I am trying to covert onnx to dlc of depth-anything https://github.com/fabio-sim/Depth-Anything-ONNX
model can be found on this location https://github.com/fabio-sim/Depth-Anything-ONNX/releases/tag/v1.0.0
snpe-onnx-to-dlc --input_network depth_anything_vitb14.onnx --input_dim image "1,3,518,518" --out_node "depth" 2024-03-18 12:26:04,665 - 240 - WARNING - Couldn't import onnx-simplifier. (<class 'ModuleNotFoundError'>: No module named 'onnxsim')
2024-03-18 12:26:04,666 - 240 - WARNING - Install the onnx-simplifier for better model compatibility: "pip3 install onnx-simplifier"
2024-03-18 12:26:04,691 - 240 - WARNING - Try simplification again by skipping optimization
2024-03-18 12:26:04,713 - 240 - WARNING - Couldn't import onnx-simplifier. (<class 'ModuleNotFoundError'>: No module named 'onnxsim')
2024-03-18 12:26:04,713 - 240 - WARNING - Install the onnx-simplifier for better model compatibility: "pip3 install onnx-simplifier"
2024-03-18 12:26:04,736 - 240 - WARNING - Resume normal conversion with unsimplified model
2024-03-18 12:26:07,032 - 240 - WARNING - Onnx shape inference failed with unexpected exception
2024-03-18 12:26:07,127 - 240 - WARNING - Skip shape inference and resume normal conversion
2024-03-18 12:26:09,343 - 235 - INFO - INFO_STATIC_RESHAPE: Applying static reshape to /Concat_output_0: new name /Reshape_output_0 new shape [-1]
2024-03-18 12:26:09,346 - 240 - WARNING - WARNING_CAST_TYPE: Only numerical type cast is supported. The op: /Cast will be interpreted at conversion time
2024-03-18 12:26:09,346 - 240 - WARNING - WARNING_CAST_TYPE: Only numerical type cast is supported. The op: /Cast_1 will be interpreted at conversion time
2024-03-18 12:26:09,347 - 240 - WARNING - WARNING_CAST_TYPE: Only numerical type cast is supported. The op: /Cast_2 will be interpreted at conversion time
2024-03-18 12:26:09,347 - 240 - WARNING - WARNING_CAST_TYPE: Only numerical type cast is supported. The op: /Cast_3 will be interpreted at conversion time
2024-03-18 12:26:09,353 - 235 - INFO - INFO_STATIC_RESHAPE: Applying static reshape to onnx::Reshape_1463: new name /Reshape_1_output_0 new shape [1, 37, 37, 768]
2024-03-18 12:26:09,358 - 235 - INFO - INFO_STATIC_RESHAPE: Applying static reshape to /Reshape_1_output_0: new name /Transpose_output_0 new shape (1, 768, 37, 37)
2024-03-18 12:26:09,366 - 240 - WARNING - WARNING_CAST_TYPE: Only numerical type cast is supported. The op: /Cast_4 will be interpreted at conversion time
Traceback (most recent call last):
File "/home/test/tools/snpe/snpe-2.17.0.231124/lib/python/qti/aisw/converters/onnx/onnx_to_ir.py", line 445, in convert
self.translations.apply_method_to_op(src_type,
File "/home/test/tools/snpe/snpe-2.17.0.231124/lib/python/qti/aisw/converters/common/converter_ir/translation.py", line 51, in apply_method_to_op
return translation.apply_method(method_name, *args, **kwargs)
File "/home/test/tools/snpe/snpe-2.17.0.231124/lib/python/qti/aisw/converters/common/converter_ir/translation.py", line 18, in apply_method
return self.indexed_methods[method_name](*args, **kwargs)
File "/home/test/tools/snpe/snpe-2.17.0.231124/lib/python/qti/aisw/converters/common/converter_ir/translation.py", line 123, in add_op
op = self.extract_parameters(src_op, context)
File "/home/test/tools/snpe/snpe-2.17.0.231124/lib/python/qti/aisw/converters/onnx/data_translations.py", line 1431, in extract_parameters
input_buf = graph.get_buffer(str(src_op.input[0]))
File "/home/test/tools/snpe/snpe-2.17.0.231124/lib/python/qti/aisw/converters/common/converter_ir/op_graph.py", line 1551, in get_buffer
return self.buffers[buffer_name] if buffer_name else self.null_buffer
KeyError: '/Transpose_output_0'
2024-03-18 12:26:09,369 - 230 - ERROR - Node /Resize: '/Transpose_output_0'
looking foward for the help what can be the error can you please explain how to resolve it.
thanks in advance.