TAUR-dev/rankalign-v6-gemma-2-2b-d0.15-e2-hc-b2d-dbl-all-tco-p0-nv1-ng1-fsx
The TAUR-dev/rankalign-v6-gemma-2-2b-d0.15-e2-hc-b2d-dbl-all-tco-p0-nv1-ng1-fsx model is a 2.6 billion parameter language model fine-tuned from Google's Gemma-2-2b architecture. It is specifically optimized for hypernym prediction tasks, demonstrating specialized performance in identifying hierarchical relationships between concepts. This model is part of the rankalign project, focusing on improving the alignment and ranking of linguistic relationships.
Loading preview...
Model Overview
This model, rankalign-v6-gemma-2-2b-d0.15-e2-hc-b2d-dbl-all-tco-p0-nv1-ng1-fsx, is a fine-tuned checkpoint derived from the google/gemma-2-2b base model, developed as part of the rankalign project. It features 2.6 billion parameters and a context length of 8192 tokens.
Key Training Details
- Base Model:
google/gemma-2-2b - Version: v6
- Task:
hypernym-concat-bananas-to-dogs-double-all - Epochs: 2
- Delta: 0.15
- Typicality Correction: Online
- Preference Loss Weight: 0
- NLL Validator Weight: 1
- NLL Generator Weight: 1
Primary Focus
This model is specifically trained and optimized for hypernym prediction tasks, which involve identifying broader categories or superordinates for given concepts. The training configuration, including the hypernym-concat-bananas-to-dogs-double-all task, indicates a specialized focus on understanding and generating hierarchical semantic relationships. It is designed for research and applications requiring precise hypernym identification within a controlled linguistic framework.