DatasetWaxal

← All examples Model summary →

Modelfacebook/omniASR-CTC-1BOpen18facebook/omniASR-CTC-300MOpen18facebook/omniASR-CTC-3BOpen18facebook/omniASR-CTC-7BOpen17facebook/omniASR-LLM-1BOpen20facebook/omniASR-LLM-300MOpen21facebook/omniASR-LLM-3BOpen20facebook/omniASR-LLM-7BOpen16aadel4/omniASR-CTC-1B-v2Open18aadel4/omniASR-CTC-300M-v2Open17openai/whisper-large-v3Open33openai/whisper-smallOpen148

What did the model get wrong?

facebook/omniASR-LLM-3BOpen   Example 99 of 100  ·  sample index 98

🎧 Listen — original audio
✅ What was actually said (Shona reference)
Nzvimboinotengeswahembedzakasiyana-siyanamunemukomanaarimo,mukamanauyuakapfekahembeineruvararutemayakanyorwanemavaramachenanechekumberi.Kunehembedzikutengeswadzimwezvineruvararutsvukudzimwedzineruvararuchenanemavaramatema.
🤖 What this model heard
inotengeswahembedzakasiyanasiyanamunemukomanaarimomukomanauyowakapfekahembeineruvararutemayakanyorwanamavaramachenanechekumberikunehembedzirukutengeswadzimwedzineruvararutsukudzimwedzineruvararuchenanamavaramatema
Correct
Wrong (different word)
Missed (skipped)
Extra (added)
11
Correct
14
Wrong
1
Missed
5
Extra
42%
Words right
0.885
WER
0.113
CER
20.14s
Duration
1.2s
Latency