Audio-driven-TalkingFace-HeadPose model.bind(data_shapes=[('data', (1, 3, image_size[0], image_size[1]))]) int Step V (Fine-tune the gan network) failed.

model.bind(data_shapes=[('data', (1, 3, image_size[0], image_size[1]))]) int Step V (Fine-tune the gan network) failed.

Open qianxinchun opened this issue 4 years ago • 1 comments

I encountered an error while running the step v: cd render-to-video/ python train_19news_1.py [person_id] [gpu_id]

The program was killed by the system when running the following line: model.bind(data_shapes=[('data', (1, 3, image_size[0], image_size[1]))])

Nov 16 '20 02:11 qianxinchun

The monitoring system says CPU memory-leak happened when running the above line. But I have already set gpu_id=0.

Nov 16 '20 04:11 qianxinchun