Downloads · 30 days
21
20% of all-time downloads
paprika-whisper-lt-v3-mlx
internetworks/paprika-whisper-lt-v3-mlx
paprika-whisper-lt-v3-mlx is a automatic speech recognition model from internetworks. Use it when you need speech turned into text. It is set up for mlx. The card lists the license as cc-by-4.0.
MLX conversion of kristijonas/paprika-whisper-lt-v3, a Lithuanian fine-tune of openai/whisper-large-v3-turbo (third generation, trained on LIEPA-3, warm-started from kristijonas/paprika-whisper-lt).
All-time downloads
104
Public
Parameters
807M
1.6 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1.6 GB · 100%
Parameter types
How the weights are stored.
F16807M · 100%
Base models
Model card
From the Hugging Face model README
paprika-whisper-lt-v3-mlx
MLX conversion of kristijonas/paprika-whisper-lt-v3, a Lithuanian fine-tune of openai/whisper-large-v3-turbo (third generation, trained on LIEPA-3, warm-started from kristijonas/paprika-whisper-lt).
Usage notes from the base model card apply: use long-form decoding, do not use chunk_length_s — the chunked path silently drops text at seams and invents text on non-speech. See the base card for details.
Use
import mlx_whisper
result = mlx_whisper.transcribe("audio.wav", path_or_hf_repo="internetworks/paprika-whisper-lt-v3-mlx")
print(result["text"])
- Task
- Automatic Speech Recognition
- Library
- mlx
- Type
- whisper
- License
- cc-by-4.0
- Languages
- lt
- Created
- Aug 30, 2026
- Updated
- Aug 30, 2026