[Pipeline] Fix error of SVD pipeline when num_videos_per_prompt > 1

[wuyushuwys]

What does this PR do?

Fixes #Issue

it should repeat latents before concatenate do_classifier_free_guidance latents to align with the CFG process after noise_pred.

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline?
• Did you read our philosophy doc (important for complex PRs)?
• Was this discussed/approved via a GitHub issue or the forum? Please add a link to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

@sayakpaul @yiyixuxu @DN6
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

[sayakpaul]

Hi,

Thanks for your PR :) Possible to add a test for this as well?

@DN6 seeing a couple of issues/PRs related to using SVD for generating more than one videos per prompt. Should we consider adding a SLOW test for it as well?

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[wuyushuwys]

Hi,

Thanks for your PR :) Possible to add a test for this as well?

@DN6 seeing a couple of issues/PRs related to using SVD for generating more than one videos per prompt. Should we consider adding a SLOW test for it as well?

Sure, I will try to add a test for this.

[wuyushuwys]

Hello , @sayakpaul

I am not familiar with pytest, so I simply followed the instructions here.
Since I modified the file in src/diffusers/pipelines/stable_video_diffusion/pipeline_stable_video_diffusion.py, I executed the following command pytest tests/pipelines/stable_video_diffusion/test_stable_video_diffusion.py and I attached the output.

$ pytest tests/pipelines/stable_video_diffusion/test_stable_video_diffusion.py 
============================================================== test session starts ===============================================================
platform linux -- Python 3.10.14, pytest-8.2.0, pluggy-1.5.0
rootdir: /home/wu.yushu/src/efficient_diffusion/fix_diffusers/diffusers
configfile: pyproject.toml
plugins: timeout-2.3.1, xdist-3.5.0, requests-mock-1.10.0
collected 26 items                                                                                                                               

tests/pipelines/stable_video_diffusion/test_stable_video_diffusion.py .s.......sss............ss                                           [100%]

================================================================ warnings summary ================================================================
diffuser_dev/lib/python3.10/site-packages/flax/core/meta.py:31
diffuser_dev/lib/python3.10/site-packages/flax/core/meta.py:31: DeprecationWarning: jax.experimental.maps and jax.experimental.maps.xmap are deprecated and will be removed in a future release. Use jax.experimental.shard_map or jax.vmap with the spmd_axis_name argument for expressing SPMD device-parallel computations. Please file an issue on https://github.com/google/jax/issues if neither jax.experimental.shard_map nor jax.vmap are suitable for your use case.
    from jax.experimental import maps

tests/pipelines/stable_video_diffusion/test_stable_video_diffusion.py::StableVideoDiffusionPipelineFastTests::test_save_load_float16
diffuser_dev/lib/python3.10/site-packages/torch/nn/modules/conv.py:605: UserWarning: Plan failed with a cudnnException: CUDNN_BACKEND_EXECUTION_PLAN_DESCRIPTOR: cudnnFinalize Descriptor Failed cudnn_status: CUDNN_STATUS_NOT_SUPPORTED (Triggered internally at ../aten/src/ATen/native/cudnn/Conv_v8.cpp:919.)
    return F.conv3d(

-- Docs: https://docs.pytest.org/en/stable/how-to/capture-warnings.html
=================================================== 20 passed, 6 skipped, 2 warnings in 22.83s ===================================================

I am not sure if it is what you expected. Thanks.

[sayakpaul]

@DN6 could you give this a look too?

[DN6]

Thanks for catching this! LGTM 👍🏽