Running Qwen3.6 and Qwen3.8 in LM Studio on an M5 Pro
Using 4-bit MLX models locally on Apple Silicon while managing unified memory, context length, and thinking settings.
Tag · 2 stories
Using 4-bit MLX models locally on Apple Silicon while managing unified memory, context length, and thinking settings.
How I turned my experience building an iOS product into a native SwiftUI system DNS manager with DoH and DoT support.