Experience

Completed a two-month remote AI/ML internship focused on language-model fine-tuning, model evaluation, and hands-on experimentation.

• Fine-tuned a LLaMA 3.2 3B instruction model on a medical reasoning dataset using Python, Hugging Face, Unsloth, and PEFT/QLoRA.
• Evaluated model responses using ROUGE-L, BERTScore, and manual review, while documenting response-quality issues and cases of hallucination.

• Python
• Hugging Face
• Unsloth
• PEFT
• QLoRA
• LLaMA 3.2

Taught Computer Science to students in Grades 6–10, explaining technical concepts in a clear and practical way and adapting lessons to different learning levels.