{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiaf77gsriuibnfknpua2puwiqf4iotstn447uwmexs53nd4xvza4e",
    "uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3mlkmutl3uyc2"
  },
  "path": "/t/proxyface-lightweight-local-llm-avatar-with-real-time-pixel-art-emotions/175909#post_1",
  "publishedAt": "2026-05-11T06:14:47.000Z",
  "site": "https://discuss.huggingface.co",
  "tags": [
    "ProxyFace.com",
    "Model weights, dataset, and notebook"
  ],
  "textContent": "We are open-sourcing **ProxyFace.com** , a browser-based system that augments LLM interactions with real-time pixel art avatars expressing 8 distinct emotions. The entire pipeline runs locally without external API dependencies.\n\n**Technical approach**\nThe emotion classifier uses **TinyBERT** , chosen for its efficiency in browser environments while maintaining solid performance on short-sentence inputs.\n\n**Research artifacts**\n\n  * **Dataset** : 3,200 labeled sentences for emotion classification\n\n  * **Training** : Google Colab notebook (free T4 tier)\n\n  * **Assets** : Model weights, dataset, and notebook\n\n\n\n\nSee screenshot:\n\nWe are sharing these resources to invite community contributions that improve both accuracy and inference speed. If you experiment with architectural changes, additional data, or optimization strategies, we welcome PRs and derived models.\n\nThank you for your attention!",
  "title": "ProxyFace: Lightweight Local LLM Avatar with Real-Time Pixel Art Emotions"
}