Search results

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.

Status: Open.

#59 In Tencent-Hunyuan/HunyuanVideo-1.5;

· UserChen666 opened on Jul 30, 2026

DreamX-World: A General-Purpose Interactive World Model

Status: Open.

#58 In Tencent-Hunyuan/HunyuanVideo-1.5;

· shuidongliu opened on Jun 18, 2026

Learnable type embedding implementation

Status: Open.

#57 In Tencent-Hunyuan/HunyuanVideo-1.5;

· adithyaiyer1999 opened on May 26, 2026

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

Status: Open.

#56 In Tencent-Hunyuan/HunyuanVideo-1.5;

· wzic opened on Apr 16, 2026

there is not data parallel in training code?

Status: Open.

#51 In Tencent-Hunyuan/HunyuanVideo-1.5;

· JoJoistheBestOne opened on Mar 10, 2026

A

Status: Open.

#50 In Tencent-Hunyuan/HunyuanVideo-1.5;

· elviolaburras-coder opened on Mar 9, 2026

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？

Status: Open.

#42 In Tencent-Hunyuan/HunyuanVideo-1.5;

· vibe-viscot opened on Dec 18, 2025

有计划发布DPO+Lora微调的脚本吗？

Status: Open.

#41 In Tencent-Hunyuan/HunyuanVideo-1.5;

· vibe-viscot opened on Dec 18, 2025

Where to find LoRAs? / Lora sharing

Status: Open.

#37 In Tencent-Hunyuan/HunyuanVideo-1.5;

· kabachuha opened on Dec 12, 2025

v2v or video editing support

Status: Open.

#36 In Tencent-Hunyuan/HunyuanVideo-1.5;

· hengrui0516 opened on Dec 10, 2025

Do you have any plans to release the post-training code, such as DPO/GRPO?

Status: Open.

#34 In Tencent-Hunyuan/HunyuanVideo-1.5;

· liming-ai opened on Dec 8, 2025

Vertical videos have worse quality than the horizonal ones?

Status: Open.

#32 In Tencent-Hunyuan/HunyuanVideo-1.5;

· kabachuha opened on Dec 6, 2025