T0mSIlver

3 ms·

1. 1.▲
   by T0mSIlver 2mo ago

   Unsloth Q4_K_M on a single 3090, llama.cpp "Generate an SVG of a pelican riding a bicycle" first try https://www.reddit.com/r/LocalLLaMA/comments/1voa3ch/comment...
2. 2.▲
   Qwen 3.8 27B Pelican(reddit.com)

   3 points by T0mSIlver 2mo ago|0 comments
3. 3.▲
   by T0mSIlver 7mo ago

   Your voxtral.c work was a big motivator for me. I built a macOS menu bar dictation app ( https://github.com/T0mSIlver/localvoxtral ) around Voxtral Realtime, currently using a voxmlx fork with an OpenAI Realtime WebSocke
4. 4.▲
   by T0mSIlver 7mo ago

   Congrats on the results. The streaming aspect is what I find most exciting here. I built a macOS dictation app ( https://github.com/T0mSIlver/localvoxtral ) on top of Voxtral Realtime, and the UX difference between strea
5. 5.▲
   by T0mSIlver 7mo ago

   Some technical context and where this is headed. Why streaming matters for dictation. Whisper and most open-source STT models use bidirectional attention, meaning they need the full audio clip before they can transcribe anything. You get y
6. 6.▲
   Show HN: Localvoxtral – Local real-time dictation on macOS with streaming STT(github.com)

   1 points by T0mSIlver 7mo ago|2 comments