#crf #conditional #fields #random #port #crfsuite #pure

crfs

Pure Rust port of CRFsuite: a fast implementation of Conditional Random Fields (CRFs)

5 releases

0.2.0 Nov 18, 2021
0.1.3 Jan 31, 2021
0.1.2 Jan 31, 2021
0.1.1 Jan 31, 2021
0.1.0 Jan 31, 2021

#493 in Machine learning

Download history 22/week @ 2024-11-20 29/week @ 2024-11-27 33/week @ 2024-12-04 17/week @ 2024-12-11 23/week @ 2024-12-18 4/week @ 2025-01-01 21/week @ 2025-01-08 13/week @ 2025-01-15 3/week @ 2025-01-22 21/week @ 2025-02-05 29/week @ 2025-02-12 42/week @ 2025-02-19 33/week @ 2025-02-26 9/week @ 2025-03-05

113 downloads per month
Used in 7 crates (5 directly)

MIT license

31KB
725 lines

crfs-rs

CI codecov Crates.io docs.rs

Pure Rust port of CRFsuite: a fast implementation of Conditional Random Fields (CRFs)

Currently only support prediction, model training is not supported. For training you can use crfsuite-rs.

Installation

Add it to your Cargo.toml:

[dependencies]
crfs = "0.1"

Performance

Performance comparsion with CRFsuite on MacBook Pro (13-inch, M1, 2020) 16GB

$ cargo bench --bench crf_bench -- --output-format bencher
test tag/crfs ... bench:        1449 ns/iter (+/- 5)
test tag/crfsuite ... bench:        2154 ns/iter (+/- 14)

Last updated on 2021-11-18.

License

This work is released under the MIT license. A copy of the license is provided in the LICENSE file.

Dependencies

~0.7–1.1MB
~20K SLoC