3.8/site-packages/flask/app.py", line 1950, in full_dispatch_request
[2021-10-28 04:20:37] rv = self.dispatch_request()
[2021-10-28 04:20:37] File "/opt/conda/lib/python3.8/site-packages/flask/app.py", line 1936, in dispatch_request
[2021-10-28 04:20:37] return self.view_functions[rule.endpoint](**req.view_args)
[2021-10-28 04:20:37] File "www/api/api_v1.py", line 368, in api_v1_inference
[2021-10-28 04:20:37] File "www/api/api_v1.py", line 255, in run_inference
[2021-10-28 04:20:37] File "www/api/api_v1.py", line 168, in run_infer
[2021-10-28 04:20:37] File "actions/inference_engine.py", line 80, in run
[2021-10-28 04:20:37] File "actions/inference_engine.py", line 141, in _run_inference
[2021-10-28 04:20:37] File "inference/triton_inference.py", line 74, in inference
[2021-10-28 04:20:37] File "inference/triton/inference.py", line 260, in sw_inference
[2021-10-28 04:20:37] File "/opt/conda/lib/python3.8/site-packages/tritonclient/http/__init__.py", line 1286, in get_result
[2021-10-28 04:20:37] _raise_if_error(response)
[2021-10-28 04:20:37] File "/opt/conda/lib/python3.8/site-packages/tritonclient/http/__init__.py", line 63, in _raise_if_error
[2021-10-28 04:20:37] raise error
[2021-10-28 04:20:37] tritonclient.utils.InferenceServerException: unexpected shape for input 'INPUT__0' for model 'clara_pt_prostate_mri_segmentation'. Expected [-1,2,192,192,64], got [1,1,192,192,64]
The prostate model probably fails because it requires two input images (T2 and ADC) and most likely the AIAA client that NVidia developers created for Slicer does not support this yet.
Please submit a feature request to NVidia developers at Issues · NVIDIA/ai-assisted-annotation-client · GitHub and copy the link here.