你即將離開本站

並前往https://mobile.twitter.com/jeremyphoward/status/901295815766941696?lang=cs

Uživatel Jeremy Howard na Twitteru: „@soumithchintala ...

Thanks! Sorry I didn't understand that. So 3 problems with sparse: it doesn't work with weight_decay, or with Adam, and it's slower here!

確定！ 回上一頁

查詢 「Weight_decay pytorch」的人也找了：

  1. Weight_decay
    2.
  2. Weight decay
    3.
  3. Optimizerstep
    4.
  4. Optimizer pytorch
    5.
  5. weight decay作用
    6.
  6. Cosine annealing scheduler
    7.
  7. Adam Optimizer
    8.
  8. Torch optim.Adam 用法
    9.