Closed RamyaRaghuraman closed 4 years ago
RamyaRaghuraman
commented
4 years ago @adynathos ,
After the current commit aa78c5d, images under gen_image, demo_with_labels and demo_with_baselines folders are not being created. And also before the current commit, only images under sem_labels and gen_image folders were being created.
adynathos
commented
4 years ago There may be an error happening during the image being written. The images are saved in a background thread, so we would not see the exception. Now I put the exception in the log, could you please run it again to see if there is something in the log?
RamyaRaghuraman
commented
4 years ago you are right. It is to do with .webp extension, but I did not face this issue before.
Traceback (most recent call last): File "C:\Users\RAR7ABT\pj-val-ml\pjval_ml\OSR\GAN\src\datasets\dataset.py", line 44, in imwrite **IMWRITE_OPTS.get(path.suffix.lower()[1:], {}), File "C:\Users\RAR7ABT\AppData\Local\conda\conda\envs\pjval\lib\site-packages\PIL\Image.py", line 1951, in save raise ValueError('unknown file extension: {}'.format(ext)) ValueError: unknown file extension: .webp ERROR |dataset.py:47| Error in saving image \abtvdfs2.de.bosch.com\ismdfs\iad\PJ-Val_ML\data\samples_osr_yolo\GAN\exp\0500_Eval\road_anomaly_test\sem_BaySegBdd\gen_image\animals03_Zebras_in_the_road_gen_image.webp Traceback (most recent call last): File "C:\Users\RAR7ABT\AppData\Local\conda\conda\envs\pjval\lib\site-packages\PIL\Image.py", line 1949, in save format = EXTENSION[ext] KeyError: '.webp'
I tried installing libwebp via "conda install -c conda-forge libwebp" in my conda environment, but I does not seem to work. I am trying to find a workaround for it.
adynathos
commented
4 years ago That is probably an outdated PIL, an update should fix it:
pip install --upgrade pillow
RamyaRaghuraman
commented
4 years ago thanks so much! Upgrading pillow worked! Will let you know if I face any other issues.
adynathos
commented
4 years ago I am glad it worked.
RamyaRaghuraman
commented
4 years ago Hi @adynathos
Sorry to reopen this, I think the following error is because of multiprocessing:
0%| | 0/51 [00:00<?, ?it/s]Traceback (most recent call last):
File "main.py", line 50, in
I also tried using argument n_proc=1 in frame_list_apply in frame.py, but I got the following:
0%| | 0/51 [00:00<?, ?it/s]Traceback (most recent call last):
File "main.py", line 50, in
adynathos
commented
4 years ago The trick to deal with the multiprocessing problem may be to restart the kernel before running the demo image generation.
If we execute the previous steps, the dataset might be in an altered state, then multiprocessing tries to serialize it and fails. So I restart the kernel, run just the top 3 cells (create dataset, create pipeline object) and then the demo image generation.
File "/home/rar7abt/checkouts/pj-val-ml/pjval_ml/OSR/GAN/src/pipeline/evaluations.py", line 32, in call self.channel.load(dset, frame, self.field_name) AttributeError: 'str' object has no attribute 'load'
https://github.com/cvlab-epfl/detecting-the-unexpected/blob/master/src/pipeline/evaluations.py That is not what should be at that line, please try updating to the newest commit.
RamyaRaghuraman
commented
4 years ago Your suggestion worked. Everything else works fine after the latest commit!
Hi @adynathos,
While saving the demo images from exec_evaluations.ipynb, I face the following error from E1_article_evaluation.py
0%| | 0/60 [00:00<?, ?it/s]Traceback (most recent call last): File "main.py", line 52, in
eval_obj.run_demo_imgs(dset, b_show=False)
File "/home/rar7abt/checkouts/pj-val-ml/pjval_ml/OSR/GAN/src/a05_differences/E1_article_evaluation.py", line 376, in run_demo_imgs
Frame.frame_list_apply(tr_make_demo_imgs, dset, n_threads=6, batch=4)
File "/home/rar7abt/checkouts/pj-val-ml/pjval_ml/OSR/GAN/src/pipeline/frame.py", line 122, in frame_list_apply
for result in pool.imap(task, frames, chunksize=batch):
File "/home/rar7abt/.conda/envs/pjval/lib/python3.7/multiprocessing/pool.py", line 325, in
return (item for chunk in result for item in chunk)
File "/home/rar7abt/.conda/envs/pjval/lib/python3.7/multiprocessing/pool.py", line 748, in next
raise value
File "/home/rar7abt/.conda/envs/pjval/lib/python3.7/multiprocessing/pool.py", line 121, in worker
result = (True, func(*args, *kwds))
File "/home/rar7abt/.conda/envs/pjval/lib/python3.7/multiprocessing/pool.py", line 44, in mapstar
return list(map(args))
File "/home/rar7abt/checkouts/pj-val-ml/pjval_ml/OSR/GAN/src/pipeline/frame.py", line 91, in frame_list_apply_worker
fr.apply(func)
File "/home/rar7abt/checkouts/pj-val-ml/pjval_ml/OSR/GAN/src/pipeline/frame.py", line 30, in apply
result = f(frame=self, self)
File "/home/rar7abt/checkouts/pj-val-ml/pjval_ml/OSR/GAN/src/pipeline/transforms.py", line 52, in call
frame.apply(tr)
File "/home/rar7abt/checkouts/pj-val-ml/pjval_ml/OSR/GAN/src/pipeline/frame.py", line 30, in apply
result = f(frame=self, self)
TypeError: tr_get_anomaly_gt() missing 1 required positional argument: 'labels_source'
0%| | 0/60 [00:01<?, ?it/s]
Any help would be appreciated. Thanks :)