Search results

Issues about convert Generation Model to Huggingface-like format

Status: Open.

#25 In modelscope/Nexus-Gen;

· cydiachen opened on Sep 12, 2025

Per-Token MSE Loss visualization

Status: Open.

#24 In modelscope/Nexus-Gen;

· jiangzhengguo opened on Aug 24, 2025

Support for Qwen-Image

Status: Open.

#23 In modelscope/Nexus-Gen;

· zheyuanliu233 opened on Aug 8, 2025

one-click setup

Status: Open.

#22 In modelscope/Nexus-Gen;

· GallonDeng opened on Aug 2, 2025

能否根据 diffsynth 的 lowvarm 和 vram_management 技术，降低显存使用，让项目能在 RTX 4090 24GB 单张显卡上运行。

Status: Open.

#21 In modelscope/Nexus-Gen;

· wikeeyang opened on Jul 19, 2025

Release Nexus-Gen Training Dataset on Hugging Face

Status: Open.

#20 In modelscope/Nexus-Gen;

· NielsRogge opened on Jul 16, 2025

How to directly edit an input image instead of generating one first?

Status: Open.

#17 In modelscope/Nexus-Gen;

· wangzrk opened on Jun 22, 2025

Issue about generation with random seed

Status: Open.

#15 In modelscope/Nexus-Gen;

· Warren118 opened on Jun 11, 2025

interleaved输出

Status: Open.

#14 In modelscope/Nexus-Gen;

· XuankeShi opened on May 20, 2025

What is the GPU memory size used for code train?

Status: Open.

#12 In modelscope/Nexus-Gen;

· itisfree opened on May 16, 2025

transformer==4.49和diffsynth==1.1.7冲突，只能下载transformers==4.46，但是会报错'DynamicCache' object has no attribute 'is_compileable'

Status: Open.

#11 In modelscope/Nexus-Gen;

· lzz-dev opened on May 13, 2025

How much GPU memory is needed

Status: Open.

#10 In modelscope/Nexus-Gen;

· ziyaxuanyi opened on May 13, 2025