err when use 'dp freeze'

[l98y3j26]

After trainng my data, I use 'dp freeze' and it goes wrong. How can I adress this problem?

Traceback (most recent call last):
File "/public1/home/sc80397/.conda/envs/EDTA/bin/dp", line 10, in
sys.exit(main())
File "/public1/home/sc80397/.conda/envs/EDTA/lib/python3.6/site-packages/deepmd/main.py", line 75, in main
freeze(args)
File "/public1/home/sc80397/.conda/envs/EDTA/lib/python3.6/site-packages/deepmd/freeze.py", line 89, in freeze
freeze_graph(args.folder, args.output, args.nodes)
File "/public1/home/sc80397/.conda/envs/EDTA/lib/python3.6/site-packages/deepmd/freeze.py", line 41, in freeze_graph
input_checkpoint = checkpoint.model_checkpoint_path
AttributeError: 'NoneType' object has no attribute 'model_checkpoint_path'

[njzjz]

Did you fail training or stop training before the first checkpoint was saved?

[l98y3j26]

thanks！