Does prisma app use something similar to come up with these cool filters?
Where can I find the Jupyter theme Jeremy is using?
Is it better to calculate f_content for a higher layer for VGG and use a lower layer for f_style sine the higher layer abstracts are captured in the higher layer and the lower layer captures textures and “style”?
i had an error running it, did anyone else have this issue?
ModuleNotFoundError: No module named 'xgboost'
@mo.shakirma I had to install these to get the notebooks running
pip install xgboost
pip install gensim
pip install keras-tqdm
I shared a link on the part 1 forum. anyway here it is: https://github.com/dunovank/jupyter-themes
for me it was:
pip install matplotlib
pip install pandas
pip install xgboost
pip install bcolz
pip install gensim
pip install nltk
pip install keras_tqdm
from utils2 import *
caused all that dependencies
I found out about Part I of this course from Import AI.
If you import the VGG model that is built into keras (keras.applications), do you still have to re-order the channels, etc.?
Shouldn’t we use something like Resnet instead of VGG (with avg pooling) since the residual blocks carry more context?
Should/do we put in batch normalization like we did in lesson 5?
Will the pre-trained weights change if we’re using Average pooling instead of max pooling?
how about vgg16_bn_avg?
MaxPool --> AvgPool change still preserves the “pretrained”?
Do you have to retrain VGG on imagenet when you change max pooling to avg pooling?