Amsterdam 创业公司 Reson8 发布语音转文字模型 Rumble-2-Flash,仅 0.5B 参数、1GB 大小,支持 45 种语言,以 Apache 2.0 协议开放权重,号称胜过体积 20 倍的模型。模型针对重口音、重叠人声、糟糕电话线路和街道噪音等音频场景,可在手机本地离线转录通话和语音备忘录,无需联网。
Amsterdam startup Reson8 has just released Rumble-2-Flash, a 1GB speech-to-text model that runs on a phone and beats models 20x its size.
Just 0.5B parameters, covers 45 languages and is free under Apache 2.0, built for the audio that breaks most transcription tools: thick accents, overlapping voices, bad phone lines and street noise.
Running on the phone itself, it would transcribe calls and voice notes with no internet connection.
Speech recognition is one of the places where very small open models are closing the gap with frontier systems fastest.
在 X 查看被引用的帖子Rumble-2-flash is open weights you can get it in all your favorite places. It’s better than models 20x it’s size small enough to run on your phone HF link below
来源:Rohan Paul · x.com