CoolFace
Datasetpublic

Trelis/eval-whisper-small-english-eka-hard-20260408-1925

Evaluation Results: whisper-small-english Evaluation results from Whisper model evaluation. Summary Model WER CER leduckhai/MultiMed-ST/asr/whisper-small-english 49.12% 35.11% Source Data Evaluation Dataset: Trelis/eka-hard Model Evaluated: leduckhai/MultiMed-ST/asr/whisper-small-english Columns Column Description audio Audio sample (if available from source dataset) reference Ground truth transcription… See the full description on the dataset page: https://huggingface.co/datasets/Trelis/eval-whisper-small-english-eka-hard-20260408-1925.

sourceHugging Faceupdated 6mo agoView on Hugging Face
0likes20downloads
Dataset Card

Evaluation Results: whisper-small-english

Evaluation results from Whisper model evaluation.

Summary

ModelWERCER
leduckhai/MultiMed-ST/asr/whisper-small-english49.12%35.11%

Source Data

Columns

ColumnDescription
audioAudio sample (if available from source dataset)
referenceGround truth transcription
predictionModel prediction
werWord Error Rate for this sample
cerCharacter Error Rate for this sample

| entities | Entity annotations from source dataset | | entity_cer | Per-sample entity CER (-1.0 if no entities) |

Entity CER

Overall Entity CER: 44.98%

CategoryCER
anatomy35.29%
biomarker44.44%
condition33.61%
drug47.98%
procedure63.64%

Generated by [Trelis Studio](https://studio.trelis.com)