s3nh/EduHelp-8B

EduHelp-8B is an 8 billion parameter decoder-only language model fine-tuned from the Qwen3-8B base model by s3nh. Optimized using PEFT (LoRA) on the Education-Young-Children dataset, it specializes in providing gentle, age-appropriate explanations and basic tutoring for young learners. This model is designed for simple explanations, basic arithmetic, and vocabulary support, primarily in English, with a context length of 32768 tokens.