HaS (Hide and Seek) is an on-device privacy model providing a complete pipeline from entity recognition to anonymization and restoration. — 📦 0.6B parameters, Q80 quantization, 639 MB — 🔒 Data never leaves device — local inference, no network required — 🌍 8 languages natively supported: Chinese, English, Portuguese, French, Spanish, German, Korean, Japanese — ⚡ Apple M4 benchmark**: prefill 1,600–2,800 tok/s, generation 96–120 tok/s Traditional anonymization (regex, Presidio, etc.) only does pattern matching. HaS is an on-device Agentic privacy pipeline — a set of composable atomic capabilities that solve multi-turn consistency, reversible restoration, and post-anonymization data usability. The same entity often appears in multiple forms. HaS automatically recognizes they refer to the same object and unifies them under one ID: This ensures anonymized text remains logically coherent — LLMs seeing multiple know it’s the same company. Critical for multi-turn conversations and long document chunking: entity IDs remain globally consistent across turns and chunks. — Listens on http://127.0.0.1:8080/v1 by default — OpenAI Chat Completions compatible API — ~1.56 GB total memory with…
Модальности:
Генерация текста
Области применения:
Диалог / чат Распознавание сущностей
Задача: Генерация текста
Автор: xuanwulab
Теги: gguf, privacy, anonymization, ner, pii, hide-and-seek, on-device, qwen3
Лайков: 3 | Загрузок: 656
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
