Reson8 open-sources Rumble-2-Flash 0.5B speech model
Start here
Reson8 has released Rumble-2-Flash, a 0.5B-parameter speech recognition model covering more than 40 languages, under the Apache 2.0 license. The weights are published on Hugging Face at Reson8/Rumble-2-Flash, and the model is also offered through the company's API. It includes automatic language detection and native support in Transformers and vLLM. On the Open ASR Leaderboard multilingual track, the model averages 6.96% error, which Reson8 says is ahead of Whisper large-v3 and Meta's omniASR-LLM 7B. On AppTek's 302-hour call-centre benchmark it reaches 7.86% word error rate, described as on par with NVIDIA's 30B Nemotron 3.5 Transcribe. The model was distilled from the larger Rumble-2 using on-policy distillation with GOLD, which aligns the differing tokenizers of teacher and student.
Generated from reports · updated 2 h ago
Report timeline
Follow the reports to see the event from different sides.
- Hugging Face BlogPicksReson8 open-sources Rumble-2-Flash, a 0.5B speech recognition model for 40+ languages
Reson8 open-sourced Rumble-2-Flash under Apache 2.0, a 0.5B-parameter speech recognition model covering 40+ languages with automatic language detection and native support in Transformers and vLLM. It averages 6.96% error on the Open ASR Leaderboard multilingual track, ahead of Whisper large-v3 and Meta's omniASR-LLM 7B, and reaches 7.86% WER on AppTek's 302-hour call-centre benchmark, on par with NVIDIA's Nemotron 3.5 Transcribe (30B). The model was distilled from the larger Rumble-2 using on-policy distillation with GOLD, which aligns differing tokenizers between teacher and student.
Heat for this event
Not enough continuous observations yet to draw a trend.