H-CODE6
commented
2 years ago The data set format is arranged according to your style. This error always appears. It's very confusing
Closed H-CODE6 closed 2 years ago
H-CODE6
commented
2 years ago The data set format is arranged according to your style. This error always appears. It's very confusing
himashi92
commented
2 years ago Is this path is correct ? /media/hkw/hkwdisk/viv-test/vizviva_brats_2021-main (1)/vizviva_brats_2021-main/BRATS2021/Data/valid/BraTS2021_00046_t1.nii.gz
H-CODE6
commented
2 years ago My path in config is BRATS_TEST_FOLDER ='/media/hkw/hkwdisk/viv-test/vizviva_brats_2021-main (1)/vizviva_brats_2021-main/BRATS2021/Data/valid' . Valid contains BraTS2021_00046 folder,I don't know why BraTS2021_00046 is missing
H-CODE6
commented
2 years ago I don't know why I don't recognize the BraTS2021_00046 folder and ask to display the files contained in it
H-CODE6
commented
2 years ago There was no problem with the training. I kept reporting this error during the test. I'm very distressed. I don't know why. I look forward to your help
himashi92
commented
2 years ago Try using get_dataset_val method in brats_train.py script. https://github.com/himashi92/vizviva_brats_2021/blob/main/dataset/brats_train.py#L113
himashi92
commented
2 years ago Change this line of code in your test.py https://github.com/himashi92/vizviva_brats_2021/blob/main/test.py#L18
from dataset.brats_train import get_datasets_val
H-CODE6
commented
2 years ago Hello, can you tell me how to modify which code and how to modify it?
H-CODE6
commented
2 years ago I added from dataset.brats_train import get_datasets_val and still reported the same error.
H-CODE6
commented
2 years ago Working with 1 GPUs
invalid literal for int() with base 10: ''
using default group number
[24, 48, 96, 192]
total number of trainable parameters 4793895
Validation Input Directory /input
Bench Test dataset number of batch: 1
=> loading checkpoint runs/brats_2021/model_1/model_best.pth.tar
=> loaded checkpoint 'runs/brats_2021/model_1/model_best.pth.tar' (epoch 3)
Traceback (most recent call last):
File "test.py", line 117, in
H-CODE6
commented
2 years ago Do not know whether to def generate_ Segmentation (data_loader, model, args) code needs to be modified?
H-CODE6
commented
2 years ago I look forward to your help
himashi92
commented
2 years ago Still, your script is referring to brats.py, not brats_train.py
H-CODE6
commented
2 years ago Can you tell me how to modify which line of code? Thank you very much
H-CODE6
commented
2 years ago Working with 1 GPUs
invalid literal for int() with base 10: ''
using default group number
[24, 48, 96, 192]
total number of trainable parameters 4793895
Traceback (most recent call last):
File "test.py", line 117, in
H-CODE6
commented
2 years ago I used from dataset.brats_train import get_datasets_val and the above problems occurred
H-CODE6
commented
2 years ago 1 was used bench_dataset = get_datasets_val(args.seed) and the following problems occurred.
Working with 1 GPUs
invalid literal for int() with base 10: ''
using default group number
[24, 48, 96, 192]
total number of trainable parameters 4793895
Train Input Directory /input
/media/hkw/hkwdisk/viv-test/vizviva_brats_2021-main (1)/vizviva_brats_2021-main/BRATS2021/Data/RSNA_ASNR_MICCAI_BraTS2021_TrainingData
first idx of train 0
first idx of test 7
Bench Test dataset number of batch: 3
=> loading checkpoint runs/brats_2021/model_1/model_best.pth.tar
=> loaded checkpoint 'runs/brats_2021/model_1/model_best.pth.tar' (epoch 3)
torch.Size([4, 137, 164, 137])
Traceback (most recent call last):
File "test.py", line 117, in
H-CODE6
commented
2 years ago I look forward to your help in solving this problem. Thank you very much and look forward to your reply
H-CODE6
commented
2 years ago I still feel that def generate_segmentations(data_loader, model, args): of test.py needs to be modified
H-CODE6
commented
2 years ago Please check test Whether the code in py needs to be updated, thank you very much
himashi92
commented
2 years ago for patient_dir in patients_dir: patient_id = patient_dir.name paths = [patient_dir / f"{patient_id}{value}.nii.gz" for value in self.patterns] patient = dict( id=patient_id, t1=paths[0], t1ce=paths[1], t2=paths[2], flair=paths[3], seg=paths[4] if not no_seg else None ) self.datas.append(patient)
Add this for loop in brats.py https://github.com/himashi92/vizviva_brats_2021/blob/main/dataset/brats.py#L28 and use get_dataset_val method in there.
This test.py is originally created for brats evaluation and according to their requirements. So there's nothing wrong with the code. It should be adjusted according to your requirements and with debugging and properly going through the code you should be able to modify the code.
H-CODE6
commented
2 years ago OK, let me add it. I don't know where to add it
himashi92
commented
2 years ago Its a for loop, so you need to add indent.
himashi92
commented
2 years ago Refer this, it's the same for loop. https://github.com/himashi92/vizviva_brats_2021/blob/main/dataset/brats_train.py#L26
H-CODE6
commented
2 years ago Traceback (most recent call last):
File "test.py", line 18, in
himashi92
commented
2 years ago Now add https://github.com/himashi92/vizviva_brats_2021/blob/main/dataset/brats_train.py#L113 this method as get_datasets_val.
This will be a combination of both brats.py and brats_train.py scripts. My advice for you is, please debug and see what's going on in the code. If you understand the code, then it'll be easy for you to modify the code. (Not all authors guide you :) )
H-CODE6
commented
2 years ago ok, thank you very much
H-CODE6
commented
2 years ago Hello, author. I'm sorry to bother you again. After testing, your code has no problem, but each path can only be displayed as/media/hkw/hkwdisk/viv-test/vizviva_brats_2021-main (1)/vizviva_brats_2021-main/BRATS2021/Data/valid/BraTS2021_00046_t1.nii.gz , and several paths cannot be recognized .nii file folders, such as Brats2021_00046, so I can only test one folder at a time, and can't test all the data at the same time. Thank you very much for helping us see where we need to optimize when you're free
H-CODE6
commented
2 years ago There is no problem with your code, but I don't understand why each brats data is not recognized in the path, but directly to the data in the folder NII file, very much looking forward to solving my doubts
himashi92
commented
2 years ago Can you send me your test.py and brats.py to my email (himashi.ama92@gmail.com)
H-CODE6
commented
2 years ago No problem. Thank you very much
H-CODE6
commented
2 years ago Hello, I have tried this method before, and it always appears
Working with 1 GPUs
invalid literal for int() with base 10: ''
using default group number
[24, 48, 96, 192]
total number of trainable parameters 4793895
Validation Input Directory /input
Traceback (most recent call last):
File "test.py", line 117, in
Hello, author, I run according to your test.py. The following problems occur. What is the reason? We look forward to you.
Working with 1 GPUs invalid literal for int() with base 10: '' using default group number [24, 48, 96, 192] total number of trainable parameters 4793895 Validation Input Directory /input Bench Test dataset number of batch: 1 => loading checkpoint runs/brats_2021/model_1/model_best.pth.tar => loaded checkpoint 'runs/brats_2021/model_1/model_best.pth.tar' (epoch 3) Traceback (most recent call last): File "test.py", line 117, in
main(arguments)
File "test.py", line 111, in main
generate_segmentations(bench_loader, model_1, args)
File "test.py", line 40, in generate_segmentations
for i, batch in enumerate(data_loader):
File "/home/hkw/anaconda3/envs/pytorch/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 530, in next
data = self._next_data()
File "/home/hkw/anaconda3/envs/pytorch/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 1224, in _next_data
return self._process_data(data)
File "/home/hkw/anaconda3/envs/pytorch/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 1250, in _process_data
data.reraise()
File "/home/hkw/anaconda3/envs/pytorch/lib/python3.7/site-packages/torch/_utils.py", line 457, in reraise
raise exception
RuntimeError: Caught RuntimeError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/home/hkw/anaconda3/envs/pytorch/lib/python3.7/site-packages/torch/utils/data/_utils/worker.py", line 287, in _worker_loop
data = fetcher.fetch(index)
File "/home/hkw/anaconda3/envs/pytorch/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 49, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/hkw/anaconda3/envs/pytorch/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 49, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/media/hkw/hkwdisk/viv-test/vizviva_brats_2021-main (1)/vizviva_brats_2021-main/dataset/brats.py", line 39, in getitem
patient_image = {key: self.load_nii(_patient[key]) for key in _patient if key not in ["id", "seg"]}
File "/media/hkw/hkwdisk/viv-test/vizviva_brats_2021-main (1)/vizviva_brats_2021-main/dataset/brats.py", line 39, in
patient_image = {key: self.load_nii(_patient[key]) for key in _patient if key not in ["id", "seg"]}
File "/media/hkw/hkwdisk/viv-test/vizviva_brats_2021-main (1)/vizviva_brats_2021-main/dataset/brats.py", line 67, in load_nii
return sitk.GetArrayFromImage(sitk.ReadImage(str(path_folder)))
File "/home/hkw/anaconda3/envs/pytorch/lib/python3.7/site-packages/SimpleITK/SimpleITK.py", line 8876, in ReadImage
return _SimpleITK.ReadImage(*args)
RuntimeError: Exception thrown in SimpleITK ReadImage: /tmp/SimpleITK/Code/IO/src/sitkImageReaderBase.cxx:99:
sitk::ERROR: The file "/media/hkw/hkwdisk/viv-test/vizviva_brats_2021-main (1)/vizviva_brats_2021-main/BRATS2021/Data/valid/BraTS2021_00046_t1.nii.gz" does not exist.