ctranslate2-4you/whisper-medium-ct2-bfloat16

🗣️ Whisper Medium — CTranslate2 (bfloat16)

This is OpenAI's whisper-medium converted to CTranslate2 format with bfloat16 precision.

Also available in other precisions: float32 · float16

---

📋 Details

Base model	openai/whisper-medium
Format	CTranslate2
Precision	bfloat16
Language	Multilingual
Task	Automatic Speech Recognition

---

⚡ Quick Start

Install the inference library:

pip install whisper-s2t-reborn

Transcribe an audio file:

import whisper_s2t

model = whisper_s2t.load_model(
    model_identifier="medium",
    compute_type="bfloat16",
    device="cuda",
)

result = model.transcribe_with_vad(
    ["audio.wav"],
    lang_codes=["en"],
    tasks=["transcribe"],
    initial_prompts=[None],
    batch_size=1,  # increase this to significantly improve throughput
)

for segment in result[0]:
    print(segment["text"])

Models are auto-downloaded from this repo the first time you run inference. No manual download required.

See the whisper-s2t-reborn repository for the full list of available parameters.

---

📦 All Available CTranslate2 Whisper Models

Every model below is hosted at huggingface.co/ctranslate2-4you and works with whisper-s2t-reborn.

🌍 Standard Whisper (Multilingual)

Model	float32	float16	bfloat16
tiny	Link	Link	Link
base	Link	Link	Link
small	Link	Link	Link
medium	Link	Link	Link
large-v3	Link	Link	Link

🇺🇸 Whisper English-Only

Model	float32	float16	bfloat16
tiny.en	Link	Link	Link
base.en	Link	Link	Link
small.en	Link	Link	Link
medium.en	Link	Link	Link

⚡ Distilled Whisper

Model	float32	float16	bfloat16
distil-small.en	Link	Link	Link
distil-medium.en	Link	Link	Link
distil-large-v3	Link	Link	Link

🚀 Whisper Large-v3 Turbo

Model	float32	float16	bfloat16
large-v3-turbo	Link	Link	Link

---

🔗 Links

• 📦 Inference library — whisper-s2t-reborn
• 🏗️ CTranslate2 — github.com/OpenNMT/CTranslate2
• 🧠 Original model — openai/whisper-medium