Qwen2.5-3B-Urdu-Ultimate-Poet: A Specialized Urdu LLM

Developed by Khurram Pervez (Khurram123), this model is a highly specialized version of the Qwen2.5-3B architecture, meticulously fine-tuned for the Urdu language. It leverages a multi-stage training process to achieve versatility across several Urdu-specific tasks.

Key Capabilities

• Urdu News & Information Summarization: Trained on the XL-Sum dataset to provide accurate and concise summaries of complex Urdu news articles.
• Urdu Instruction Following & Chat: Integrated with Urdu-Alpaca data, enabling it to follow user commands, answer questions, and engage in natural conversations in Urdu.
• Classical & Modern Urdu Poetry Generation: Deeply fine-tuned on the Urdu-Poetry-Dataset, allowing it to compose ghazals and nazms with an understanding of Urdu poetic rhythm, vocabulary, and various styles (e.g., Firaq Gorakhpuri).

Good for

• Generating high-quality, contextually relevant Urdu text.
• Summarizing Urdu news and informational content.
• Developing Urdu-speaking chatbots and conversational AI.
• Creating original Urdu poetry, including ghazals and nazms.
• Applications requiring nuanced understanding and generation of the Urdu language, especially in creative writing and information processing.