一些建议some advice。

[jjandnn]

H： hello！

test_multipose.py:

ngpus = opt.device_count
render_gpu_ids = list(range(ngpus - opt.render_thread, ngpus))
render_layer_list = []
for gpu in render_gpu_ids:
    opt.gpu_ids = gpu
    render_layer = TestRender(opt)
    render_layer_list.append(render_layer)

opt.gpu_ids = list(range(0, ngpus - opt.render_thread))？？？？？？？？？？？？？？？

这句有问题，如果device_count=1,ngpus就等于1，这时候，opt.render_thread（你默认为2）假设等于1呢？这个时候opt.gpu_ids = list(range(0, ngpus - opt.render_thread)) 就等于【】，没有值的空表。 so：opt.gpu_ids = list(range(0, ngpus)才能继续下去。

我觉得作者的码基于自身8个gpu，多线程的设备环境。 你可以在read me说明清晰。 另外，眼前，一个项目的流行，进一步的发展很大程度上依赖于colab，paddle等云端平台，开发者不会天天conda的，大家习惯，先扔进云端走一遍。 作者可以自己先走一遍，这样对于后续有极大的好处。 也可以让你避免代码中出现你之前的use_BG、 use_VAE之类改动而忘记删码的不必要的麻烦， 这些问题会导致明明挺优秀的项目无法流行，这么大量的参数，没有说明，连test都无法顺利向前走（不是大家走不过去，而是要花时间），会很可惜的。

祝顺利！

There is a problem with this sentence. If device_count = 1, ngpus is equal to 1, at this time, suppose that opt.render_thread (you default is 2) is equal to 1? At this time opt.gpu_ids = list (range (0, ngpus-opt.render_thread)) is equal to [], an empty table with no value.

so: opt.gpu_ids = list (range (0, ngpus) to continue.

I think the author's code is based on its 8 gpu, multi-threaded device environment.

You can explain clearly in read me.

In addition, in front of us, the popularity and further development of a project relies heavily on cloud platforms such as colab and paddle. Developers will not conda every day. Everyone is used to throwing them into the cloud first.

The author can go through it by himself first, which has great benefits for the follow-up.

It also allows you to avoid your previous use_BG,

Use_VAE and other changes and forget the unnecessary trouble of deleting the code,

These problems will lead to the failure of popular projects that are clearly outstanding. With so many parameters and no description, even the test cannot go forward smoothly (not everyone can't go through it, but it takes time), it will be a pity.

Good luck

[Hangz-nju-cuhk]

Hi @jjandnn these are indeed great suggestions. We have tested before, but some new bugs are created accidentally by me after releasing the code.

You should see the other popular repo of mine, which is even worse:(. Maintaining projects is a bit difficult for a lazy researcher like me. As we will try making our repo better, I am curious about the "Developers will not conda every day" statement. Maybe it is time for us to step into the next era.

[jjandnn]

Hi @jjandnn these are indeed great suggestions. We have tested before, but some new bugs are created accidentally by me after releasing the code.

You should see the other popular repo of mine, which is even worse:(. Maintaining projects is a bit difficult for a lazy researcher like me. As we will try making our repo better, I am curious about the "Developers will not conda every day" statement. Maybe it is time for us to step into the next era.

感谢您的答复，我们会持续关注您的作品！加油！ 祝你顺利！ Thanks for your reply, we will continue to pay attention to your work! Come on!

Good luck to you!

[jjandnn]

@Hangz-nju-cuhk 有关‘conda’这句话的意思是大家在测试自己感兴趣的项目时，会本能的去找特别干净现成的主流环境，而不是在自己的主机上花时间反复搭建环境，其效能会比较高。这也是kaggle，colab之类的云端服务流行的原因。你去看最近的很多机器学习的流行项目，都在往更模块化的方向靠拢。这是趋势。 The meaning of 'conda' means that when testing projects that interest you, you will instinctively find a particularly clean and ready-made mainstream environment, instead of spending time on your own host to build the environment repeatedly, its performance will be relatively high . This is why cloud services such as kaggle and colab are popular. You go to see many recent popular projects of machine learning, which are moving towards a more modular direction. This is the trend.

[yuemengyuan]

@jjandnn 您好，由于latest_net_G.pth在谷歌云盘，我翻墙下载非常不稳定，经常断开，请问您可以分享一下百度云盘的预训练模型吗？谢谢

[jolt2017]

H： hello！

test_multipose.py:

ngpus = opt.device_count
render_gpu_ids = list(range(ngpus - opt.render_thread, ngpus))
render_layer_list = []
for gpu in render_gpu_ids:
    opt.gpu_ids = gpu
    render_layer = TestRender(opt)
    render_layer_list.append(render_layer)

opt.gpu_ids = list(range(0, ngpus - opt.render_thread))？？？？？？？？？？？？？？？

这句有问题，如果device_count=1,ngpus就等于1，这时候，opt.render_thread（你默认为2）假设等于1呢？这个时候opt.gpu_ids = list(range(0, ngpus - opt.render_thread)) 就等于【】，没有值的空表。 so：opt.gpu_ids = list(range(0, ngpus)才能继续下去。

我觉得作者的码基于自身8个gpu，多线程的设备环境。 你可以在read me说明清晰。 另外，眼前，一个项目的流行，进一步的发展很大程度上依赖于colab，paddle等云端平台，开发者不会天天conda的，大家习惯，先扔进云端走一遍。 作者可以自己先走一遍，这样对于后续有极大的好处。 也可以让你避免代码中出现你之前的use_BG、 use_VAE之类改动而忘记删码的不必要的麻烦， 这些问题会导致明明挺优秀的项目无法流行，这么大量的参数，没有说明，连test都无法顺利向前走（不是大家走不过去，而是要花时间），会很可惜的。

祝顺利！

There is a problem with this sentence. If device_count = 1, ngpus is equal to 1, at this time, suppose that opt.render_thread (you default is 2) is equal to 1? At this time opt.gpu_ids = list (range (0, ngpus-opt.render_thread)) is equal to [], an empty table with no value.

so: opt.gpu_ids = list (range (0, ngpus) to continue.

I think the author's code is based on its 8 gpu, multi-threaded device environment.

You can explain clearly in read me.

In addition, in front of us, the popularity and further development of a project relies heavily on cloud platforms such as colab and paddle. Developers will not conda every day. Everyone is used to throwing them into the cloud first.

The author can go through it by himself first, which has great benefits for the follow-up.

It also allows you to avoid your previous use_BG,

Use_VAE and other changes and forget the unnecessary trouble of deleting the code,

These problems will lead to the failure of popular projects that are clearly outstanding. With so many parameters and no description, even the test cannot go forward smoothly (not everyone can't go through it, but it takes time), it will be a pity.

Good luck

你好，请问你在自己数据集上 训练成功 了么？我这里遇到 判别器损失函数为 0的情况，不知道怎么解决