---
name: "ngrammake"
section: "1"
platform: "linux"
package: "libngram-tools"
version: "1.3.2-3"
release: "Debian 10 (Buster)"
platforms: ["linux"]
url: "https://new.carta.tech/ngrammake"
---

# ngrammake(1)

Make ngram model from input count file.

## NAME

ngrammake - Make ngram model from input count file.

## DESCRIPTION

Make ngram model from input count file.

-

  ## SYNOPSIS

ngrammake \[--options\] \[in.fst \[out.fst\]\]

## OPTIONS

**--backoff**: type = bool, default = false

-

  Use backoff smoothing (default: method dependent

**--backoff\_label**: type = int64, default = 0

-

  Backoff label

**--bins**: type = int64, default = **-1**

-

  Number of bins for katz or absolute discounting

**--check\_consistency**: type = bool, default = false

-

  Check model consistency

**--count\_of\_counts**: type = string, default = ""

-

  Read count-of-counts from file

**--discount\_D**: type = double, default = **-1**

-

  Absolute discount value D to use

**--interpolate**: type = bool, default = false

-

  Use interpolated smoothing (default: method dependent)

**--method**: type = string, default = "katz"

-

  One of: "absolute", "katz", "kneser\_ney", "presmoothed", "unsmoothed", "katz\_frac", "witten\_bell"

**--norm\_eps**: type = double, default = 0.001

-

  Normalization check epsilon

**--witten\_bell\_k**: type = double, default = 1

-

  Witten-Bell hyperparameter K

**--help**: type = bool, default = false

-

  show usage information

**--helpshort**: type = bool, default = false

-

  show brief usage information

**--tmpdir**: type = string, default = "/tmp"

-

  temporary directory

**--v**: type = int32, default = 0

-

  verbosity level

**--fst\_align**: type = bool, default = false

-

  Write FST data aligned where appropriate

**--fst\_default\_cache\_gc**: type = bool, default = true

-

  Enable garbage collection of cache

**--fst\_default\_cache\_gc\_limit**: type = int64, default = 1048576

-

  Cache byte size that triggers garbage collection

**--fst\_read\_mode**: type = string, default = "read"

-

  Default file reading mode for mappable files

**--fst\_verify\_properties**: type = bool, default = false

-

  Verify FST properties queried by TestProperties

**--save\_relabel\_ipairs**: type = string, default = ""

-

  Save input relabel pairs to file

**--save\_relabel\_opairs**: type = string, default = ""

-

  Save output relabel pairs to file

**--end\_symbol**: type = string, default = "\</s\>"

-

  Class label for sentence start

**--start\_symbol**: type = string, default = "\<s\>"

-

  Class label for sentence start

**--fst\_compat\_symbols**: type = bool, default = true

-

  Require symbol tables to match when appropriate

**--fst\_field\_separator**: type = string, default = "      "

-

  Set of characters used as a separator between printed fields

**--fst\_error\_fatal**: type = bool, default = true

-

  FST errors are fatal; o.w. return objects flagged as bad: e.g., FSTs: kError property set, FST weights: not a Member()

**--ngram\_error\_fatal**: type = bool, default = true

-

  NGram errors are fatal; o.w. return objects flagged as bad:  e.g., NGramModel::Error() returns true.

**--fst\_weight\_parentheses**: type = string, default = ""

-

  Characters enclosing the first weight of a printed composite weight (e.g., pair weight, tuple weight and derived classes) to ensure proper I/O of nested composite weights; must have size 0 (none) or 2 (open and close parenthesis)

**--fst\_weight\_separator**: type = string, default = ","

-

  Character separator between printed composite weights; must be a single character