whisper-small-eole / README.md
davidmeikle's picture
Create README.md
699d6e9 verified
metadata
language:
  - en
  - zh
  - de
  - es
  - ru
  - ko
  - fr
  - ja
  - pt
  - tr
  - pl
  - ca
  - nl
  - ar
  - sv
  - it
  - id
  - hi
  - fi
  - vi
  - he
  - uk
  - el
  - ms
  - cs
  - ro
  - da
  - hu
  - ta
  - 'no'
  - th
  - ur
  - hr
  - bg
  - lt
  - la
  - mi
  - ml
  - cy
  - sk
  - te
  - fa
  - lv
  - bn
  - sr
  - az
  - sl
  - kn
  - et
  - mk
  - br
  - eu
  - is
  - hy
  - ne
  - mn
  - bs
  - kk
  - sq
  - sw
  - gl
  - mr
  - pa
  - si
  - km
  - sn
  - yo
  - so
  - af
  - oc
  - ka
  - be
  - tg
  - sd
  - gu
  - am
  - yi
  - lo
  - uz
  - fo
  - ht
  - ps
  - tk
  - nn
  - mt
  - sa
  - lb
  - my
  - bo
  - tl
  - mg
  - as
  - tt
  - haw
  - ln
  - ha
  - ba
  - jw
  - su
tags:
  - audio
  - automatic-speech-recognition
  - eole
  - whisper
license: apache-2.0
base_model: openai/whisper-small
pipeline_tag: automatic-speech-recognition

Whisper Small (eole)

This is openai/whisper-small converted to eole format using eole convert --model_dir openai/whisper-small.

No weights were modified — this is a format conversion only.

Model details

Original model openai/whisper-small
Parameters 244M
Encoder layers 12
Decoder layers 12
Hidden size 768
Attention heads 12
Mel bins 80
Vocab size 51,865
License Apache 2.0

Usage

pip install eole[wer]

Transcribe

eole predict \
  -config eval_config.yaml \
  -model_path whisper-small-eole \
  -src audio_files.txt \
  -output transcriptions.txt \
  -language en \
  -task transcribe \
  -gpu_ranks 0

Evaluation

All evaluations use beam size 5.

Benchmark WER
LibriSpeech test-clean 3.30%

Conversion

eole convert --model_dir openai/whisper-small --output whisper-small-eole

Citation

@misc{radford2023robust,
      title={Robust Speech Recognition via Large-Scale Weak Supervision},
      author={Alec Radford and Jong Wook Kim and Tao Xu and Greg Brockman and Christine McLeavey and Ilya Sutskever},
      year={2023},
      eprint={2212.04356},
      archivePrefix={arXiv},
      primaryClass={eess.AS}
}