suvoooo/learn-tensorflow

Learning Tensorflow Step by Step:: Concepts, Examples & Applications

learn-tensorflow's Issues

Images Not Loading

First of all, thank you for your amazing explanation along with the code. But the images you provided are not loading. So it's difficult to follow your explanation. Please update the notebook so that it will load the images.

`fit()` missing in your VGG-like model in Facial_Keypoint_Kaggle.ipynb script

Hi,
I am a Tensorflow noob and was wondering what is missing after the following compilation in the script:

# face_key_model.compile(loss='mse', 
#                        optimizer=sgd, 
#                        metrics=['acc'])

I was trying to fit with the following parameters:

face_key_model.fit(train_ims_clean, 
                    clean_keypoints_arr, 
                    validation_split= 0.05, 
                    batch_size=64, 
                    epochs=300, 
                    callbacks=[customCallbacks(), reduce_lr, lrdecay, earlystop],
                    verbose=0)

But I get an error message when I want to include earlystop:

  File "C:\Users\xxx\anaconda3\envs\Keras\lib\site-packages\keras\callbacks.py", line 285, in set_model
    callback.set_model(model)

AttributeError: 'function' object has no attribute 'set_model'

Once I omit earlystop, training does work, but the network fails miserably after a few epochs of training as presumably the weights shoot up to infinity (or minus infinity) and so does the loss..
Any help?

Question about resnet implementation

from 2nd-5th stage why do we use the res_identity more than once?
`#2nd stage

frm here on only conv block and identity block, no pooling

x = res_conv(x, s=1, filters=(64, 256))
x = res_identity(x, filters=(64, 256))
x = res_identity(x, filters=(64, 256))

3rd stage

x = res_conv(x, s=2, filters=(128, 512))
x = res_identity(x, filters=(128, 512))
x = res_identity(x, filters=(128, 512))
x = res_identity(x, filters=(128, 512))

4th stage

x = res_conv(x, s=2, filters=(256, 1024))
x = res_identity(x, filters=(256, 1024))
x = res_identity(x, filters=(256, 1024))
x = res_identity(x, filters=(256, 1024))
x = res_identity(x, filters=(256, 1024))
x = res_identity(x, filters=(256, 1024))

5th stage

x = res_conv(x, s=2, filters=(512, 2048))
x = res_identity(x, filters=(512, 2048))
x = res_identity(x, filters=(512, 2048))
`