small-models-for-glam/Qwen3-0.6B-SFT-name-parser-yaml

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:0.8BQuant:BF16Ctx Length:32kPublished:Oct 1, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Warm

The small-models-for-glam/Qwen3-0.6B-SFT-name-parser-yaml is a 0.6 billion parameter Qwen3-based causal language model developed by small-models-for-glam. Fine-tuned using Supervised Fine-Tuning (SFT) with TRL, this specialized model excels at parsing cultural heritage person names into a structured YAML format. It accurately extracts first name, last name, middle names, temporal data, titles, and extra information from diverse international name patterns, making it ideal for digital humanities, library science, and archive management applications.

Loading preview...

Overview

This model, small-models-for-glam/Qwen3-0.6B-SFT-name-parser-yaml, is a specialized fine-tuned version of the Qwen3-0.6B base model. Developed by small-models-for-glam, it is specifically designed to parse person names from cultural heritage contexts (libraries, archives, museums) into a structured YAML format. The model was trained using Supervised Fine-Tuning (SFT) with Hugging Face's TRL library.

Key Capabilities

  • Structured Name Parsing: Extracts first_name, last_name, middle_names, temporal information (birth/death/flourished dates), titles, and extra_info into a consistent YAML output.
  • Diverse Name Pattern Support: Handles a wide variety of name formats, including basic, complex (e.g., Baron William Henry Ashe A'Court Heytesbury, c. 1809-1891), mononyms, initials, diacritics, and non-Western names (e.g., Chinese).
  • Multi-cultural Training: Trained on synthetic data covering names from English, French, German, Italian, Spanish, Dutch, Arabic, and Chinese contexts, along with diverse temporal data and titles.
  • Robust Performance: Demonstrates strong performance in identifying and structuring diverse international name formats, temporal data, titles, and complex surname patterns.

Intended Use Cases

This model is primarily intended for applications within cultural heritage domains:

  • Digital Humanities: Processing historical person names found in manuscripts and documents.
  • Library Science: Cataloging and standardizing author names in bibliographic records.
  • Archive Management: Structuring person names within archival finding aids.
  • Museum Collections: Organizing creator and subject names in cultural heritage databases.

Limitations

  • Primarily trained on Western and East Asian name patterns; may struggle with very rare or highly specialized naming conventions.
  • Temporal date parsing assumes Gregorian calendar years and has limited support for ancient or historical dating systems (e.g., BCE, regnal years).

Finetunes
1 models