I have some issue with runing app.py

[EURAISA]

I don't know where should I put the data on, create a new directory?

what's this mean of green code, what's the format of data, should I creat 3 directory of those green name?
could I use chinese to share my problem

[adamz799]

I don't know where should I put the data on, create a new directory?

what's this mean of green code, what's the format of data, should I creat 3 directory of those green name?
could I use chinese to share my problem

Yes, you need to create 3 folders as the code shows if you want to train the neuralnet, 'bw/' and 'color/' are the directories for input data(black&white and corresponding color images). But i don`t remeber what 'block/' means, because I haven`t worked on Deep Learning for over a year. Anyway, all these directories are only used in traning process. To run the demo, you need to prepare a black&white lineart and a reference image to fetch color and follow the instruction:

[EURAISA]

感谢您的解答，我想block应该是指作为参考的色块，app.py是demo的话utils_app.py就是用来训练的吗？

[EURAISA]

如果可以的话我希望你能告诉哦我如何训练这个网络，我看了一遍代码只看见了使用参考图片和线稿路径的代码

[EURAISA]

I have try to run the demo not train the neural but it still warning me of this , why and I have creat 3 folders

[adamz799]

如果可以的话我希望你能告诉哦我如何训练这个网络，我看了一遍代码只看见了使用参考图片和线稿路径的代码

不好意思，最近工作比较忙，没来得及弄这个，训练的python文件我也不确定是哪个，因为当时用了很多种训练方法，可能需要你自己看一下。我记得当时比较有效的办法是把色域从RGB转换到HSV，然后用GAN分通道训练。文件路径的话，就是按照readme建立路径，如果找不到路径，有可能是启动位置不对，可以试试直接在cmd启动。代码我上传到train文件夹了，可以参考一下，有问题欢迎继续留言。

[EURAISA]

感谢上传，我是一名研究生，刚进入这个领域有很多还不懂，谢谢你的回复