Exception: expecting input to have 3 dimensions, model 'densenet_onnx' input has 4

[mahesh11T]

python /workspace/client/src/python/examples/image_client.py image_filename=/workspace/images/mug.jpg -m densenet_onnx -b=0 -s=INCEPTION -v

    if len(input_metadata.shape) != expected_input_dims:
        raise Exception(
            "expecting input to have {} dimensions, model '{}' input has {}".format(
                expected_input_dims, model_metadata.name, len(input_metadata.shape)
            )
        )

line 116 in image_client.py is expecting (3,2..x, 2.3xx) shape, but its getting (1,3,2.x,2.x) , y is this happening?

[Yobol]

Hi. You can modify 'max_batch_size param' in config.pbtxt of densenet_onnx. Default value is 0. It must be more than 0. https://github.com/triton-inference-server/server/blob/main/docs/examples/model_repository/densenet_onnx/config.pbtxt

[mahesh11T]

Hi. You can modify 'max_batch_size param' in config.pbtxt of densenet_onnx. Default value is 0. It must be more than 0. https://github.com/triton-inference-server/server/blob/main/docs/examples/model_repository/densenet_onnx/config.pbtxt

I had set it to 1 itself and it was giving this error.

[Yobol]

I encountered this error yesterday. I solve it because I found config.pbtxt was missing.