<html>
<head>
<meta content="text/html; charset=gbk" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
On 5/14/2012 8:02 PM, mvp-songyoung wrote:
<blockquote
cite="mid:1af95c8d.6d67.1374e72c757.Coremail.mvp-songyoung@163.com"
type="cite">
<div
style="line-height:1.7;color:#000000;font-size:14px;font-family:arial">
<div>I have tried the -simple-classes option. It seems that my
models do not satisfy its requirements as I get such warning:</div>
<div>
<div>...</div>
<div>./LM/xx.class: line 6122: word holidays has multiple class memberships<br>
./LM/xx.class: line 6122: word still has multiple class memberships<br>
./LM/xx.class: line 6122: word five has multiple class memberships<br>
./LM/xx.class: line 6122: word form has multiple class memberships
</div>
</div>
<div>...</div>
<div>I merged word classes for LM1 LM2 and LM3 from three
different corpus separately. So it can not avoid that they
have some duplicate words between each other. And I want to
use interpolated class-based LM in my decoding task. How to
operate it? Thank you</div>
<br>
</div>
</blockquote>
In that case you could try<br>
<br>
- converting the class-based LM into a purely word-based LM (ngram
-expand-classes)<br>
- generating N-best from lattices and rescoring those instead <br>
<br>
Andreas<br>
<br>
</body>
</html>