the GPU usage

[Along-wang1]

Hello！ Could you please explain why the GPU usage remains at 0% during model training?

[Along-wang1]

CUDA_VISIBLE_DEVICES=0 python main_HTNet.py lr=0.000050, epochs=800, device=cuda 442 ['037', '035', '023', '011', 's11', 'sub02', 's01', 'sub05', 'sub14', 'sub11', '019', '013', '028', 's12', 's19', 'sub25', 'sub22', 's09', '033', 'sub26', 's18', '030', '021', 'sub15', '009', 'sub04', '016', 'sub16', '018', 's02', 'sub23', '020', 's03', 's15', '026', 's06', '015', 'sub19', '010', '036', '006', 'sub21', '024', 'sub17', 'sub24', 's13', 'sub13', 's14', 's20', '012', 'sub06', 'sub01', '031', '022', 'sub20', 'sub12', '032', 'sub07', 's04', '017', 's08', 'sub09', 'sub03', '007', 'sub08', 's05', '034', '014'] Subject: 037 train Best Predicted : [0] Ground Truth : [0] Evaluation until this subject:

This is the instruction I called to train the model and some of the output

[wangzhifengharrison]

You can use this command 'python main_HTNet.py --train True'

[Along-wang1]

Thank you very much for your response. Additionally, I would like to inquire about the duration of training for my code, which takes over 40 hours for a single training session. Is this duration considered normal? How long does it typically take for you to train the model for one iteration?

[wangzhifengharrison]

Using 3090 for training process normally costs 10 hours.

[Along-wang1]

I sincerely thank you for your willingness to dedicate time to address my questions. Your assistance is greatly appreciated.