Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
LLaSA: Scaling Train-time and Inference-time Compute for LLaMA-based Speech Synthesis
| Date | Stars |
|---|---|
| 2026-07-24 | 661 |
| 2026-07-25 | 661 |
| 2026-07-28 | 661 |
| 2026-07-30 | 661 |
| 2026-08-06 | 661 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
[](https://arxiv.org/abs/2502.04128) ## Directly used on Hugging Face **Codec**: [xcodec2](https://huggingface.co/HKUST-Audio/xcodec2) **TTS inference: Llasa-collections**: [Llasa-collections](https://huggingface.co/collections/HKUSTAudio/llasa-679b87dbd06ac556cc0e0f44) **Update (2025-06-09):** Add Llasa finetune instruction. Awesome work based on Llasa - [LLaSa GRPO tuning](https://github.com/channel-io/ch-tts-llasa-rl-grpo) : Special thanks to [Seungyoun Shin](https://github.com/SeungyounShin) from [Channel Corp.](https://channel.io/en) for sharing the Llasa GRPO RL-tuning script based on Verl. You can try the finetuning results here: - [LLaSA 1B Multi-Speakers (Genshin-zh-en-ja-ko)](https://huggingface.co/spaces/HKUST-Audio/Llasa-1B-multi-speakers-genshin-zh-en-ja-ko) - [LLaSA 1B Finetuned for Two Speakers](https://huggingface.co/spaces/HKUST-Audio/Llasa-1B-finetuned-for-two-speakers) **Update (2025-02-07):** Our paper has been released! Llasa 1b Multilingual version released! ## Training ```bash torchrun --nproc_per_node=8 train_tts.py config.json ``` or ```bash sbatch run_slurm.sh ``` ## Data You can download tokenized open-source speech data [here](https://huggingface.co/datasets/HKUST-Audio/Llasa_opensource_speech_data_160k_hours_tokenized/tree/main). This includes LibriHeavy, Emilia (in both Chinese and English), and WenetSpeech4TTS, totaling approximately 160,000 hours of open-source data. Our models are trained on 250,000 hours of speech data. Of this, 160,000 hours come from the open-source datasets mentioned above, while the remaining 90,000 hours are from internal datasets, which are not yet available for open-source release. ## Data instruction [Text_sequence](https://github.com/zhenye234/LLaSA_training/blob/5ffcddee243f0aa594ebfc089f4327a24f7cac6f/train_tts.py#L111) is encoded by the text tokenizer from Llama, for example, [Llama-3.2-1B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct) [Speech_sequence](https://github.com/zhenye234/LLaSA_training/blob/5ffcddee243f0aa594ebfc089f4327a24f7cac6f/train_tts.py#L112) is extrated through [X-codec2](https://github.com/zhenye234/X-Codec-2.0) We change the value of speech tokens by adding len(text tokenizer) +8 [special tokens](https://github.com/zhenye234/LLaSA_training/blob/1d65cf3e34c0d5b508404d67ff41b3b6fb1ecab7/train_tts.py#L67) thereby forming a unified tokenizer that encompasses both speech and text.
Excerpt of 2,521 characters
Read on GitHub
27
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:da54e61461d09fe7, topic:llm, topic:llama
matched fp:da54e61461d09fe7, topic:text-to-speech, desc:speech synthesis