Yi Wang

Results 78 comments of Yi Wang

@sneaxiy This issue mixed up two topics -- (1) removing global variables and (2) redesign the API. Let us keep (2) at https://github.com/sql-machine-learning/playground/issues/42 and make this issue focusing on (1)....

Thanks to @brightcoder01 for raising this question. I reorganized the comparison in the above comment into a table. | | Dask | Mars | |---|------|------| | Kubernetes | Yes |...

我建议 @wangyang59 是不是按照 @luotao1 的第一个建议,规范一下目录结构,我们就先merge这个PR。内容的修改我估计是一个精益求精的过程,可以随后再发PR修改?

另外,conditional GAN 是不是应该分出来单做一章?我理解一个notebook里只能放一个示范程序,或者放一个示范程序比较易于读者参考的修改吧?

这个change是为了能浮动出二级菜单(对应每个chapter里的sections)吗?如果是这样,在手机上也能有同样的效果吗?

感谢贡献!可能需要请 @luotao1 看看这个PR?

> We can check in the pre-compiled binary files of PS in the repo. In the building `.whl` process, we can directly use the binary files. No. The basic rule...

I am trying to understand https://github.com/wangkuiyi/elasticdl/issues/1006#issue-478372383. This follows my understanding. Currently, the function `elasticdl.train` requires the following parameters: - `model_def`: the model definition as a `tf.keras.Model`-derived class or a function...

ElasticDL can handle very large models using its general parameter server in Go, which is based on the previous design we explained in Google Developer Day 2019, but with many...

![image](https://user-images.githubusercontent.com/1548775/85955683-27c74f00-b935-11ea-9630-ac8b80b6a16c.png) This one looks a connected graph of processes in a distributed job.