Repository Issue Activity (beta)

#59 按照readme中的demo进行推理时报错 ValueError: Cannot generate a cpu tensor from a generator of type cuda.

#58 DreamX-World: A General-Purpose Interactive World Model

#57 Learnable type embedding implementation

#56 Can we directly use Qwen3.5 as the text encoder ?

#46 Any plans for reference-to-video similar to Hunyuan Custom?

#36 v2v or video editing support

#54 Questions about gradient checkpointing

#52 Test

#51 there is not data parallel in training code?

#50 A

#47 multi-node finetune error when saving checkpoint

#48 [Usage] Wrong output Inference with source code

#45 fixed

#44 BUG in apply_fsdp

#43 Training GPU Memory

#42 我想要一个既能在我的任务上表现良好，又能少步推理（步数蒸馏）的权重。我该怎么做？

#41 有计划发布DPO+Lora微调的脚本吗？

#39 Wan 2.2 version in comparison

#38 About training i2v

#33 H800 8卡复现需要41s？

#37 Where to find LoRAs? / Lora sharing

#31 Inference Time of i2v model

#12 Users, please share your experiences - what GPUs are "enough" for meaningful usage?

#27 Install fails in ComfyUI

#30 Could you please let us know when training code will be released?