Error of "model.predict_generator"

[ycjing]

Hi, thanks for the great code. It is great for a beginner of Keras, like me. But I found some errors when I run the code. In gap.py, the terminal reports that "AttributeError: 'DirectoryIterator' object has no attribute 'nb_sample'". I wonder if it is due to the different versions of keras and is there any solutions to this? Thanks ahead of time!

[ypwhs]

我用的版本是 Keras 1.2.2，你可以用这个命令安装这个版本的Keras:

pip install keras==1.2.2

[ycjing]

@ypwhs Thanks!

[dysdsyd]

@ycjing, Replace nb_sample with samples.

[elitezhe]

@syeddanish41 Thanks for your reply. But I find that the train_generator.samples return value of 25000, that is number of total pictures.
However, we generate feature by batch, so predict_generator will run batch_size times more, this seems make the progress never end.

I tried replace nb_sample with samples//batch_size, it works.(but cannot generate all the 25000, since the batch size i chose is 16)

[leeguandong]

你好，我运行了下，出现了两个错误 1、OSError: [Errno 22] Invalid argument: 'train2\cat\cat.0.jpg' 2、TypeError: Error when checking : data should be a Numpy array, or list/dict of Numpy arrays. Found: None...
没法导出后h5文件？？