Error in postprocess function

        # Get input and output binding indices
        input_binding_idx = self.engine.get_binding_index('images')
        output_binding_num_idx = self.engine.get_binding_index('num')
        output_binding_boxes_idx = self.engine.get_binding_index('boxes')
        output_binding_scores_idx = self.engine.get_binding_index('scores')
        output_binding_classes_idx = self.engine.get_binding_index('classes')

        # Get input and output shapes
        input_shape = self.engine.get_binding_shape(input_binding_idx)
        output_shape_num = self.engine.get_binding_shape(output_binding_num_idx)
        output_shape_boxes = self.engine.get_binding_shape(output_binding_boxes_idx)
        output_shape_scores = self.engine.get_binding_shape(output_binding_scores_idx)
        output_shape_classes = self.engine.get_binding_shape(output_binding_classes_idx)

        # If buff is of type torch.Tensor, convert to a numpy array
        if isinstance(buff, torch.Tensor):
            buff = buff.cpu().numpy()

        # Allocate input and output memory
        d_input = cuda.mem_alloc(buff.nbytes)
        d_output_num = cuda.mem_alloc(trt.volume(output_shape_num) * buff.dtype.itemsize)
        d_output_boxes = cuda.mem_alloc(trt.volume(output_shape_boxes) * buff.dtype.itemsize)
        d_output_scores = cuda.mem_alloc(trt.volume(output_shape_scores) * buff.dtype.itemsize)
        d_output_classes = cuda.mem_alloc(trt.volume(output_shape_classes) * buff.dtype.itemsize)
        bindings = [int(d_input), int(d_output_num), int(d_output_boxes), int(d_output_scores), int(d_output_classes)]


        # If buff is of type torch.Tensor, convert to a numpy array
        if isinstance(buff, torch.Tensor):
            buff = buff.cpu().numpy()

        # Check buff type and size
        print(f"buff type: {type(buff)}")
        print(f"buff shape: {buff.shape}")


        # Copy input data to GPU
        cuda.memcpy_htod(d_input, buff)


        # Copy input data to GPU
        # cuda.memcpy_htod(d_input, buff.numpy())


        # Secure output data
        output_num = np.empty(output_shape_num, dtype=trt.nptype(self.engine.get_binding_dtype(output_binding_num_idx)))
        output_boxes = np.empty(output_shape_boxes, dtype=trt.nptype(self.engine.get_binding_dtype(output_binding_boxes_idx)))
        output_scores = np.empty(output_shape_scores, dtype=trt.nptype(self.engine.get_binding_dtype(output_binding_scores_idx)))
        output_classes = np.empty(output_shape_classes, dtype=trt.nptype(self.engine.get_binding_dtype(output_binding_classes_idx)))

        # Secure output data
        self.context.execute_v2(bindings)

        # Copy output data from GPU to host
        cuda.memcpy_dtoh(output_num, d_output_num)
        cuda.memcpy_dtoh(output_boxes, d_output_boxes)
        cuda.memcpy_dtoh(output_scores, d_output_scores)
        cuda.memcpy_dtoh(output_classes, d_output_classes)

        # Post-processing of inference results
        outputs = postprocess(
            output_num,
            output_boxes,
            output_scores,
            output_classes,
            self.exp.num_classes,
            self.exp.test_conf,
            self.exp.nmsthre,
            class_agnostic=True
        )

When executing the postprocess function in the “Post-processing inference results” section of this program, I get the exception “postprocess() got multiple values for argument ‘class_agnostic’”.
I have tried everything, but cannot solve the problem.
outputs = postprocess(
output_num, output_boxes, output_boxes
output_boxes, output_scores, output_boxes
output_scores, output_boxes, output_classes, output_scores
output_classes, self.exp.num_classes, self.
self.exp.num_classes, self.exp.test_conf,
self.exp.test_conf, self.exp.test_conf,
self.exp.nmsthre, self.exp.test_conf, self.exp.nmsthre
True
)
and ,
outputs = postprocess(
output_boxes=output_boxes, output_boxes
output_scores=output_scores, output_classes=output_classes, output_classes
output_classes=output_classes, output_classes
num_classes=self.exp.num_classes,
conf_thre=self.exp.test_conf,
nms_thre=self.exp.nmsthre,
class_agnostic=True
)
I have tried the following, but the error persists.
I would appreciate it if you could answer my question.
Thank you in advance.

Dear @gecreative.kasa,
Does this issue still need support? May I know the function definition signature?

Thank you for your answer. I was able to solve the issue on my own. Thank you very much. If I have any other questions, I hope to ask for your advice again.

Thanks.

Dear @gecreative.kasa,
Could you share what was the mistake to help others in community.