abdulmateenchitrali/TorkhowGPT-v2

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:1.5BQuant:BF16Ctx Length:32kPublished:May 14, 2026Architecture:Transformer Warm

TorkhowGPT-v2 by Abdul Mateen is a 1.5 billion parameter large language model, notable as the world's first LLM specifically developed for the Khowar language. With a 32768 token context length, it specializes in Khowar language processing, including translation tasks. This model is primarily designed to serve the Khowar-speaking community in Chitral and Gilgit-Baltistan, Pakistan.

Loading preview...

TorkhowGPT-v2: The First Khowar Language LLM

TorkhowGPT-v2, developed by Abdul Mateen, marks a significant advancement as the world's first large language model (LLM) built specifically for the Khowar language. With 1.5 billion parameters and a substantial 32768 token context length, this model is engineered to address the unique linguistic needs of the Khowar-speaking population.

Key Capabilities

  • Khowar Language Processing: Specialized in understanding and generating text in Khowar.
  • Translation: Capable of translating between Khowar and English, demonstrated by examples like "water" to Khowar and "loth" (Khowar for 'water') to English.
  • Community Focus: Directly supports the Khowar language, spoken by over 500,000 people in Chitral and Gilgit-Baltistan, Pakistan.

Good For

  • Khowar Language Translation: Ideal for applications requiring translation services for Khowar speakers.
  • Linguistic Research: Valuable for researchers studying low-resource languages and their computational modeling.
  • Local Content Generation: Useful for creating digital content in the Khowar language, fostering digital inclusion for the community.