# BrikerMan/Kashgari

Kashgari is a production-level NLP Transfer learning framework built on top of tf.keras for text-labeling and text-classification, includes Word2Vec, BERT, and GPT2 Language Embedding.

Repository: https://github.com/BrikerMan/Kashgari
Canonical: https://ross.abutalabs.com/products/kashgari
Homepage: http://kashgari.readthedocs.io/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: nlp, sequence-labeling, text-classification, bert-model, ner, machine-learning, nlp-framework, named-entity-recognition, gpt-2, transfer-learning, seq2seq, bert, text-labeling
Last push: 2024-09-03T21:05:29+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2784, "days_push": 729, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2381, forks 431 (observed 2026-08-28T04:06:42.292378+00:00)

## What it is
Kashgari is a Keras/TensorFlow 2-based NLP transfer learning framework for building text labeling (NER, PoS) and text classification models quickly. It bundles pre-trained BERT, GPT-2, and Word2Vec embeddings and supports exporting models in SavedModel format for production deployment.

## Use cases
- train a named entity recognition model with BERT embeddings
- build a text classifier in a few lines of Keras code
- do part-of-speech tagging with pre-trained language models
- fine-tune BERT or GPT-2 embeddings for sequence labeling
- export an NLP model as SavedModel for production serving
- experiment with different embeddings and model architectures for NLP tasks

## When to choose
- you want a simple, human-friendly Keras API for NER, PoS, or text classification
- you need built-in BERT/GPT-2/Word2Vec transfer learning without wiring transformers yourself
- you want to export trained models to TensorFlow SavedModel for production

## When to avoid
- you need PyTorch or the latest Hugging Face transformers ecosystem
- you require cutting-edge model architectures beyond sequence labeling and classification
- you need a project with frequent recent development activity

## Facets
- artifact type: framework
- maturity: maintenance
- function: nlp, machine-learning, deep-learning
- domain: machine-learning, deep-learning
- platform: python
- tags: transfer-learning, named-entity-recognition, text-classification, sequence-labeling, bert, gpt-2, word2vec, keras, tensorflow, natural-language-processing

## Member repositories
- BrikerMan/Kashgari (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:42.292378+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:34:44.223231+00:00, confidence not recorded.
  - readme: https://github.com/BrikerMan/Kashgari (fetched 2026-08-28T04:06:42.292378+00:00, sha 680567b473dc)
  - registry_pypi: https://pypi.org/pypi/kashgari/json (fetched 2026-08-29T10:16:10.904540+00:00, sha 153581112d09)
- Data as of 2026-08-30T08:39:29.467469+00:00.