dp compress error

[Lewis-YL]

Hello,

I have problem compressing my model with fparam=1

Here is the error info:

DEEPMD INFO training with 1 frame parameter(s)
DEEPMD INFO training data with lower boundary: -0.22030017292961623
DEEPMD INFO training data with upper boundary: 57.81506323533096
Traceback (most recent call last):
File "/home/ptxlab/.local/bin/dp", line 8, in
sys.exit(main())
File "/home/ptxlab/.local/lib/python3.9/site-packages/deepmd/entrypoints/main.py", line 483, in main
compress(**dict_args)
File "/home/ptxlab/.local/lib/python3.9/site-packages/deepmd/entrypoints/compress.py", line 118, in compress
train(
File "/home/ptxlab/.local/lib/python3.9/site-packages/deepmd/entrypoints/train.py", line 106, in train
_do_work(jdata, run_opt, is_compress)
File "/home/ptxlab/.local/lib/python3.9/site-packages/deepmd/entrypoints/train.py", line 162, in _do_work
model.build(train_data, stop_batch)
File "/home/ptxlab/.local/lib/python3.9/site-packages/deepmd/train/trainer.py", line 300, in build
self.fitting.init_variables(self.model_param['compress']['model_file'])
TypeError: init_variables() missing 1 required positional argument: 'graph_def'

Is there a way to solve the problem?

My deepmd version is 2.1.0

[njzjz]

Did you change the code? graph_def is an argument in v2.1.1

[Lewis-YL]

Yes,

I had some issues with restart model before,

so I changed the code following #1692.

What should I do next to make compress work again?

[njzjz]

graph_def is provided by #1610. But it's a huge work.
I will suggest upgrading to v2.1.1 and then apply #1692.

[Lewis-YL]

Sounds good!

I will upgrade to v2.1.1 first and see whether it works.

[Lewis-YL]

Thank you, problem solved in 2.1.1