Error occurred on loras

[Crd1140234468]

Hi, thanks for releasing the nice work.
When I try to run the Train MotionDirector on a single video, using the provided data ''./test_data/car_turn/car-turn-24.mp4'', the code encounters an error
Traceback (most recent call last):
File "MotionDirector_train.py", line 1024, in
main(**OmegaConf.load(args.config))
File "MotionDirector_train.py", line 894, in main
loss_spatial, loss_temporal, latents, init_noise = finetune_unet(batch, step, mask_spatial_lora=mask_spatial_lora, mask_temporal_lora=mask_temporal_lora)
File "MotionDirector_train.py", line 820, in finetune_unet
loras[lora_idx + step].scale = 1.
IndexError: list index out of range

What is going on?

[ruizhaocv]

Hi. Which config file did you use? Did you set the num_frames to 24?

[Crd1140234468]

Hi. Which config file did you use? Did you set the num_frames to 24?

I am using config_single_video_24.yaml, and when only replacing the training video, I found that using my own video will detect num_examples=2, what is the reason?

[Crd1140234468]

But after I restored all the default settings of ''config_single_video_24.yaml'', the demo''car_turn/car-turn-24.mp4'' can be trained.

[ruizhaocv]

Hi. This is because your video has more than 24 frames. The training script has been updated for this issue. You can try the new script now.

[Crd1140234468]

Thank you for your reply. I have successfully trained my own video according to the newly updated scripts. In addition, if the training video used exceeds 24 frames, will it affect the training effect?

[ruizhaocv]

It's OK. I think it will not significantly impact the results.