Downloads · 30 days
0
mktub-whisper-codeswitch-ggml
eliasbatshon/mktub-whisper-codeswitch-ggml
mktub-whisper-codeswitch-ggml is a automatic speech recognition model from eliasbatshon. Use it when you need speech turned into text. It is set up for whisper.cpp. The card lists the license as apache-2.0.
A GGML build for whisper.cpp of Seif-Eldeen-Sameh/whisper-medium-arabic-codeswitched, quantized to q80, for on-device Egyptian Arabic + English code-switch speech-to-text (Arabic stays Arabic, inline English stays in…
Access
Public
Updated Jul 8, 2026
Repo size
823 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin823 MB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- whisper.cpp
- License
- apache-2.0
- Access
- Public
- Created
- Jul 8, 2026
- Updated
- Jul 8, 2026
- SHA
- ec432297
Base models
Model card
From the Hugging Face model README
MKTUB — Whisper Medium Arabic Code-Switch (GGML, q8_0)
A GGML build for whisper.cpp of
Seif-Eldeen-Sameh/whisper-medium-arabic-codeswitched,
quantized to q8_0, for on-device Egyptian Arabic + English code-switch speech-to-text
(Arabic stays Arabic, inline English stays in Latin letters).
- File:
ggml-codeswitch-medium-q8_0.bin(~785 MB) - Base model:
Seif-Eldeen-Sameh/whisper-medium-arabic-codeswitched(Apache-2.0) - Runtime: whisper.cpp (Metal / CUDA / CPU)
- Used by: the MKTUB captions plugin for Adobe Premiere Pro (downloaded on first launch).
License & attribution
Apache-2.0, inherited from the base model. All credit for the underlying weights goes to the original author (Seif-Eldeen-Sameh); this repo only redistributes a format-converted, quantized copy for whisper.cpp. See the base model card for full details.
- Task
- Automatic Speech Recognition
- Library
- whisper.cpp
- License
- apache-2.0
- Languages
- ar, en
- Created
- Jul 8, 2026
- Updated
- Jul 8, 2026