[Most-ai-contest] New dispatcher module on latest training data

[kysu]

Test set:

         Top1 acc: 0.9842

         Top2 acc: 1.0   

         Top2 coverage: 0.9421

         Type acc: 0.9368

>Top2 acc: one of the top2 predicted modes in golden answer modes

>Top2 coverage: the ratio that the top2 predicted modes cover the golden modes

 

What is the difference between “Top2 acc” and “Top2 coverage”? They should be the same in usual setting.

 

KY

 

From: most-ai-contest-bounces at iis.sinica.edu.tw [mailto:most-ai-contest-bounces at iis.sinica.edu.tw] On Behalf Of 蔡仕竑
Sent: Saturday, April 4, 2020 9:04 PM
To: 范正忠 <jjfan at iis.sinica.edu.tw>
Cc: Most-ai Contest <Most-ai-contest at iis.sinica.edu.tw>
Subject: [Most-ai-contest] New dispatcher module on latest training data

 

Dear all,

 

The following link directs to the latest dispatcher module with its performance in the attachment.

https://drive.google.com/file/d/1FsH_1gYMCH6KKkAK3152esL1qax8GIYm/view?usp=sharing

 

We made some adjustment to its answer mode prediction 

so that it won't output the unused module name "commonsense" as answer mode.

Please find the attached ppt to modify the dispatcher code accordingly.

 

Sincerely,

Shih-hong 

 

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.iis.sinica.edu.tw/pipermail/most-ai-contest/attachments/20200404/0c2317de/attachment.html>