paprika-whisper-lt-mlx

MLX conversion of kristijonas/paprika-whisper-lt, a Lithuanian fine-tune of openai/whisper-large-v3-turbo (via svogunas/whisper-large-v3-turbo-lt).

This repo fixes the feature extractor filename from the nonstandard processor_config.json to the standard preprocessor_config.json, so transformers.WhisperProcessor.from_pretrained() works without patching.

Use

import mlx_whisper

result = mlx_whisper.transcribe("audio.wav", path_or_hf_repo="internetworks/paprika-whisper-lt-mlx")
print(result["text"])
Downloads last month
-
Safetensors
Model size
0.8B params
Tensor type
I64
·
F16
·
MLX
Hardware compatibility
Log In to add your hardware

Quantized

Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for internetworks/paprika-whisper-lt-mlx