search

taigw / brats17

Brain tumor segmentation for MICCAI 2017 BraTS challenge
BSD 3-Clause "New" or "Revised" License
324 stars 130 forks source link

Train:MemoryError occuried #9

Closed cb632482135qq closed 6 years ago

cb632482135qq commented 6 years ago

My computer is:

windows 10 GTX1060 6G 16G DDR4 2400Mhz

When I run the" python train.py train_wt_ax.txt" , i meet this problem:

Traceback (most recent call last): File "D:\TF\brats17-master\train.py", line 123, in train(config_file) File "D:\TF\brats17-master\train.py", line 82, in train dataloader.load_data() File "D:\TF\brats17-master\util\data_loader.py", line 112, in load_data volume = itensity_normalize_one_volume(volume) File "D:\TF\brats17-master\util\data_process.py", line 95, in itensity_normalize_one_volume out = (volume - mean)/std MemoryError

At the same time, I found that my memory utilization is close to 100% from the Task Manager.

Please tell me whether i can get some solutions to this problems. Any help will be appreciated sincerely!

taigw commented 6 years ago

@cb632482135qq Unfortunately, the current implementation requires loading all the training images into the memory at the beginning. You may consider to use a machine with a larger memory or use the IO in NiftyNet to avoid the memory error.