Faster Whisper transcription with CTranslate2

This repository demonstrates how to implement the Whisper transcription using CTranslate2, which is a fast inference engine for Transformer models.

This implementation is up to 4 times faster than openai/whisper for the same accuracy while using less memory. The efficiency can be further improved with 8-bit quantization on both CPU and GPU.

Benchmark

For reference, here's the time and memory usage that are required to transcribe 13 minutes of audio using different implementations:

Large-v2 model on GPU

Implementation	Precision	Beam size	Time	Max. GPU memory	Max. CPU memory
openai/whisper	fp16	5	4m30s	11325MB	9439MB
faster-whisper	fp16	5	54s	4755MB	3244MB
faster-whisper	int8	5	59s	3091MB	3117MB

Executed with CUDA 11.7.1 on a NVIDIA Tesla V100S.

Small model on CPU

Implementation	Precision	Beam size	Time	Max. memory
openai/whisper	fp32	5	10m31s	3101MB
whisper.cpp	fp32	5	17m42s	1581MB
whisper.cpp	fp16	5	12m39s	873MB
faster-whisper	fp32	5	2m44s	1675MB
faster-whisper	int8	5	2m04s	995MB

Executed with 8 threads on a Intel(R) Xeon(R) Gold 6226R.

Installation

pip install faster-whisper

The model conversion script requires the modules transformers and torch which can be installed with the [conversion] extra requirement:

pip install faster-whisper[conversion]

Other installation methods:

# Install the master branch:
pip install --force-reinstall "faster-whisper @ https://github.com/guillaumekln/faster-whisper/archive/refs/heads/master.tar.gz"

# Install a specific commit:
pip install --force-reinstall "faster-whisper @ https://github.com/guillaumekln/faster-whisper/archive/a4f1cc8f11433e454c3934442b5e1a4ed5e865c3.tar.gz"

# Install for development:
git clone https://github.com/guillaumekln/faster-whisper.git
pip install -e faster-whisper/

GPU support

GPU execution requires the NVIDIA libraries cuBLAS 11.x and cuDNN 8.x to be installed on the system. Please refer to the CTranslate2 documentation.

Usage

Model conversion

A Whisper model should be first converted into the CTranslate2 format. We provide a script to download and convert models from the Hugging Face model repository.

For example the command below converts the "large-v2" Whisper model and saves the weights in FP16:

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

If the option --copy_files tokenizer.json is not used, the tokenizer configuration is automatically downloaded when the model is loaded later.

Models can also be converted from the code. See the conversion API.

Transcription

from faster_whisper import WhisperModel

model_path = "whisper-large-v2-ct2/"

# Run on GPU with FP16
model = WhisperModel(model_path, device="cuda", compute_type="float16")

# or run on GPU with INT8
# model = WhisperModel(model_path, device="cuda", compute_type="int8_float16")
# or run on CPU with INT8
# model = WhisperModel(model_path, device="cpu", compute_type="int8")

segments, info = model.transcribe("audio.mp3", beam_size=5)

print("Detected language '%s' with probability %f" % (info.language, info.language_probability))

for segment in segments:
    print("[%.2fs -> %.2fs] %s" % (segment.start, segment.end, segment.text))

Word-level timestamps

segments, _ = model.transcribe("audio.mp3", word_timestamps=True)

for segment in segments:
    for word in segment.words:
        print("[%.2fs -> %.2fs] %s" % (word.start, word.end, word.word))

See more model and transcription options in the WhisperModel class implementation.

Comparing performance against other implementations

If you are comparing the performance against other Whisper implementations, you should make sure to run the comparison with similar settings. In particular:

Verify that the same transcription options are used, especially the same beam size. For example in openai/whisper, model.transcribe uses a default beam size of 1 but here we use a default beam size of 5.
When running on CPU, make sure to set the same number of threads. Many frameworks will read the environment variable OMP_NUM_THREADS, which can be set when running your script:

OMP_NUM_THREADS=4 python3 my_script.py

haloha123/faster-whisper