Rohan Paul

@rohanpaul_ai

The interesting experiment is basically this: find where the 3.54 WER breaks. Run HojoAI/Hojo-ASR-Multi-V1 against the ASR stack already being used, publish the audio conditions, hardware, settings, ground truth, and raw outputs, then show exactly where each model fails.
打开原帖#511482
  1. Industry

    Rohan Paul: An open multilingual ASR model just posted 3.54 average WER across 5…
  2. Industry

    Arthur Mensch: Should Europe leave frontier AI to the Americans and the Chinese? The…
  3. Industry

    Charles Frye: our clocks are finally free from the stars only took a few gigasecond…