Lahgtna Levantine TTS — Synthetic Levantine Arabic & Code-Switching

🎙️ Lahgtna Levantine TTS Synthetic Levantine Arabic + English Code-Switching speech dataset. Generated using Lahgtna-OmniVoice, a fine-tuned zero-shot TTS model for Levantine Arabic dialect. 📊 Dataset Statistics Metric Value Total utterances 50,000 Total speakers 10 (5 male, 5 female) Pure Levantine Arabic 44,154 utterances Code-switching (AR+EN) 5,846 utterances Sampling rate 24,000 Hz Estimated total duration ~66.8 hours… See the full description on the dataset page:

Source: Hugging Face Hub (mohammedaly22/lahgtna-levantine-tts). Metadata imported from the dataset’s Hub tags.