pengyige123

[Bug] The low-noise model weights trained in WAN2.2 are then used for further low-noise VSA training, and the loss becomes Nan.

7

### Describe the bug The low-noise model weights trained in WAN2.2 are then used for further low-noise VSA training, and the loss becomes Nan. Our investigation revealed that the Nan...

[Bug] Why is VSA slower than Flash_attention when I run the training script examples/training/finetune/Wan2.1-VSA/Wan-Syn-Data/T2V-14B-VSA.slurm?

2

### Describe the bug The phenomenon is as follows: vsa When using VSA, each step takes 11.91 seconds. Fa When using Flash attention, each step takes 9.39 seconds. thank you...