Will you release the text-to-token model?

[hbwu-ntu]

No description provided.

[hbwu-ntu]

And by the way, does the text-to-token model support batch inference?

[sixsixcoder]

Please check the technical report. The current version of the model is in the warehouse and there is no major upgrade plan for the time being.

[hbwu-ntu]

Hi thanks a lot for the response. I traced the entire GitHub repo and only found you released these three models: speech tokenizer, chat model, speech decoder. I may miss some codes in the repo about the text-to-token model.

May I ask if the text-to-token model (1.5B) mentioned in Section 2.2 of https://arxiv.org/abs/2411.17607 will be released? If so, could you direct me to corresponding scripts or checkpoints to use that model?

Thank you so much.

[ChenYuhan00]

Have you found the text2token model? I haven't found it either.