div0-space's picture
Upload README.md with huggingface_hub
cf73191 verified
metadata
license: mit
language:
  - en
  - pl
  - de
  - fr
  - es
  - it
  - pt
  - nl
  - ru
  - zh
  - ja
  - ko
  - ar
  - hi
  - multilingual
tags:
  - whisper
  - speech-to-text
  - mlx
  - quantized
  - q8
  - apple-silicon
library_name: transformers
pipeline_tag: automatic-speech-recognition
base_model: openai/whisper-large-v3

Whisper Large V3 - MLX Q8 Quantized

8-bit quantized version of OpenAI's Whisper Large V3, optimized for Apple Silicon with MLX.

Model Details

Property Value
Original Model openai/whisper-large-v3
Parameters ~1.55B
Quantization INT8 (Q8)
Size ~1.6GB
Decoder Layers 32

Best accuracy among Whisper models. Use large-v3-turbo for faster inference with slightly lower accuracy.

Other Whisper Models

Model Size Speed Link
small ~300MB Fastest LibraxisAI/whisper-small-mlx-q8
medium ~800MB Fast LibraxisAI/whisper-medium-mlx-q8
large-v3 (this) ~1.6GB Slow -
large-v3 q4 ~900MB Medium LibraxisAI/whisper-large-v3-q4
large-v3-turbo ~900MB Fast LibraxisAI/whisper-large-v3-turbo-mlx-q8

Usage

import mlx_whisper

result = mlx_whisper.transcribe(
    "audio.wav",
    path_or_hf_repo="LibraxisAI/whisper-large-v3-mlx-q8"
)
print(result["text"])

Supported Languages

Full multilingual support: English, Polish, German, French, Spanish, Italian, Portuguese, Dutch, Russian, Chinese, Japanese, Korean, Arabic, Hindi, and 90+ additional languages.

Hardware Requirements

  • Apple Silicon Mac (M1/M2/M3/M4)
  • Minimum 16GB RAM recommended

License

MIT - inherited from OpenAI Whisper.


Converted by LibraxisAI using mlx-whisper