Error in model Compressing: google.protobuf.message.DecodeError: Error parsing message

[Epikurz]

DP developers:

The NN potential has been trained in two round. In the first round, the model has been compressed sucessfully. Then the model was
trained wiht "dp train -i model.ckpt". An error raised in compressing the second model as follows:

DEEPMD INFO stage 2: freeze the model
Traceback (most recent call last):
File "/home/user/deepmd-kit_cpu/bin/dp", line 10, in
sys.exit(main())
File "/home/user/deepmd-kit_cpu/lib/python3.9/site-packages/deepmd/entrypoints/main.py", line 447, in main
compress(**dict_args)
File "/home/user/deepmd-kit_cpu/lib/python3.9/site-packages/deepmd/entrypoints/compress.py", line 141, in compress
freeze(checkpoint_folder=checkpoint_folder, output=output, node_names=None)
File "/home/user/deepmd-kit_cpu/lib/python3.9/site-packages/deepmd/entrypoints/freeze.py", line 173, in freeze
input_graph_def = graph.as_graph_def()
File "/home/user/deepmd-kit_cpu/lib/python3.9/site-packages/tensorflow/python/framework/ops.py", line 3374, in as_graph_def
result, _ = self._as_graph_def(from_version, add_shapes)
File "/home/user/deepmd-kit_cpu/lib/python3.9/site-packages/tensorflow/python/framework/ops.py", line 3291, in _as_graph_def
graph.ParseFromString(compat.as_bytes(data))
google.protobuf.message.DecodeError: Error parsing message

Is it a bug in compress module? May be I can send my model to someone by email?

[njzjz]

See #787 and #790. What's your version? A message has been added there.

[Epikurz]

Both gpu and cpu 2.03 version have been applied.
According to #790, I confirmed this error was caused by quite large "-s" parameter. The model can be compressed successfully with 0.03.

[njzjz]

Surprised to see you get the same error from a place different from #790... I'll also add some tips here.