OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Berkeley Neural Parser

Berkeley Neural Parser is a Python NLP parser that generates constituency trees using pretrained neural models for 11 languages. Distributed as benepar, it integrates with spaCy for raw-text pipelines and also provides an NLTK interface for prepared sentences and treebanks.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Berkeley Neural Parser — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when an NLP workflow needs phrase-structure trees. For raw text, the README recommends its spaCy integration; the NLTK interface is intended for workflows that already tokenize and split sentences.

Who it is for

  • NLP developers adding phrase-structure information to text-processing pipelines.
  • Researchers working with multilingual parsing, pre-tokenized data, or treebanks.
  • Teams building syntax-aware text analysis that can use constituency trees.

Documented uses

  • Generate constituency trees for sentences in a supported language.
  • Retrieve parse strings, labels, constituents, and parent-child relationships through spaCy extensions.
  • Parse pre-tokenized sentences or batches of sentences through the NLTK interface.
  • Train or use a custom parser based on a selected pretrained model.
Before you choose

Requirements & limits

The repository specifies an MIT license. Installation requires Python 3.6 or newer and PyTorch 1.6 or newer; parsing models are downloaded separately.

  • The README lists pretrained models for 11 languages; select a parsing model separately for the intended language.
  • For raw text, spaCy is the recommended route because it handles tokenization and sentence segmentation.
  • The NLTK interface expects prepared sentence input for accuracy-oriented use; raw string parsing is described as suitable only for debugging or interactive use.
  • Inference uses PyTorch, with GPU use dependent on a GPU being available to PyTorch.
  • The project supports custom parser training, with pretrained model choice as part of that workflow.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up the parser with spaCy for English

Install benepar, prepare an English spaCy pipeline, and load a parsing model as described in the README.

Before you start

  • Python 3.6 or newer
  • PyTorch 1.6 or newer
  • spaCy and an English spaCy model; the README example uses en_core_web_md

Steps

  1. Install the parser with pip install benepar.

  2. Install the English spaCy model with python -m spacy download en_core_web_md.

  3. Download a parsing model in Python with benepar.download('benepar_en3').

  4. Load the spaCy model and add the benepar component using the configuration appropriate to the installed spaCy major version.

Expected outcome

A spaCy pipeline configured with a constituency parser; the README shows sentence parse strings and labels accessible through spaCy extensions.

Guide limitations

  • The spaCy model supplies tokenization and sentence segmentation; it is separate from the benepar parsing model.
  • The parsing model must be downloaded separately.
  • The README gives different component-registration code for spaCy 2 and later versions.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
17 October 2026
Original website title
nikitakit/self-attentive-parser · Official homepage title · retained in its source language
How we review sourcesSuggest a correction