When T try to run 'python train.py' ,upto 50% training processing working well,now here is the error:
(Pytorch) C:\Users\mujta\Desktop\Pytorch>python train.py
loading annotations into memory...
Done (t=6.68s)
creating index...
index created!
loading annotations into memory...
Done (t=0.26s)
creating index...
index created!
train images: 64115
val images: 2693
50%|████████████████████████████████████▍ | 8014/16029 [3:37:20<2:39:58, 1.20s/it]C:\Users\mujta\Desktop\Pytorch\openpose.py:417: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use arr[tuple(seq)] instead of arr[seq]. In the future this will be interpreted as an array index, arr[np.array(seq)], which will result either in an error or a different result.
paf_in_edge = np.hstack([paf[0][np.hsplit(integ_points, 2)], paf[1][np.hsplit(integ_points, 2)]])
50%|████████████████████████████████████▍ | 8014/16029 [3:40:22<3:40:24, 1.65s/it]
Traceback (most recent call last):
File "train.py", line 16, in
openpose.train()
File "C:\Users\mujta\Desktop\Pytorch\openpose.py", line 287, in train
self.save_state(val_loss)
File "C:\Users\mujta\Desktop\Pytorch\openpose.py", line 105, in savestate
('model{}_val_loss:{}_step:{}.pth'.format(time, val_loss, self.step)))
File "C:\Users\mujta\anaconda3\envs\Pytorch\lib\site-packages\torch\serialization.py", line 376, in save
with _open_file_like(f, 'wb') as opened_file:
File "C:\Users\mujta\anaconda3\envs\Pytorch\lib\site-packages\torch\serialization.py", line 230, in _open_file_like
return _open_file(name_or_buffer, mode)
File "C:\Users\mujta\anaconda3\envs\Pytorch\lib\site-packages\torch\serialization.py", line 211, in init
super(_open_file, self).init(open(name, mode))
OSError: [Errno 22] Invalid argument: 'work_space\model\model_2021-10-15-23-34_val_loss:0.02189785609021783_step:8014.pth'
before 50% is dine well,but why showing OSError? How to solve this issue? anyone help please!
When T try to run 'python train.py' ,upto 50% training processing working well,now here is the error: (Pytorch) C:\Users\mujta\Desktop\Pytorch>python train.py loading annotations into memory... Done (t=6.68s) creating index... index created! loading annotations into memory... Done (t=0.26s) creating index... index created! train images: 64115 val images: 2693 50%|████████████████████████████████████▍ | 8014/16029 [3:37:20<2:39:58, 1.20s/it]C:\Users\mujta\Desktop\Pytorch\openpose.py:417: FutureWarning: Using a non-tuple sequence for multidimensional indexing is deprecated; use arr[tuple(seq)] instead of arr[seq]. In the future this will be interpreted as an array index, arr[np.array(seq)], which will result either in an error or a different result. paf_in_edge = np.hstack([paf[0][np.hsplit(integ_points, 2)], paf[1][np.hsplit(integ_points, 2)]]) 50%|████████████████████████████████████▍ | 8014/16029 [3:40:22<3:40:24, 1.65s/it] Traceback (most recent call last): File "train.py", line 16, in
openpose.train()
File "C:\Users\mujta\Desktop\Pytorch\openpose.py", line 287, in train
self.save_state(val_loss)
File "C:\Users\mujta\Desktop\Pytorch\openpose.py", line 105, in savestate
('model{}_val_loss:{}_step:{}.pth'.format(time, val_loss, self.step)))
File "C:\Users\mujta\anaconda3\envs\Pytorch\lib\site-packages\torch\serialization.py", line 376, in save
with _open_file_like(f, 'wb') as opened_file:
File "C:\Users\mujta\anaconda3\envs\Pytorch\lib\site-packages\torch\serialization.py", line 230, in _open_file_like
return _open_file(name_or_buffer, mode)
File "C:\Users\mujta\anaconda3\envs\Pytorch\lib\site-packages\torch\serialization.py", line 211, in init
super(_open_file, self).init(open(name, mode))
OSError: [Errno 22] Invalid argument: 'work_space\model\model_2021-10-15-23-34_val_loss:0.02189785609021783_step:8014.pth'
before 50% is dine well,but why showing OSError? How to solve this issue? anyone help please!