sonktx/gpt-oss-20b-vi-qa-mxfp4

TEXT GENERATIONConcurrency Cost:1Model Size:20BQuant:FP8Ctx Length:32kPublished:May 21, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The sonktx/gpt-oss-20b-vi-qa-mxfp4 is a 20 billion parameter GPT-OSS model developed by sonktx, fine-tuned for Vietnamese Question Answering tasks. This model was trained using Unsloth and Huggingface's TRL library, enabling faster fine-tuning. It is optimized for efficient performance in question answering within the Vietnamese language domain.

Loading preview...

Model Overview

The sonktx/gpt-oss-20b-vi-qa-mxfp4 is a 20 billion parameter language model developed by sonktx. It is a fine-tuned variant of the unsloth/gpt-oss-20b-unsloth-bnb-4bit base model, specifically optimized for Vietnamese Question Answering (QA) tasks.

Key Capabilities

  • Vietnamese Question Answering: Specialized in understanding and generating responses to questions posed in Vietnamese.
  • Efficient Fine-tuning: Leverages Unsloth and Huggingface's TRL library for accelerated training, achieving approximately 2x faster fine-tuning compared to standard methods.
  • GPT-OSS Architecture: Built upon the GPT-OSS model family, providing a robust foundation for language understanding and generation.

Good For

  • Vietnamese QA Systems: Ideal for applications requiring accurate and contextually relevant answers to questions in Vietnamese.
  • Research and Development: Suitable for researchers and developers exploring efficient fine-tuning techniques for large language models, particularly in low-resource or specialized language domains.
  • Resource-Efficient Deployment: The use of Unsloth suggests potential for more efficient inference and deployment, making it a candidate for applications where computational resources are a consideration.