Model weights and loading instructions are coming soon.

[zhw-zhang]

Thanks for your work, I look forward to trying out your models. I would like to ask when will the model weight be released?

[Lijun-Yu]

Hi, we are going through some internal process, but will make sure to get them out as soon as possible.

[howardgriffin]

Looking for the model too!

[SoftologyPro]

Hi, we are going through some internal process, but will make sure to get them out as soon as possible.

2 months later. Still no models and instructions?

[IIIIIIIllllllllIIIII]

Model release eta?

[JosephZZ]

also follow up

[roadjiang]

Thank you all for your interest and patience. Each time you ask, I consult with our internal lawyers and seek the necessary approvals. We genuinely hope to release the models trained on public benchmarks to the research community. However, there are procedural obstacles we must address, which are beyond our control.

[JosephZZ]

could you give an estimation of :
1, is it still possible to release?
2, how long will it be release (a loose time range would be fine)?

I definitely understand the company issue is out of your control. For me personally as a grad student researcher, I'm just concerned about whether I should still use your model as my base model. If it's possible with a reasonable amount of time, i'll most likely wait for it it. Otherwise I need to consider whether to train one myself or use other alternatives. Speaking of this, how long does it take to train the vqvae base model? and the generation model? How much gpu memory is needed?