NAME¶

evaluate-ngram - evaluates n-gram language model

SYNOPSIS¶

evaluate-ngram [Options]

DESCRIPTION¶

Evaluates the performance of an n-gram language model. It also supports various n-gram language model conversions, including changes in order, vocabulary, and file format.

Filename argument can be an ASCII file, a compressed file (ending in .Z or .gz), or '-' to indicate stdin/stdout.

OPTIONS¶

-h, -help

Print this message.

-verbose <int>

Set verbosity level.

Default: 1

-o, -order <int>

Set the n-gram order of the estimated LM.

Default: 3

-v, -vocab <file>

Fix the vocab to only words from the specified file.

-l, -lm <file>

Load specified LM.

-cl, -compile-lattices <file>

[SLS] Compile lattices into a binary format.

-wb, -write-binary <boolean>

Write LM/counts files in binary format.

Default: false

-wv, -write-vocab <file>

Write LM vocab to file.

-wl, -write-lm <file>

Write ARPA backoff LM to file.

-ep, -eval-perp <files>

Compute test set perplexity.

-ew, -eval-wer <files>

Compute test set lattice word error rate.

-em, -eval-margin <files>

Compute test set lattice margin.

SEE ALSO¶

estimate-ngram(1), interpolate-ngram(1)