PyTorch convert function for op 'norm' not implemented.

[vivisl]

❓PyTorch convert function for op 'norm' not implemented.

Problem description

• Hi, I am using this depth estimation algorithm BTS to predict my pics and it works well on my Ubuntu 18,04.
• But I have encountered a problem when I try to convert the pytorch model to coreml model:

mlmodel = ct.convert(traced_model, inputs=[input])

• Hope for all of your replies. THANKS A LOT !

System Info

• python 3.7
• opencv 4.4
• pytorch 1.6
• coremltools 4.0

Converting code

import coremltools as ct
example_input = torch.rand(1, 3, 256, 512).cuda()
traced_model = torch.jit.trace(model.module, example_input)
# input = ct.ImageType(shape=(1, 3, 256, 512), bias=[-1,-1,-1], scale=1/127)
input = ct.TensorType(name='bts_input', shape=(1,3,256,512))
mlmodel = ct.convert(traced_model, inputs=[input])
mlmodel.save("bts_coremltool.mlmodel")

Error info

Converting Frontend ==> MIL Ops:  76%|█████████████████▍     | 1581/2084 [00:01<00:00, 1302.10 ops/s]
Traceback (most recent call last):
  File "bts_test.py", line 289, in <module>
    test(args)
  File "bts_test.py", line 185, in test
    mlmodel = ct.convert(traced_model, inputs=[input])
  File "/home/vivi/anaconda3/envs/pt16/lib/python3.7/site-packages/coremltools/converters/_converters_entry.py", line 183, in convert
    **kwargs
  File "/home/vivi/anaconda3/envs/pt16/lib/python3.7/site-packages/coremltools/converters/mil/converter.py", line 129, in mil_convert
    ConverterRegistry, **kwargs)
  File "/home/vivi/anaconda3/envs/pt16/lib/python3.7/site-packages/coremltools/converters/mil/converter.py", line 171, in mil_convert_to_proto
    prog = frontend_converter(model, **kwargs)
  File "/home/vivi/anaconda3/envs/pt16/lib/python3.7/site-packages/coremltools/converters/mil/converter.py", line 85, in __call__
    return load(*args, **kwargs)
  File "/home/vivi/anaconda3/envs/pt16/lib/python3.7/site-packages/coremltools/converters/mil/frontend/torch/load.py", line 83, in load
    raise e
  File "/home/vivi/anaconda3/envs/pt16/lib/python3.7/site-packages/coremltools/converters/mil/frontend/torch/load.py", line 75, in load
    prog = converter.convert()
  File "/home/vivi/anaconda3/envs/pt16/lib/python3.7/site-packages/coremltools/converters/mil/frontend/torch/converter.py", line 224, in convert
    convert_nodes(self.context, self.graph)
  File "/home/vivi/anaconda3/envs/pt16/lib/python3.7/site-packages/coremltools/converters/mil/frontend/torch/ops.py", line 53, in convert_nodes
    "PyTorch convert function for op '{}' not implemented.".format(node.kind)
RuntimeError: PyTorch convert function for op 'norm' not implemented.

[AtomLee4279]

I encountered the same problem.
Has anyone solved it?

[qqpann]

+1 on this

[TobyRoseman]

With the release of coremltools 5.1, we now support norm.