Model Overview

Suchinthana/Sinhala-Translate-and-Dolly-Llama-7b is a 7 billion parameter language model built upon the Llama architecture. It has been specifically fine-tuned to enhance its capabilities in Sinhala-English translation.

Key Capabilities

• Sinhala-English Translation: The model was fine-tuned using a 20,000-entry Sinhala-English translation dataset, applied in both English-to-Sinhala and Sinhala-to-English directions.
• Instruction Following: Initial fine-tuning with the Dolly 3k dataset provides a foundation for general instruction-following, which was then specialized for translation.

Use Cases

This model is particularly well-suited for applications requiring translation between Sinhala and English. Its specialized training makes it a strong candidate for:

• Translating text from Sinhala to English.
• Translating text from English to Sinhala.
• Developing applications that need to process or generate content in both languages.