Welcome to nuwa-pytorch Discussions!

[lucidrains]

👋 Welcome!

We’re using Discussions as a place to connect with other members of our community. We hope that you:

• Ask questions you’re wondering about.
• Share ideas.
• Engage with other community members.
• Welcome others and are open-minded. Remember that this is a community we
  build together 💪.

To get started, comment below with an introduction of yourself and tell us about what you do with this community.

[YashVardhan-AI]

Oh my god you are the best.

[corranmac]

Missing you in the discord! Looking forward to keep watching your progress on this amazing project though 😊.

Was wondering how you would feel about having discussions similar to our old ones here with some of the other people?
I totally understand if you'd rather keep discussuion here strictly on topic though. All the best.

[fractaldna22]

tell us about what you do with this community.
i bring vqgan as close as it can come to nuwa functions without actually training anything lol. am following this intensely

[lucidrains]

small NUWA primed with "3 and 0" - cherrypicked of course :D

follow along in the training here https://tinyurl.com/yc4jy5zt

[lucidrains]

"2 and 0" - it is starting to learn zero :)

[lucidrains]

also figured out that the NUWA 3d nearby attention wasn't the most efficient, as I could use padding to do a causal - conv-like attention. A bit too late to redo the training for this test run, but have it addressed in the latest version of NUWA 649a916

[lucidrains]

"6 and 5" - Seeing a 5 for a brief second :)

[tjthejuggler]

With what is available here are we able to actually create video from my own text? If so, then after I run the training file, what file do i run with my text in, and where do i put my text? Thanks so much!

[lucidrains]

try following the instructions in the second half of this section https://github.com/lucidrains/nuwa-pytorch#trainers