IBM/transition-amr-parser

SoTA Abstract Meaning Representation (AMR) parsing with word-node alignments in Pytorch. Includes checkpoints and other tools such as statistical significance Smatch.

50
/ 100
Established

Implements transition-based graph decoding with Structured-BART and multilingual variants (mBART) to jointly model AMR graph structures and word-node alignments in Penman notation. Provides significance testing via Smatch bootstrapping and Maximum Bayesian Smatch Ensemble (MBSE) for robust evaluation, with pretrained checkpoints for sentence-level, document-level (handling coreference via `:same-as` edges), and multilingual parsing across English, German, Chinese, and Italian.

273 stars. No commits in the last 6 months.

Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 22 / 25

How are scores calculated?

Stars

273

Forks

56

Language

Python

License

Apache-2.0

Last pushed

Sep 17, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/nlp/IBM/transition-amr-parser"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.