CLIP-BERT Multi-Label Classifier
Multi-label image classification combining CLIP vision-text embeddings with BERT for enhanced semantic understanding. Uses label GCN and co-occurrence matrix to model tag correlations.
AI/ML Engineer
Crafting intelligent systems with deep learning, LLMs, and modern AI. Currently actively seeking internships 2026.
I'm a CS student who loves building AI systems — from fine-tuning LLMs to RAG pipelines and computer vision projects.
Still learning, always building. Currently looking for internships where I can contribute and grow.
suzu
no current activity
Multi-label image classification combining CLIP vision-text embeddings with BERT for enhanced semantic understanding. Uses label GCN and co-occurrence matrix to model tag correlations.
Conversational AI tutor for IELTS preparation powered by OpenAI GPT and ElevenLabs voice synthesis. Supports adaptive band scoring from 1–9 with one-shot prompt engineering.
Fine-tuned YOLOv10 on a custom fruit dataset with Streamlit inference UI. Includes training pipeline, evaluation metrics, and model weight export.
Open to internships and collaborations. Just say hi.