Ninja2005/FoodExtract-gemma-3-270m-fine-tune-v1

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:0.3BQuant:BF16Context Size:32kPublished:Sep 8, 2026License:gemmaArchitecture:Transformer Featherless Exclusive Cold

Ninja2005/FoodExtract-gemma-3-270m-fine-tune-v1 is a 0.3 billion parameter language model, fine-tuned from Google's Gemma 3 270M, specifically designed for extracting food and drink related information from raw text. It classifies text as food/drink, tags items, and extracts lists of edible food and drink items. This model excels at filtering large text datasets, such as image captions, for food and drink content.

Loading preview...

FoodExtract-v1: Specialized Food and Drink Information Extraction

This model, fine-tuned from Google's Gemma 3 270M, is a compact 0.3 billion parameter language model engineered for precise extraction of food and drink related data from unstructured text. It processes raw text inputs to perform several key functions, making it highly suitable for data filtering and organization tasks.

Key Capabilities

  • Classification: Determines if input text is related to food or drink.
  • Tagging: Assigns relevant tags (e.g., 'ingredient list', 'menu', 'food items', 'drink items') to the text for fast categorization.
  • Item Extraction: Identifies and lists specific edible food and drink items mentioned in the text.
  • Optimized Output: Generates condensed, structured outputs rather than verbose JSON, reducing token generation and improving efficiency. Helper functions are provided for condensing and uncondensing outputs.

Training and Dataset

The model was trained using Supervised Fine-Tuning (SFT) via Hugging Face's TRL library on the FoodExtract-1k dataset. This dataset comprises 1400 samples of raw text paired with structured food extractions, originally generated by gpt-oss-120b.

Good For

  • Filtering large datasets: Ideal for sifting through extensive text corpora, such as image caption datasets (e.g., DataComp-1B), to isolate food and drink related entries.
  • Structured Data Generation: Transforming unstructured text into organized, extractable food and drink lists and classifications.
  • Resource-constrained environments: Its small parameter count (0.3B) makes it efficient for deployment where computational resources are limited.

Finetunes
1 models