Fl sah:异步-FL模型(asynchronous federated learning)

如何使用 TFF 框架执行异步模型训练?

我回顾迭代训练过程循环,但我不知道如何知道哪些客户模型收到。

2

很有可能在 TFF 中模拟类似于“异步 FL”的东西。考虑这一点的一种方法可能是在概念上将模拟时间挂钟时间解耦。

每轮采样不同数量的客户端(而不是通常做的统一的K客户端),也许有一些分布,根据预期训练的时间来加权客户端,可以模拟异步 FL。研究人员可以首先只处理一部分选定的客户端,研究人员可以根据需要自由地分割数据 / 计算。

Python-esque 伪代码演示了两种技术,不同的客户端采样和延迟梯度应用程序:

state = fed_avg_iter_proc.initialize()
for round_num in range(NUM_ROUNDS):
  # Here we conceptualize a "round" as a block of time, rather than a synchronous
  # round. We have a function that determines which clients will "finish" within 
  # our configured block of time. This might even return only a single client.
  partints = get_next_clients(time_window=timedelta(minutes=30))
  num_partints = len(partints)
  # Here we only process the first half, and then updated the global model.
  state2, metrics = fed_avg_iter_proc.next(state, partints[:num_partints/2])
  # Now process the second half of the selected clients. 
  # Note: this is now apply the 'pseudo-gradient' that was computed on clients
  # (the difference between the original `state` and their local training result),
  # to the model that has already taken one step (`state2`). This possibly has
  # undesirable effects on the optimisation process, or may be improved with
  # techniques that handle "stale" gradients.
  state3, metrics = fed_avg_iter_proc.next(state2, partints[num_partints/2:])
  # Finally update the state for the next for-loop of the simulation.
  state = state3

本站系公益性非盈利分享网址,本文来自用户投稿,不代表边看边学立场,如若转载,请注明出处

(268)
怎么添加开机启动程序:添加应用程序组而不在Xcode上添加开发人员帐户
上一篇
Changhng电视如何投屏:如何为三星电视应用程序创建闪屏
下一篇

相关推荐

发表评论

登录 后才能评论

评论列表(17条)