Llasa - a HKUSTAudio Collection

HKUSTAudio 's Collections

Llasa

YuE

Llasa

updated May 11

TTS foundation model compatible with Llama framework (160k hours tokenized speech data released)

HKUSTAudio/xcodec2

Audio-to-Audio • 0.8B • Updated Feb 23 • 164k • 78
HKUSTAudio/Llasa-1B

Text-to-Speech • 1B • Updated May 10 • 8.98k • 98
HKUSTAudio/Llasa-3B

Text-to-Speech • 4B • Updated May 10 • 3.33k • • 511
HKUSTAudio/Llasa-8B

Text-to-Speech • 9B • Updated Mar 9 • 5.12k • 94
HKUSTAudio/Llasa-1B-Multilingual

Text-to-Speech • 2B • Updated Mar 5 • 62.4k • 38
HKUSTAudio/Llasa_opensource_speech_data_160k_hours_tokenized

Updated Feb 13 • 913 • 29
Llasa: Scaling Train-Time and Inference-Time Compute for Llama-based Speech Synthesis

Paper • 2502.04128 • Published Feb 6 • 26
HKUSTAudio/Llasa-3B-Preserve-TextChat

Text-to-Speech • 4B • Updated Feb 13 • 3 • 2
HKUSTAudio/Llasa-1B-Preserve-TextChat

Text-to-Speech • 2B • Updated Feb 13 • 511 • 2
HKUSTAudio/Llasa-1B-multi-speakers-genshin-zh-en-ja-ko

Text-to-Speech • 2B • Updated Feb 13 • 1k • 4
HKUSTAudio/Llasa-1B-two-speakers-kore-puck

Text-to-Speech • 2B • Updated Feb 13 • 8 • 4