Loading the leaderboard…
Loading the leaderboard…
Model profile · Multilingual open
First sub-1B audio-LLM on the board — strongest code-switching recall of the verified runs
WER family in %, lower is better · MTR higher is better · RTFx = audio s ÷ wall s, higher is faster · normalizer v0.1
The Hard split holds 1.38 h of rare and unseen code-switched terms. A positive Δ (amber) means the model degrades off the common-term distribution.
Per-topic WER is unavailable for this row — this submission did not include a per-domain breakdown. Platform-verified runs break WER down across the five ViMedCSS domains:
WER is comparable only within a normalizer version · bootstrap 95% CIs on headline metrics.