Overview

Suchinthana/databricks-dolly-15k-sinhala is a Llama 2-based large language model, featuring 7 billion parameters. It has been fine-tuned by Suchinthana using a specialized dataset to enhance its capabilities in the Sinhala language. The fine-tuning process involved 3000 data points from the databricks-dolly-15k-sinhala dataset, executed over 200 steps, which approximates to 1.01 epochs of training.

Key Capabilities

• Sinhala Language Processing: Optimized for understanding and generating text in Sinhala.
• Instruction Following: Benefits from the instruction-tuned nature of the base Dolly-15k dataset, adapted for Sinhala.

Good For

• Applications requiring natural language understanding in Sinhala.
• Generating Sinhala text based on prompts or instructions.
• Research and development in Sinhala NLP.