Failure to compress a model

[XTFan07]

I use 'module load deepmd/compress dp compress input.json -i 1.3-model.pb -o compressed-model.pb' to compress a model and it goes wrong. How can I adress this problem?

Traceback (most recent call last):
File "/data/share/apps/deepmd/compress/bin/dp", line 8, in
sys.exit(main())
File "/data/share/apps/deepmd/compress/lib/python3.8/site-packages/deepmd/main.py", line 327, in main
compress(**dict_args)
File "/data/share/apps/deepmd/compress/lib/python3.8/site-packages/deepmd/entrypoints/compress.py", line 108, in compress
freeze(checkpoint_folder=checkpoint_folder, output=output, node_names=None)
File "/data/share/apps/deepmd/compress/lib/python3.8/site-packages/deepmd/entrypoints/freeze.py", line 147, in freeze
input_graph_def = graph.as_graph_def()
File "/data/share/apps/deepmd/compress/lib/python3.8/site-packages/tensorflow/python/framework/ops.py", line 3294, in as_graph_def
result, _ = self._as_graph_def(from_version, add_shapes)
File "/data/share/apps/deepmd/compress/lib/python3.8/site-packages/tensorflow/python/framework/ops.py", line 3211, in _as_graph_def
graph.ParseFromString(compat.as_bytes(data))
google.protobuf.message.DecodeError: Error parsing message

[njzjz]

Please report the version you are using

[XTFan07]

OK. Thanks so much. I will have a try.

[XTFan07]

It works! Thanks!

[XTFan07]

Although the model has been successfully compressed, the MD simulation using the compressed model gives incorrect structures. Is this due to the lower accuracy of the smaller model size? How to set the step value more reasonably?

[felix5572]

Yes, try and error for now... In my experience. The MD trace will be nearly the same for at least the first 100 steps.(energy, pressure, temperature.the difference is within 1e-4 )

You could compare the lammps.log file with compressed and uncompressed model. (shouble be the same for the first frame in MD.).

[XTFan07]

OK. Thanks!

[njzjz]

This error cannot be fixed due to tensorflow limitations. But I'm adding a message in #839.