URajinda/ShweYon-Qwen2.5-Burmese-1.5B-v1.0

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:1.5BQuant:BF16Ctx Length:32kPublished:Dec 30, 2025License:apache-2.0Architecture:Transformer Open Weights Warm

URajinda/ShweYon-Qwen2.5-Burmese-1.5B-v1.0 is a 1.5 billion parameter base language model built upon the Qwen 2.5 architecture, specifically optimized for the Burmese language. This model integrates Burmese tokens directly into its embedding, eliminating the need for a separate tokenizer. It serves as a foundational base model for developing Burmese-centric NLP applications, including chatbots and question-answering systems, through further fine-tuning.

Loading preview...

Overview

URajinda/ShweYon-Qwen2.5-Burmese-1.5B-v1.0 is a 1.5 billion parameter base language model developed by URajinda, built on the Qwen 2.5 architecture. It represents a significant step in the "ShweYon" project, focusing on enhancing the efficiency of Burmese script processing.

Key Differentiators

  • Burmese-centric Optimization: Specifically designed and optimized for the Burmese language.
  • Integrated Tokenization: Unlike previous versions, this model directly incorporates Burmese tokens into its embedding, removing the need for a separate tokenizer.

Purpose and Use Cases

This model is intended as a Foundation Base Model for Burmese language applications. It provides a strong starting point for:

  • Developing Chatbots
  • Building Question Answering systems
  • Other Downstream NLP Tasks requiring further fine-tuning (SFT/RLHF).

Important Note

As a base model, ShweYon-Qwen2.5-Burmese-1.5B-v1.0 requires additional chat fine-tuning to effectively engage in human-like conversations or follow instructions.