When I run isonet.py predict tomoset.star demo_results/model_iter30.h5 --gpuID 0,1 --cube_size 48 --crop_size 64, I got an error.
And then I tried adding the following to the start of isonet.py.
'''
physical_devices = tf.config.experimental.list_physical_devices('GPU')
assert len(physical_devices) > 0, "Not enough GPU hardware devices available"
for gpu in physical_devices:
tf.config.experimental.set_memory_growth(gpu, True)
'''
The previous error has been solved and the prediction can be made, but new problems have appeared and the predicted result is not accurate, the error and result as follow:
predict result:
I have checked the intermediate result of the training and it is normal. So why ? can anyone help solve this problem?
When I run isonet.py predict tomoset.star demo_results/model_iter30.h5 --gpuID 0,1 --cube_size 48 --crop_size 64, I got an error.
And then I tried adding the following to the start of isonet.py. ''' physical_devices = tf.config.experimental.list_physical_devices('GPU') assert len(physical_devices) > 0, "Not enough GPU hardware devices available" for gpu in physical_devices: tf.config.experimental.set_memory_growth(gpu, True) ''' The previous error has been solved and the prediction can be made, but new problems have appeared and the predicted result is not accurate, the error and result as follow:
predict result:
I have checked the intermediate result of the training and it is normal. So why ? can anyone help solve this problem?