Latte的实时微信讨论组

Vchitect / Latte

Latte: Latent Diffusion Transformer for Video Generation.

Apache License 2.0

1.45k stars 147 forks source link

Latte的实时微信讨论组 #28

Open akebest opened 4 months ago

akebest commented 4 months ago

我刚试跑了一下t2v demo，效果还是有一定的差距的在想几个事情， 1、是不是加大训练量能够很大改善，极大缩小跟sora 的差距 2、素材怎么搞得更好质量，比如类似dalle e 视频生成文字。

有没有实时的讨论组啊，比如微信群啥的，要不我建个

latte

maxin-cn commented 4 months ago

我刚试跑了一下t2v demo，效果还是有一定的差距的在想几个事情， 1、是不是加大训练量能够很大改善，极大缩小跟sora 的差距 2、素材怎么搞得更好质量，比如类似dalle e 视频生成文字。

有没有实时的讨论组啊，比如微信群啥的，要不我建个

Thank you for your interest and suggestions. I created a discord channel for discussion. Look forward to communicating with you to make the performance of LatteT2V model better ~

uniqzheng commented 4 months ago

二维码过期了，求加入讨论群

pokameng commented 3 months ago

二维码过期了，求加入讨论群

tianyma commented 3 months ago

discord channel cannot be used and the QR code is out of date.

maxin-cn commented 3 months ago

discord channel cannot be used and the QR code is out of date.

Use this: https://discord.gg/RguYqhVU92

tianyma commented 3 months ago

Thank you for your quick reply! Can you respond to my discord question?