← Latest products

I trained a 9M speech model to fix my Mandarin tones

A tiny on-device speech model to grade Mandarin pronunciation and tones. Built this because tones are killing my spoken Mandarin and I can't reliably hear my own mistakes.

Visit simedw.comexternal site

Reactions (1)

Similar products

by description similarity

  1. I built a tiny LLM to demystify how language models work

    github.com

    A ~9M parameter LLM that talks like a small fish.

    Hacker News915 pts

  2. I built a Chinese learning app that teaches through sentence patterns

    doudou-chinese.com

    Two apps for learning Mandarin Chinese: 句式 teaches essential sentence patterns with AI feedback, 听听 trains tones and HSK vocabulary, and 句 loops level-appropriate phrases for audio immersion.

    Hacker News9 pts

  3. A real-time speech-language model for $10 of training

    tincans.ai

    Hacker News5 pts

  4. I build an AI Language Speaking that make 3k a month

    fluentpal.app

    Practise real conversations out loud with friendly AI characters. Get a pronunciation score on every sentence, instant translations and bite-size courses in English, German, Chinese, Korean and Japanese.

    Hacker News14 pts

  5. I built a sub-500ms latency voice agent from scratch

    ntik.me

    I built a voice agent from scratch that averages ~400ms end-to-end latency (phone stop → first syllable).

    Hacker News570 pts