v2.2.0 (2026-01-08): spaCy Pipeline Integration

v2.2.0 (2026-01-08)

✨ New Features

spaCy Pipeline Integration

Hashformers can now be used directly as a spaCy pipeline component! This enables seamless integration with spaCy workflows and makes it easy to combine word segmentation with other NLP tasks.

Installation:

pip install hashformers[spacy]

Usage:

import spacy
import hashformers.spacy  # registers the 'hashformers' component

nlp = spacy.blank("en")
nlp.add_pipe("hashformers", config={"model": "distilgpt2"})

doc = nlp("#weneedanationalpark")
print(doc._.segmented)  # 'we need a national park'

Configuration options:

• model: Hugging Face model name (default: "distilgpt2")
• device: Device to run on, "cuda" or "cpu" (default: "cuda")
• gpu_batch_size: Batch size for GPU processing (default: 1000)

📦 Package Improvements

• Added long_description from README for better PyPI presentation
• Added PyPI classifiers for improved discoverability
• Added python_requires=">=3.8" specification
• Added keywords for search optimization
• Added optional [spacy] extras for spaCy dependency

📚 Documentation

• Added spaCy integration section to README
• Prepared spaCy Universe submission

---

v2.1.0

• Previous stable release
• Core word segmentation with TransformerWordSegmenter
• Beam search algorithm with optional reranker
• Support for any Hugging Face autoregressive model