Whisper large-v3 model for CTranslate2

This repository contains the conversion of Whisper large-v3 to the CTranslate2 model format.

This model can be used in CTranslate2 or projects based on CTranslate2 such as faster-whisper.

Example

from faster_whisper import WhisperModel

model = WhisperModel("flyingleafe/faster-whisper-large-v3")

segments, info = model.transcribe("audio.mp3")
for segment in segments:
    print("[%.2fs -> %.2fs] %s" % (segment.start, segment.end, segment.text))

Conversion details

The original model was converted the following way:

# use Transformers convertation to HF format
python transformers/src/transformers/models/whisper/convert_openai_to_hf.py \
    --checkpoint_path large-v3 --pytorch_dump_folder_path ./whisper-large-v3 --convert_tokenizer True

# ... some manual convertation to get `tokenizer.json` via `WhisperTokenizerFast` class ...

ct2-transformers-converter --model ./whisper-large-v3 --output_dir faster-whisper-large-v2 \
    --copy_files tokenizer.json preprocessor_config.json --quantization float16

Note that the model weights are saved in FP16. This type can be changed when the model is loaded using the compute_type option in CTranslate2.

Downloads last month
419
Inference Examples
Inference API (serverless) does not yet support ctranslate2 models for this pipeline type.