Use my own data for training, but there is no loss value about deep-motion-editing HOT 4 CLOSED

You can use --verbose=1 to see if training is really working. Regarding the tensorboard problem, I'm not quite sure what happened. Maybe check if the logs directory is empty and make sure you are using the correct path for tensorboard.

from deep-motion-editing.

Thank you very much for your reply. I have solved this problem.
But I use the new training model to test, there are some errors, how to solve it?

loading from ./pretrained/models/topology0
loading from epoch 2000......
load succeed!
loading from ./pretrained/models/topology1
loading from epoch 2000......
load succeed!
Traceback (most recent call last):
File "evel_single_pair_1.py", line 98, in
main()
File "evel_single_pair_1.py", line 92, in main
model.test()
File "/home/wuxiaoliang/docker/motion_retarget/deep-motion-editing/retargeting/models/base_model.py", line 95, in test
self.forward()
File "/home/wuxiaoliang/docker/motion_retarget/deep-motion-editing/retargeting/models/architecture.py", line 103, in forward
latent, res = self.models[i].auto_encoder(motion, offsets)
File "/home/wuxiaoliang/anaconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 550, in call
result = self.forward(*input, **kwargs)
File "/home/wuxiaoliang/docker/motion_retarget/deep-motion-editing/retargeting/models/enc_and_dec.py", line 133, in forward
latent = self.enc(input, offset)
File "/home/wuxiaoliang/anaconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 550, in call
result = self.forward(*input, **kwargs)
File "/home/wuxiaoliang/docker/motion_retarget/deep-motion-editing/retargeting/models/enc_and_dec.py", line 66, in forward
input = layer(input)
File "/home/wuxiaoliang/anaconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 550, in call
result = self.forward(*input, **kwargs)
File "/home/wuxiaoliang/anaconda3/lib/python3.7/site-packages/torch/nn/modules/container.py", line 100, in forward
input = module(input)
File "/home/wuxiaoliang/anaconda3/lib/python3.7/site-packages/torch/nn/modules/module.py", line 550, in call
result = self.forward(*input, **kwargs)
File "/home/wuxiaoliang/docker/motion_retarget/deep-motion-editing/retargeting/models/skeleton.py", line 95, in forward
res = F.conv1d(F.pad(input, self._padding_repeated_twice, mode=self.padding_mode),
File "/home/wuxiaoliang/anaconda3/lib/python3.7/site-packages/torch/nn/functional.py", line 3397, in _pad
return torch._C._nn.reflection_pad1d(input, pad)
RuntimeError: Argument #4: Padding size should be less than the corresponding input dimension, but got: padding (7, 7) at dimension 2 of input [1, 112, 7]
@PeizhuoLi

from deep-motion-editing.

Hi, sorry for the inconvenience. The default hyper-parameter of number of layers was mistakenly set to 3. It should be 2. You can add --num_layer=2 in training parameter or pull the newest code and retrain your model.

from deep-motion-editing.

Thank you very much for your answer to my question。

from deep-motion-editing.