RyeAI
/

ekko-v1-tiny

Automatic Speech Recognition

speech-recognition

local-inference

word-timestamps

Eval Results (legacy)

Model card Files Files and versions

emilstabil commited on Oct 9

Commit

09a7edf

·

verified ·

1 Parent(s): 27ff726

Correct Common Voice benchmark naming

Files changed (1) hide show

README.md +3 -3

README.md CHANGED Viewed

@@ -57,8 +57,8 @@ model-index:
   - task:
       type: automatic-speech-recognition
     dataset:
-      name: Common Voice 17 Danish
-      type: mozilla-foundation/common_voice_17_0
       config: da
       split: test
     metrics:
@@ -115,7 +115,7 @@ removal, and whitespace collapse. Lower is better.
 |---|---|---:|
 | CoRal conversation | Spontaneous conversation | 24.15% |
 | CoRal read aloud | Read speech | 11.87% |
-| Common Voice 17 | Crowd-sourced read speech | 8.44% |
 | FLEURS `da_dk` | Read news-style speech | 9.75% |
 | FTSpeech | Parliamentary speech | 7.62% |
 | **Equal-weight mean** | | **12.37%** |

   - task:
       type: automatic-speech-recognition
     dataset:
+      name: Common Voice Danish (CV 25.0 test subset)
+      type: mozilla-foundation/common_voice_27_0
       config: da
       split: test
     metrics:
 |---|---|---:|
 | CoRal conversation | Spontaneous conversation | 24.15% |
 | CoRal read aloud | Read speech | 11.87% |
+| Common Voice Danish (CV 25.0 test subset) | Crowd-sourced read speech | 8.44% |
 | FLEURS `da_dk` | Read news-style speech | 9.75% |
 | FTSpeech | Parliamentary speech | 7.62% |
 | **Equal-weight mean** | | **12.37%** |