Lesson 1 (Week 1) OSError: Is a directory


(Patrick Suzuki) #1

I am getting the following error after trying to run lesson1.ipynb

Reading the following post, it seems like an issue due to '.ipynb_checkpoints’: How to remove .ipynb checkpoint

However, I cannot find the above specific notebook from the Jupyter GUI. I’m not 100% if this is causing the issue so please point me to the right direction. If the above is in fact the cause then is there a way to delete from GUI or do I need to delete from terminal? If from the terminal, what is the best way to access Jupyter Notebook that is running on Gradient? Thank you all!

OSError: Is a directory: data/magicornot/train/gourmet/.ipynb_checkpoints


Part 1: OS Error
(Martin) #2

Is it not an easy option for you to just simply delete it all (or rename it) and download it again from GitHub?
I presume this is an option for you because you have a problem with “lesson1.ipynb”… correct me if this isn’t an option.


(Patrick Suzuki) #3

That is what I thought, so I created a clean Jupyter Notebook environment with Gradient and re-uploaded my images but I still encounter the same error. It seems like it doesn’t matter if I start from a clean slate or new data set since “.ipynb_checkpoints” get created dynamically. And yes, you are correct that this is an issue with “lesson1.ipynb”.

Including code that is raising error:
arch=resnet34 data = ImageClassifierData.from_paths(PATH, tfms=tfms_from_model(arch, sz)) learn = ConvLearner.pretrained(arch, data, precompute=True) learn.fit(0.01, 2)

Including the error snippet:

OSError                                   Traceback (most recent call last)
<ipython-input-16-e6c87b20ce86> in <module>()
      1 arch=resnet34
      2 data = ImageClassifierData.from_paths(PATH, tfms=tfms_from_model(arch, sz))
----> 3 learn = ConvLearner.pretrained(arch, data, precompute=True)
      4 learn.fit(0.01, 2)

/notebooks/courses/dl1/fastai/conv_learner.py in pretrained(cls, f, data, ps, xtra_fc, xtra_cut, custom_head, precompute, pretrained, **kwargs)
    112         models = ConvnetBuilder(f, data.c, data.is_multi, data.is_reg,
    113             ps=ps, xtra_fc=xtra_fc, xtra_cut=xtra_cut, custom_head=custom_head, pretrained=pretrained)
--> 114         return cls(data, models, precompute, **kwargs)
    115 
    116     @classmethod

/notebooks/courses/dl1/fastai/conv_learner.py in __init__(self, data, models, precompute, **kwargs)
     98         if hasattr(data, 'is_multi') and not data.is_reg and self.metrics is None:
     99             self.metrics = [accuracy_thresh(0.5)] if self.data.is_multi else [accuracy]
--> 100         if precompute: self.save_fc1()
    101         self.freeze()
    102         self.precompute = precompute

/notebooks/courses/dl1/fastai/conv_learner.py in save_fc1(self)
    166         m=self.models.top_model
    167         if len(self.activations[0])!=len(self.data.trn_ds):
--> 168             predict_to_bcolz(m, self.data.fix_dl, act)
    169         if len(self.activations[1])!=len(self.data.val_ds):
    170             predict_to_bcolz(m, self.data.val_dl, val_act)

/notebooks/courses/dl1/fastai/model.py in predict_to_bcolz(m, gen, arr, workers)
     15     lock=threading.Lock()
     16     m.eval()
---> 17     for x,*_ in tqdm(gen):
     18         y = to_np(m(VV(x)).data)
     19         with lock:

/opt/conda/envs/fastai/lib/python3.6/site-packages/tqdm/_tqdm.py in __iter__(self)
    928  fp_write=getattr(self.fp, 'write', sys.stderr.write))
    929 
--> 930             for obj in iterable:
    931                 yield obj
    932                 # Update and possibly print the progressbar.

/notebooks/courses/dl1/fastai/dataloader.py in __iter__(self)
     86                 # avoid py3.6 issue where queue is infinite and can result in memory exhaustion
     87                 for c in chunk_iter(iter(self.batch_sampler), self.num_workers*10):
---> 88                     for batch in e.map(self.get_batch, c):
     89                         yield get_tensor(batch, self.pin_memory, self.half)
     90 

/opt/conda/envs/fastai/lib/python3.6/concurrent/futures/_base.py in result_iterator()
    584                     # Careful not to keep a reference to the popped future
    585                     if timeout is None:
--> 586                         yield fs.pop().result()
    587                     else:
    588                         yield fs.pop().result(end_time - time.time())

/opt/conda/envs/fastai/lib/python3.6/concurrent/futures/_base.py in result(self, timeout)
    430                 raise CancelledError()
    431             elif self._state == FINISHED:
--> 432                 return self.__get_result()
    433             else:
    434                 raise TimeoutError()

/opt/conda/envs/fastai/lib/python3.6/concurrent/futures/_base.py in __get_result(self)
    382     def __get_result(self):
    383         if self._exception:
--> 384             raise self._exception
    385         else:
    386             return self._result

/opt/conda/envs/fastai/lib/python3.6/concurrent/futures/thread.py in run(self)
     54 
     55         try:
---> 56             result = self.fn(*self.args, **self.kwargs)
     57         except BaseException as exc:
     58             self.future.set_exception(exc)

/notebooks/courses/dl1/fastai/dataloader.py in get_batch(self, indices)
     73 
     74     def get_batch(self, indices):
---> 75         res = self.np_collate([self.dataset[i] for i in indices])
     76         if self.transpose:   res[0] = res[0].T
     77         if self.transpose_y: res[1] = res[1].T

/notebooks/courses/dl1/fastai/dataloader.py in <listcomp>(.0)
     73 
     74     def get_batch(self, indices):
---> 75         res = self.np_collate([self.dataset[i] for i in indices])
     76         if self.transpose:   res[0] = res[0].T
     77         if self.transpose_y: res[1] = res[1].T

/notebooks/courses/dl1/fastai/dataset.py in __getitem__(self, idx)
    166             xs,ys = zip(*[self.get1item(i) for i in range(*idx.indices(self.n))])
    167             return np.stack(xs),ys
--> 168         return self.get1item(idx)
    169 
    170     def __len__(self): return self.n

/notebooks/courses/dl1/fastai/dataset.py in get1item(self, idx)
    159 
    160     def get1item(self, idx):
--> 161         x,y = self.get_x(idx),self.get_y(idx)
    162         return self.get(self.transform, x, y)
    163 

/notebooks/courses/dl1/fastai/dataset.py in get_x(self, i)
    238         super().__init__(transform)
    239     def get_sz(self): return self.transform.sz
--> 240     def get_x(self, i): return open_image(os.path.join(self.path, self.fnames[i]))
    241     def get_n(self): return len(self.fnames)
    242 

/notebooks/courses/dl1/fastai/dataset.py in open_image(fn)
    221         raise OSError('No such file or directory: {}'.format(fn))
    222     elif os.path.isdir(fn):
--> 223         raise OSError('Is a directory: {}'.format(fn))
    224     else:
    225         #res = np.array(Image.open(fn), dtype=np.float32)/255

OSError: Is a directory: /storage/magicornot/train/gourmet/.ipynb_checkpoints```

(Martin) #4

Oh! The problem is with your data.

To test this I would run:

batch = next(iter(data.trn_dl))

This is essentially what happens to get a batch when we learn.fit.

This should cause the same error but with a shorter error and should be easier to test solutions :slight_smile:

data = ImageClassifierData.from_paths(PATH, tfms=tfms_from_model(arch, sz))

I don’t think the error is because of tfms so it is most likely what is in PATH.

I then would look at the proper use of ImageClassifierData.from_paths by running:
??ImageClassifierData.from_paths

That should tell you what arguments to use for ImageClassifierData.from_paths.


(Patrick Suzuki) #5

Ran the following script to remove the directories that were being auto-generated.