Huaying Yuan

Thank you for the guidance. However, I'm curious about the '80% speedup' mentioned in the readme. Could you clarify how to achieve this performance improvement?

I share the same concern. I can't locate the code for the '225 * 90 * 90' setting.

> Sora technical report claims that 1920 * 1080 resolution (1080P) within any generation, but also able to generate 1 minute of video [not sure whether it is a one-time...

Another question is, why flashattn and layernorm_kernel is forbidden during sampling? (https://github.com/NUS-HPC-AI-Lab/OpenDiT/blob/c15d82b738d0efb7f8f9e79c2f5277cbb417c8e2/sample.py#L70) Looking forward to you reply. Thanks in advance.

> Hi! We are still working on adapting Fastseq to the Latte model and will release it in the future. You can manually set the `enable_flashattn` to `True` when sampling....

> For example, sequence parallel size=2, and we have 2 gpus, the `qkv.weight` is sliced into 2 slices; On rank0, the first half of qkv.weight has grad, the second part...

> not for the current time because our latte model does not exacty match the origin one Could you release your pretrained weight of latte?

李老师是高中班主任型授课。讲课比较清晰，会帮助同学复习上一节课的知识，课堂会经常提问，目的在于让每一个同学都能掌握知识。给分客观

Sorry, I dont find the solution either ------------------ 原始邮件 ------------------ 发件人: "kssteven418/LTP" ***@***.***>; 发送时间: 2022年7月11日(星期一) 上午7:25 ***@***.***>; ***@***.******@***.***>; 主题: Re: [kssteven418/LTP] Where to get the pretrained model with max-seq-length over 512? (Issue #7) I...

OK, thank you for your reply!! ------------------ 原始邮件 ------------------ 发件人: "Sehoon ***@***.***>; 发送时间: 2022年7月12日(星期二) 凌晨1:20 收件人: ***@***.***>; 抄送: ***@***.***>; ***@***.***>; 主题: Re: [kssteven418/LTP] Where to get the pretrained model with max-seq-length over...