Modelpublic

CheshireCC/faster-whisper-large-v3-float32

sourceHugging Facemitupdated 3y agoView on Hugging Face

15likes73downloads

Model Card

openai-Whisper model large-v3, CTranslate2 format

convert from flax model files

download flax_model.msgpack of float32 and other config files from https://huggingface.co/openai/whisper-large-v3

from ctranslate2.converters import TransformersConverter as cvter
model_name_or_path = <your folder with model files>
output_dir = <target folder to save model files>
cvter_01 = cvter(model_name_or_path=model_name_or_path)
cvter_01.convert(output_dir=output_dir, quantization="float32", force=True)