| .TH SPHINX_LM_EVAL 1 "2008-05-12" |
| .SH NAME |
| sphinx_lm_eval \- Evaluate perplexity of a transcription |
| .SH SYNOPSIS |
| .B sphinx_lm_eval |
| [\fI options \fR]... |
| .SH DESCRIPTION |
| .PP |
| This program evaluates the perplexity of a text file according to a |
| given language model. The text file is assumed to be in transcript |
| format, i.e. one utterance per line, delimited by <s> and </s>. |
| .TP |
| .B \-help |
| Shows the usage of the tool |
| .TP |
| .B \-lm |
| model file |
| .TP |
| .B \-lmctlfn |
| file listing a set of language models |
| .TP |
| .B \-lmname |
| of language model in \fB\-lmctlfn\fR to use for all utterances |
| .TP |
| .B \-logbase |
| Base in which all log-likelihoods calculated |
| .TP |
| .B \-lsn |
| file to evaluate |
| .TP |
| .B \-lw |
| Language model weight |
| .TP |
| .B \-mmap |
| Use memory-mapped I/O for reading binary LM files |
| .TP |
| .B \-probdef |
| definition file for classes in LM |
| .TP |
| .B \-text |
| string to evaluate |
| .TP |
| .B \-uw |
| Unigram probability weight (interpolated with uniform distribution) |
| .TP |
| .B \-verbose |
| Print details of perplexity calculation |
| .TP |
| .B \-wip |
| Word insertion probability |
| .SH AUTHOR |
| David Huggins-Daines <dhdaines@gmail.com> |
| .SH COPYRIGHT |
| Copyright \(co 2007-2008 Carnegie Mellon University. See the file |
| \fICOPYING\fR included with this package for more information. |
| .br |
|
|
