Zhuang Weiming

@954514762 可以本地运行一个后端 https://github.com/weimingwill/fashion-mall-server

`app.wpy` 里的 `baseUrl` 是 null，没有模拟数据，起不来。

@GugaLiz 除非他们开放，不然没办法解决。 可以参考我的另一个比较简单的商城 https://github.com/weimingwill/fashion-mall 搭配了本地可以运行的服务器 https://github.com/weimingwill/fashion-mall-server

@yanchuan1990 可以用 https://github.com/weimingwill/fashion-mall-server 启动一个本地服务器，改一下地址就能收到数据哦~

@coolhwm 你好，请问是关闭了 easy_mock 的模拟数据对接吗？build 之后，运行起来 auth 有问题。不知道能不能给个 mock 的 baseUrl 能够参考学习？谢谢

最后怎么实现定时爬去的？ @liujunyi

@marcomilanesio just remove `group=0` in `dist.all_reduce(param.grad.data, op=dist.ReduceOp.SUM, group=0)`

Hi @SunTongtongtong, maybe you can try running it with only one dataset first to check whether the performance match with "Local Training" in the paper.

Hi @J-BING , these feedbacks are really useful. We will incorporate and revise related issues. It'll be even better if you have time to create a pull request for the...

Yes. It supports the transformer model. You just need to register the transformer model following this tutorial https://easyfl.readthedocs.io/en/latest/tutorials/model.html