search

zye1996 / 3DSSD-torch

Apache License 2.0
16 stars 1 forks source link

TypeError: 'NoneType' object is not iterable #3

Closed Tocmac closed 3 years ago

Tocmac commented 3 years ago

Hello, I have successfully configured the environment, but when I trained on the server, I was reported the following error:

train: 100%|██████████| 441/441 [06:03<00:00, 1.39it/s, total_it=882] epochs: 1%| | 1/120 [12:11<12:10:14, 368.19s/it, total_loss=1.8] epochs: 1%| | 1/120 [12:11<12:10:14, 368.19s/it, total_loss=1.8]

eval: 0%| | 0/111 [00:00<?, ?it/s] epochs: 1%| | 1/120 [12:18<24:24:44, 738.53s/it, total_loss=1.8]

Traceback (most recent call last): File "lib/core/trainer.py", line 238, in cur_trainer.train() File "lib/core/trainer.py", line 194, in train self.val() File "lib/core/trainer.py", line 209, in val end_points = post_process(end_points) File "/data0/3DSSD-torch/lib/modeling/single_stage_detector.py", line 304, in post_process pred_score = end_points['post_pred_score'] KeyError: 'post_pred_score' Exception ignored in: <object repr() failed> Traceback (most recent call last): File "/usr/local/lib/python3.6/dist-packages/tqdm/std.py", line 1134, in del File "/usr/local/lib/python3.6/dist-packages/tqdm/std.py", line 1281, in close File "/usr/local/lib/python3.6/dist-packages/tqdm/std.py", line 1467, in display File "/usr/local/lib/python3.6/dist-packages/tqdm/std.py", line 1137, in repr File "/usr/local/lib/python3.6/dist-packages/tqdm/std.py", line 1424, in format_dict TypeError: 'NoneType' object is not iterable

I would appreciate it if you could give me some help.

zye1996 commented 3 years ago

I am sorry I made a mistake in trainer.py where the validation is performed. I have pushed the commit to comment those lines out and I will push another to do eval during training. Now you can only eval after finishing training.

Tocmac commented 3 years ago

I am sorry I made a mistake in trainer.py where the validation is performed. I have pushed the commit to comment those lines out and I will push another to do eval during training. Now you can only eval after finishing training.

Thank you so much !

zye1996 commented 3 years ago

I am sorry I made a mistake in trainer.py where the validation is performed. I have pushed the commit to comment those lines out and I will push another to do eval during training. Now you can only eval after finishing training.

Thank you so much !

codebase updated, issue closed