TAUR-dev/rankalign-v6-gemma-2-2b-d0.15-e2-hc-b2d-dbl-all-tcs-p0-nv1-ng1-fsx-sm0.1

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:2.6BQuant:BF16Ctx Length:8kPublished:Apr 6, 2026Architecture:Transformer Warm

The TAUR-dev/rankalign-v6-gemma-2-2b-d0.15-e2-hc-b2d-dbl-all-tcs-p0-nv1-ng1-fsx-sm0.1 model is a 2.6 billion parameter Gemma-2-2b base model fine-tuned within the rankalign project. This model is specifically optimized for hypernym-concat-bananas-to-dogs-double-all tasks, focusing on semantic relationship understanding. It is designed for specialized natural language processing applications requiring precise hierarchical concept identification.

Loading preview...

Model Overview

This model, rankalign-v6-gemma-2-2b-d0.15-e2-hc-b2d-dbl-all-tcs-p0-nv1-ng1-fsx-sm0.1, is a fine-tuned checkpoint derived from the google/gemma-2-2b base model. It is part of the rankalign project, which focuses on advanced alignment techniques for language models.

Key Training Details

The model underwent specific fine-tuning for a task identified as hypernym-concat-bananas-to-dogs-double-all. Training involved:

  • Base Model: google/gemma-2-2b
  • Version: v6 of the rankalign project's fine-tuning process.
  • Epochs: Trained for 2 epochs.
  • Delta: A delta value of 0.15 was applied during training.
  • Typicality Correction: Utilized a 'self' typicality correction method.
  • Loss Weights: Preference loss weight was 0, while NLL validator and generator weights were both 1.
  • Semi-supervised Ratio: A semi-supervised training ratio of 0.1 was used.

Use Cases

This model is particularly suited for research and development in:

  • Hypernym Detection: Identifying hierarchical relationships between concepts, as indicated by its training task.
  • Semantic Relationship Analysis: Tasks requiring a nuanced understanding of how words and concepts relate to each other.
  • Specialized NLP Research: Exploring the effects of specific fine-tuning strategies on base models for targeted linguistic tasks.