Muhammad Fahad

I’m a Computer Science graduate specializing in deep learning and applied machine learning, with a strong focus on generative models, Vision Transformers, and vision-language modeling.

I’ve built generative models (GANs, Diffusion) and vision-language models from base PyTorch, and I’ve fine-tuned large language models (BERT, GPT-2, T5) and VLMs (Qwen2-VL) using LoRA/QLoRA on multi-GPU setups.

My work also includes research on Bayesian deep learning for uncertainty quantification in medical signal processing—aiming for calibrated, auditable predictions in clinically reliable settings.

I care deeply about end-to-end quality: model training, evaluation, and deployment. I’ve shipped interactive experiences too—Gradio apps for DDPM and Qwen2-VL inference, plus Streamlit deployment for an LLM fine-tuning suite—because I like closing the loop from research to real use.