deep-head-pose
deep-head-pose copied to clipboard
natanielruiz
torch version on gpu
the version of torch must >=1.1.0 when run on gpu? with torch=0.4.0(cuda=0.8.61) ,I made an error 'Aborted(core dump)' but on cpu is well
I tried both 0.4.1 and 1.1 (of PyTorch) and it works on both with very minor changes in the code.
I tried both 0.4.1 and 1.1 (of PyTorch) and it works on both with very minor changes in the code.
@shahroudy hi, could you help tell where to change? when use 1.1 pytorch?thanks!
I tried both 0.4.1 and 1.1 (of PyTorch) and it works on both with very minor changes in the code.
@shahroudy hi, could you help tell where to change? when use 1.1 pytorch?thanks!
You can find it in my forked repository.
@shahroudy I was able to use your fork with Pytorch 1.3, but I did run into the following error that was fixed with casting...
Environment: PyTorch version: 1.3.0 Is debug build: No CUDA used to build PyTorch: 10.1.243
OS: Ubuntu 16.04.6 LTS GCC version: (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609 CMake version: version 3.5.1
Python version: 3.6 Is CUDA available: Yes CUDA runtime version: 10.1.243 GPU models and configuration: GPU 0: Tesla K80 Nvidia driver version: 418.87.00
Error:
(dl4cvpy36-pytorch1.3) ubuntu@ip-172-31-0-76:~/cviz/deep-head-pose$ python code/test_on_video_dlib.py --snapshot models/hopenet_snapshot_a1.pkl --face_model mmod_human_face_detector.dat --video 2019-11-26-173255.webm.mp4 --n_frames 100 --fps 15
Loading snapshot.
Loading data.
/home/ubuntu/.virtualenvs/dl4cvpy36-pytorch1.3/lib/python3.6/site-packages/torchvision/transforms/transforms.py:211: UserWarning: The use of the transforms.Scale transform is deprecated, please use transforms.Resize instead.
"please use transforms.Resize instead.")
Ready to test network.
1
Traceback (most recent call last):
File "code/test_on_video_dlib.py", line 171, in
Fix: img = cv2_frame[int(y_min):int(y_max),int(x_min):int(x_max)]
@shahroudy I was able to use your fork with Pytorch 1.3, but I did run into the following error that was fixed with casting...
Environment: PyTorch version: 1.3.0 Is debug build: No CUDA used to build PyTorch: 10.1.243
OS: Ubuntu 16.04.6 LTS GCC version: (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609 CMake version: version 3.5.1
Python version: 3.6 Is CUDA available: Yes CUDA runtime version: 10.1.243 GPU models and configuration: GPU 0: Tesla K80 Nvidia driver version: 418.87.00
Error: (dl4cvpy36-pytorch1.3) ubuntu@ip-172-31-0-76:~/cviz/deep-head-pose$ python code/test_on_video_dlib.py --snapshot models/hopenet_snapshot_a1.pkl --face_model mmod_human_face_detector.dat --video 2019-11-26-173255.webm.mp4 --n_frames 100 --fps 15 Loading snapshot. Loading data. /home/ubuntu/.virtualenvs/dl4cvpy36-pytorch1.3/lib/python3.6/site-packages/torchvision/transforms/transforms.py:211: UserWarning: The use of the transforms.Scale transform is deprecated, please use transforms.Resize instead. "please use transforms.Resize instead.") Ready to test network. 1 Traceback (most recent call last): File "code/test_on_video_dlib.py", line 171, in img = cv2_frame[y_min:y_max,x_min:x_max] TypeError: slice indices must be integers or None or have an index method
Fix: img = cv2_frame[int(y_min):int(y_max),int(x_min):int(x_max)]
Your result is so great! How many epochs did you train? I trained with 100 epochs but then I had got the same result of (yaw, pitch , roll) on different inputs images
