thedatageek
(Mayukh Sarkar)
February 25, 2020, 10:10am
1
Hello,
I have follwoed the instruction properly and I have installed pillow-simd
as mentioned here . I am getting the following error on importing from fastai.vision import *
This is how the error looks like
Suprisingly it is using Pillow in anaconda3/lib/python3.7
not from the local environment which is python3.6 not python 3.7.
Please help.
Thanks in advance
philchu
(Phillip Chu)
February 25, 2020, 10:19am
2
It is a known problem with the latest version of Pillow (7.0?) Is that what you’re using?
Hi everyone! Each semester I lecture at the University of West Florida through one of our clubs on fastai. Normally I keep it offline but this year I have decided to live stream it along with have a dedicated megathread to discussing the course, asking questions, and networking together! This semester’s version will be using the version 2 of the fastai library (thank you to @Jeremy and @sgugger for all the amazing work you have put in!). This course is designed to be intro friendly (it’s geared …
opened 09:04AM - 11 Oct 19 UTC
closed 10:25AM - 19 Nov 19 UTC
Installation
Linux
Anaconda
I got this error while importing Image
```python
from PIL import Image
Trac… eback (most recent call last):
File "<ipython-input-13-b7f01c2f8cfe>", line 1, in <module>
from PIL import Image
File "/lib/python3.7/site-packages/PIL/Image.py", line 40, in <module>
from . import PILLOW_VERSION, ImageMode, TiffTags, __version__, _plugins
**ImportError:** cannot import name 'PILLOW_VERSION' from 'PIL' (unknown location)
```
I tried to remove it by myself but then got the same problem with `__version__`, then `_plugins`
Updating doesn't work for me. Can you help please?
thedatageek
(Mayukh Sarkar)
February 25, 2020, 10:30am
3
@jeremy @stas
Actually to be honest, I am not using Pillow
. I am using pillow simd
as mentened in fastai website. But I am getting this error. Here is a detailed output of conda list
# Name Version Build Channel
_libgcc_mutex 0.1 main
asn1crypto 1.3.0 py36_0
blas 1.0 mkl
ca-certificates 2020.1.1 0
certifi 2019.11.28 py36_0
cffi 1.14.0 py36h2e261b9_0
chardet 3.0.4 py36_1003
conda 4.8.2 py36_0
conda-package-handling 1.6.0 py36h7b6447c_0
cryptography 2.8 py36h1ba5d50_0
idna 2.8 py36_0
intel-openmp 2020.0 166
jpeg 9b h024ee3a_2
ld_impl_linux-64 2.33.1 h53a641e_7
libedit 3.1.20181209 hc058e9b_0
libffi 3.2.1 hd88cf55_4
libgcc-ng 9.1.0 hdf63c60_0
libgfortran-ng 7.3.0 hdf63c60_0
libjpeg-turbo 2.0.3 h516909a_1 conda-forge
libstdcxx-ng 9.1.0 hdf63c60_0
libtiff 4.1.0 h2733197_0
mkl 2020.0 166
mkl-service 2.3.0 py36he904b0f_0
mkl_fft 1.0.15 py36ha843d7b_0
mkl_random 1.1.0 py36hd6b4f25_0
ncurses 6.2 he6710b0_0
numpy 1.18.1 py36h4f9e942_0
numpy-base 1.18.1 py36hde5b4d6_1
openssl 1.1.1d h7b6447c_4
pillow-simd 7.0.0.post3 pypi_0 pypi
pip 20.0.2 py36_1
pycosat 0.6.3 py36h7b6447c_0
pycparser 2.19 py36_0
pyopenssl 19.1.0 py36_0
pysocks 1.7.1 py36_0
python 3.6.10 h0371630_0
readline 7.0 h7b6447c_5
requests 2.22.0 py36_1
ruamel_yaml 0.15.87 py36h7b6447c_0
scipy 1.4.1 py36h0b6359f_0
setuptools 45.2.0 py36_0
six 1.14.0 py36_0
sqlite 3.31.1 h7b6447c_0
tk 8.6.8 hbc83047_0
tqdm 4.42.1 py_0
urllib3 1.25.8 py36_0
wheel 0.34.2 py36_0
xz 5.2.4 h14c3975_4
yaml 0.1.7 had09818_2
zlib 1.2.11 h7b6447c_3
zstd 1.3.7 h0b5b093_0
You can see I am using pillow-simd 7.0.0.post3.
I am using exactly what is mentioned in the website. Do I need to do something different? As I mentioned already, it is not using pillow-simd from the python36 environment but using PIL from other python37 environment.
Try using something like pip install -U pillow_simd==6.2.2.post1
. That fixed it for me.
stas
(Stas Bekman)
February 25, 2020, 8:20pm
5
pillow-simd
is the same as pillow
with a few optimizations. pillow-7 now uses PIL.Image.__version__
instead of PIL.Image.PILLOW_VERSION
. (docs should be updated shortly)
If you get your repos crossed, perhaps start with a fresh conda repository? It’s possible that over time you got some mixups. I trust that if you start afresh everything will work.
But your error has nothing to do with fastai, as you can see the problem is in torchvision, you need to update it, as you can see it has been fixed a few months ago:
from __future__ import division
import torch
import sys
import math
from PIL import Image, ImageOps, ImageEnhance, __version__ as PILLOW_VERSION
try:
import accimage
except ImportError:
accimage = None
import numpy as np
from numpy import sin, cos, tan
import numbers
import collections
import warnings
committed 08:09PM - 18 Oct 19 UTC
* doc-build fixed
* deprecation fixes
* deprecation updates
You need to install torchvision 0.5.0 or higher which you can do by just updating conda or pip
e.g.: pip install torchvision
liuyao
(Yao Liu)
February 25, 2020, 8:38pm
6
a “hack” I learned from here: include these two lines before you import torchvision
or fastai
import PIL
PIL.PILLOW_VERSION = PIL.__version__
3 Likes
paidion
(Jeremiah YS Chung)
April 10, 2020, 12:25pm
7
I edit and change PILLOW_VERSION to version in L5 of funtional.py. It seems working too.