Hi, I'm Huang
RSS FeedI build things for the mobile, for the web, for language learners, and for fun. Below are the projects I work on — I also write notes about what I learn along the way.
Social Links:
Projects

OctoCounts
octocounts.comA free SLOC counter for public GitHub repositories, as a web app and a browser extension.
| Type | Web app + browser extension (Chrome / Edge / Firefox) |
|---|---|
| Price | Free |
| Backend | Rust (Axum + Tokio) |
| Counting engine | tokei, across 200+ languages |
| Caching | Reports cached by commit SHA — repeat queries are instant |
| Export | JSON / plain text / PNG badge |

EchoPod
echopod.clothpath.comA language-learning app that turns your favorite podcasts into interactive bilingual study material with AI subtitles.
| Platforms | iOS (App Store); Android (alpha program) |
|---|---|
| Core features | Synced AI subtitles, guided echoing sessions |
| Trains | Listening and speaking |

Ketsuin 結印
ketsuin.clothpath.comA web-based input method that lets you type with ninja hand signs, detected entirely in your browser.
| Input | 12 classic ninja seals via webcam |
|---|---|
| Detection | YOLOX-Nano on ONNX Runtime WebAssembly, fully in-browser |
| Privacy | No data leaves your device |
| Text engine | T9 predictive engine maps seal sequences to text |
| Game mode | Challenge mode with a global leaderboard |

QwenASR
github.com/huanglizhuo/QwenASRA fast, pure-Rust, CPU-only inference engine for Qwen3-ASR speech-to-text, tuned for Apple Silicon.
| Tech stack | Pure Rust, CPU-only — no Python, no GPU, no tensor framework, only libc |
|---|---|
| Optimization | Hand-written NEON / Accelerate / AMX-aware kernels for Apple Silicon |
| Performance | Transcribes a 28-second clip in 613 ms on an M5 Pro (46× realtime) |
| Features | Offline & streaming transcription, live capture with VAD, SRT/VTT subtitles, structured JSON, forced alignment |
| Install | cargo install qwen-asr-cli |
Recent Posts
Hello World
First post on this blog.