Automatic Speech Recognition on CommonVoice 17.0 (test)
2.8Word Error Rate (WER)granite-speech-4.0-1b
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| granite-speech-4.0-1bDecoding Regime=High Accuracy SSD, tau_CTC=0.7, tau_SLM=0.2, Target Language=Portuguese2026.03 | 2.8 | 1,066 | 5 | 69 | |
| granite-speech-4.0-1bDecoding Regime=Full AR, Target Language=Portuguese2026.03 | 2.82 | 866 | — | — | |
| granite-speech-4.0-1bDecoding Regime=High RTFx SSD, tau_CTC=3.0, tau_SLM=0.1, Target Language=Portuguese2026.03 | 3.83 | 2,344 | — | — | |
| granite-speech-4.0-1bDecoding Regime=High Accuracy SSD, tau_CTC=0.7, tau_SLM=0.2, Target Language=Spanish2026.03 | 3.98 | 890 | 1 | 59 | |
| granite-speech-4.0-1bDecoding Regime=Full AR, Target Language=Spanish2026.03 | 4.04 | 752 | — | — | |
| granite-speech-4.0-1bDecoding Regime=High Accuracy SSD, tau_CTC=0.7, tau_SLM=0.2, Target Language=German2026.03 | 4.71 | 998 | 0 | 59 | |
| granite-speech-4.0-1bDecoding Regime=Full AR, Target Language=German2026.03 | 4.79 | 872 | — | — | |
| granite-speech-4.0-1bDecoding Regime=High Accuracy SSD, tau_CTC=0.7, tau_SLM=0.2, Target Language=Average2026.03 | 5.03 | 916 | 7 | 53 | |
| granite-speech-4.0-1bDecoding Regime=Full AR, Target Language=Average2026.03 | 5.06 | 824 | — | — | |
| granite-speech-4.0-1bDecoding Regime=High RTFx SSD, tau_CTC=3.0, tau_SLM=0.1, Target Language=Spanish2026.03 | 5.42 | 2,526 | — | — | |
| granite-speech-4.0-1bDecoding Regime=High RTFx SSD, tau_CTC=3.0, tau_SLM=0.1, Target Language=German2026.03 | 6.34 | 2,309 | — | — | |
| granite-speech-4.0-1bDecoding Regime=Full AR, Target Language=English2026.03 | 6.52 | 954 | — | — | |
| granite-speech-4.0-1bDecoding Regime=High Accuracy SSD, tau_CTC=0.7, tau_SLM=0.2, Target Language=English2026.03 | 6.56 | 716 | 28 | 31 | |
| granite-speech-4.0-1bDecoding Regime=High Accuracy SSD, tau_CTC=0.7, tau_SLM=0.2, Target Language=French2026.03 | 7.1 | 910 | 0 | 46 | |
| granite-speech-4.0-1bDecoding Regime=High RTFx SSD, tau_CTC=3.0, tau_SLM=0.1, Target Language=Average2026.03 | 7.11 | 2,393 | — | — | |
| granite-speech-4.0-1bDecoding Regime=Full AR, Target Language=French2026.03 | 7.15 | 674 | — | — | |
| granite-speech-4.0-1bDecoding Regime=High RTFx SSD, tau_CTC=3.0, tau_SLM=0.1, Target Language=English2026.03 | 9.3 | 2,446 | — | — | |
| granite-speech-4.0-1bDecoding Regime=High RTFx SSD, tau_CTC=3.0, tau_SLM=0.1, Target Language=French2026.03 | 10.69 | 2,341 | — | — |