An expressive speech dataset for LoRA finetuning based on the character Tahm Kench from League of Legends.
Total of 765.94 seconds of audio.
Each entry contains:
Text transcript
Audio file (24kHz, mono)
Single speaker (speaker_id=0)
Formatted for use with the Sesame CSM model.
Credits: Riot Games (original voice lines)
