kaldi-asr/kaldi

nnet3 : Reduce unnecessary outputs (file or email) during training

Open

#1,349 建立於 2017年1月17日

在 GitHub 查看
 (2 留言) (0 反應) (0 負責人)Shell (15,392 star) (5,359 fork)batch import
enhancementhelp wantednnet3stale

描述

(Created to ensure we keep track of a low priority issue.)

Status update emails are sent after every iteration if an email id is specified during training. However status information uses output of train/validation accuracy/log-probability computation jobs which are run in the background. In many cases these might be very slow to compute, depending on the model, as these are run on the CPU. This would mean that there are uninformative status mails which do not provide any info other than training time.

e.g.

 %Iter   duration        train_loss      valid_loss      difference
Total training time is 0:59:53

貢獻者指南

技術棧
cppshell
領域
machine learning
議題類型
feature
難度面向新貢獻者的預計實作難度，1 表示很小改動，5 表示專家級工作。
3
預計時間有經驗貢獻者完成調查、實作、測試並準備 pull request 的粗略時間範圍。
half day
活動狀態議題目前的可參與程度：新鮮、活躍、陳舊、阻塞或等待維護者輸入。
stale
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
mostly clear
前置要求
Understanding of Kaldi training scriptsBasic shell scripting
新手友善度1-100 的估計分數，表示該議題對首次貢獻者的友善程度。
40
研究方向
Investigate the training scripts in 'steps/nnet3/' to understand how status emails are generated. Look for the logic that decides whether to send email and how the validation accuracy is computed. Consider adding a condition to skip sending if validation results are not ready or are uninformative. Check existing comments or related issues for previous discussion.