buddhist-nlp/gemma-2-mitra-it

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:9BQuant:FP8Ctx Length:16kPublished:Nov 7, 2024Architecture:Transformer0.0K Warm

The buddhist-nlp/gemma-2-mitra-it is a 9 billion parameter instruction-tuned language model based on the Gemma 2 architecture, specifically fine-tuned for translation tasks. It leverages the gemma2-mitra-base model and is optimized to translate input sentences into a specified target language. This model is designed for developers requiring a specialized translation LLM with a 16384 token context length.

Loading preview...

Overview

The buddhist-nlp/gemma-2-mitra-it model is a 9 billion parameter instruction-tuned language model built upon the Gemma 2 architecture. It is specifically derived from gemma2-mitra-base and has undergone further fine-tuning with a focus on translation instructions. The model is designed to perform language translation tasks efficiently.

Key Capabilities

  • Specialized Translation: The primary function of this model is to translate text from one language to another, based on explicit instructions.
  • Instruction-Tuned: It responds to a specific prompting template for translation, making its usage straightforward for its intended purpose.
  • Gemma 2 Architecture: Benefits from the underlying capabilities and performance characteristics of the Gemma 2 model family.

Usage and Prompting

Users should employ a specific template for prompting the model for translation:

Please translate into <target_language>: <input_sentence> 🔽 Translation::

Line breaks within the input should be replaced with the '🔽' character, and '#' serves as a stop token during generation. For general operational details and how to run Gemma 2 models, users are directed to the google/gemma-2-9b repository.