Nicotine

output with shape [1, 28, 28] doesn't match the broadcast shape [3, 28, 28]

30

When I run GAN code, I got a runtime error output with shape [1, 28, 28] doesn't match the broadcast shape [3, 28, 28] how fix it

首先谢谢提供代码。 按照代码思路运行之后，在第2000轮基本就收敛了，但是生成的图片好像不太好，网络加深好像也不太能行，这是怎么回事呢？

你好，看到你用export_savemodel的方式导出模型，想问下predict文件该怎么写呢，是否还要包括feature column那些预处理？

I think the key point of wide network is the crossing of category features but it's not reflected in this code. https://github.com/keras-team/keras-io/blob/master/examples/structured_data/wide_deep_cross_networks.py Maybe the CategoryCrossing method can achieve it.