run sv3d_u success but output mp4 can't play

[hengshan123]

hi, I run the script success: python scripts/sampling/simple_video_sample.py --input_path assets/000.jpg --version sv3d_u

Initialized embedder #0: FrozenOpenCLIPImagePredictionEmbedder with 683800065 params. Trainable: False Initialized embedder #1: VideoPredictionEmbedderWithEncoder with 83653863 params. Trainable: False Initialized embedder #2: ConcatTimestepEmbedderND with 0 params. Trainable: False Restored from checkpoints/sv3d_u.safetensors with 0 missing and 0 unexpected keys ############################## Sampling setting ############################## Sampler: EulerEDMSampler Discretization: EDMDiscretization Guider: TrianglePredictionGuider Sampling with EulerEDMSampler for 51 steps: 0%| | 0/51 [00:00<?, ?it/s]/home/j00668593/anaconda3/envs/py310/lib/python3.10/site-packages/torch/utils/checkpoint.py:31: UserWarning: None of the inputs have requires_grad=True. Gradients will be None warnings.warn("None of the inputs have requires_grad=True. Gradients will be None") Sampling with EulerEDMSampler for 51 steps: 98%|██████████████████████████████████████████████████████████████████████████████████████████▏ | 50/51 [00:59<00:01, 1.19s/it]

And i get the output in outputs/simple_video_sample/sv3d_u. But the 000000.mp4 can't be played.

[wzds2015]

Same here for sv3d_p

[zijinY]

same here for sv3d_u

[hengshan123]

I check the origin result vid, this is valid. the problem maybe in imageio

[wzds2015]

confirmed the issue is from imageio. Switch to cv2 solved the problem. fix: https://github.com/Stability-AI/generative-models/pull/327

[zijinY]

confirmed the issue is from imageio. Switch to cv2 solved the problem. fix: #327

Thanks for your advise, the issue is fixed !

[hengshan123]

after install imageio-ffmpeg the problem is solved

[wyweiyangweiGh]

same question