All issues

Search results

麻烦问下，qwen image 2.1 如何进行图像编辑的训练？没看到相关数据集

Status: Open.

#1709 In modelscope/DiffSynth-Studio;

· whyPeanutbutter opened on Sep 24, 2026

WanToDance: hardcoded device='cuda' in music encoder construction crashes model loading on Ascend NPU

Status: Open.

#1707 In modelscope/DiffSynth-Studio;

· li-lizhe opened on Sep 23, 2026

qwen image 2.1 lora/full 训出来的模型 出图全是噪声

Status: Open.

#1706 In modelscope/DiffSynth-Studio;

· sjf18 opened on Sep 22, 2026

[Bug] Qwen-Image-2.1 validate_full 未加载训练权重，实际验证的是基座模型

Status: Open.

#1702 In modelscope/DiffSynth-Studio;

· qiu-pinggaizi opened on Sep 21, 2026

Qwen-Image-2.1 2K输出，多参考图 <=5, 推理时OOM

Status: Open.

#1701 In modelscope/DiffSynth-Studio;

· gobigrassland opened on Sep 21, 2026

怎么能使用minimax h3 turbo lora、我测试了最新的 1.2版本、输出视频都是噪声

Status: Open.

#1691 In modelscope/DiffSynth-Studio;

· 01-111162-173 opened on Sep 14, 2026

Silently slower attention kernel when training MiniMax-H3 LoRA on B200

Status: Open.

#1687 In modelscope/DiffSynth-Studio;

· TarzanZhao opened on Sep 11, 2026

FLUX LoRA Fine-Tuning: Border Outpainting and Pixel-Level Misalignment Issues in Image Editing

Status: Open.

#1685 In modelscope/DiffSynth-Studio;

· lllllialois opened on Sep 11, 2026

oom on minimax h3

Status: Open.

#1682 In modelscope/DiffSynth-Studio;

· perryyang-hue opened on Sep 10, 2026

Partial download surfaces as "Cannot detect the model type" rather than a download error

Status: Open.

#1668 In modelscope/DiffSynth-Studio;

· rakhimovv opened on Sep 2, 2026

LingBot-Video 30B LoRA training OOM on H100

Status: Open.

#1664 In modelscope/DiffSynth-Studio;

· Zanghu-ze opened on Sep 2, 2026

train_scale_cfg

Status: Open.

#1663 In modelscope/DiffSynth-Studio;

· smith-fugui opened on Sep 2, 2026