Qwen3 Speculative Decoding on the DGX Spark: Two Models, Four Methods, One Surprising LessonA beginner's guide to one of the most practical speed techniques for AI inference, tested on real hardware with real numbers.15 min read·shaunliew.hashnode.devMay 18
Running Qwen3-VL on DGX Spark: Transformers vs vLLM vs SGLangA hands-on inference benchmark across three frameworks on a desk-sized AI supercomputer.14 min read·shaunliew.hashnode.devMay 16
Fixing CUDA PTX Error When Running Qwen3-VL with vLLM on H200Running vision-language models like Qwen3-VL with vLLM on high-end GPUs should be straightforward. Except when it's not. The Problem I was setting up Qwen3-VL-8B-Instruct on our H200 cluster (8x H200, 143GB VRAM each) when I hit this error: vllm ser...4 min read·shaunliew.hashnode.devDec 7, 2025
Understanding neural style transferNeural Style Transfer: Turning the Mona Lisa into a Picasso Have you ever wondered what the Mona Lisa would look like if Picasso had painted it? Neural style transfer makes this artistic dream a reality by combining the content of one image with the ...9 min read·shaunliew.hashnode.devJul 29, 2025
Performing an adversarial attack on imagesExploring adversarial attacks - where tiny, imperceptible modifications can completely deceive even the smartest neural networks From Autoencoders to Adversarial Attacks: A New Kind of Magic After my fascinating journey with Variational Autoencoders,...10 min read·shaunliew.hashnode.devJul 29, 2025