Search results

How to run original Wan-2.1-T2V-14b 720p on a single RTX 5090

Status: Open.

#123 In thu-ml/TurboDiffusion;

· Jordanyang opened on Mar 9, 2026

net.load_state_dict(state_dict, assign=True) 失败

Status: Open.

#122 In thu-ml/TurboDiffusion;

· JackBlue302 opened on Mar 9, 2026

Generalization of the merging operation

Status: Open.

#121 In thu-ml/TurboDiffusion;

· toyot-li opened on Mar 9, 2026

hello,请教下关于模型兼容性的问题

Status: Open.

#120 In thu-ml/TurboDiffusion;

· godwenbin opened on Mar 5, 2026

Illegal Memory Access on 5090

Status: Open.

#119 In thu-ml/TurboDiffusion;

· loaydatrain opened on Mar 5, 2026

QA What benchmark/dataset is used for evaluations?

Status: Open.

#118 In thu-ml/TurboDiffusion;

· mikasa955 opened on Feb 25, 2026

sla是全参微调吗，还是只训练SparseLinearAttention？

Status: Open.

#117 In thu-ml/TurboDiffusion;

· wzr0108 opened on Feb 25, 2026

Question about different patch embedding implementation compared to official Wan-Video

Status: Open.

#114 In thu-ml/TurboDiffusion;

· ChenJiongxu355 opened on Feb 14, 2026

How is the performance if we train with sCM loss only?

Status: Open.

#111 In thu-ml/TurboDiffusion;

· meng-ustc opened on Feb 1, 2026

Sparse Linear Layer weights are 0 for 14B 480p model

Status: Open.

#109 In thu-ml/TurboDiffusion;

· ulucsahin opened on Jan 27, 2026

帧率比较低？

Status: Open.

#108 In thu-ml/TurboDiffusion;

· ScottishFold007 opened on Jan 27, 2026

推理效果

Status: Open.

#107 In thu-ml/TurboDiffusion;

· marmalade-s opened on Jan 27, 2026