mirror of
https://github.com/huggingface/diffusers.git
synced 2026-06-16 00:35:12 +08:00
2d8a41cae8
* Add wanx pipeline, model and example * wanx_merged_v1 * change WanX into Wan * fix i2v fp32 oom error Link: https://code.alibaba-inc.com/open_wanx2/diffusers/codereview/20607813 * support t2v load fp32 ckpt * add example * final merge v1 * Update autoencoder_kl_wan.py * up * update middle, test up_block * up up * one less nn.sequential * up more * up * more * [refactor] [wip] Wan transformer/pipeline (#10926) * update * update * refactor rope * refactor pipeline * make fix-copies * add transformer test * update * update * make style * update tests * tests * conversion script * conversion script * update * docs * remove unused code * fix _toctree.yml * update dtype * fix test * fix tests: scale * up * more * Apply suggestions from code review * Apply suggestions from code review * style * Update scripts/convert_wan_to_diffusers.py * update docs * fix --------- Co-authored-by: Yitong Huang <huangyitong.hyt@alibaba-inc.com> Co-authored-by: 亚森 <wangjiayu.wjy@alibaba-inc.com> Co-authored-by: Aryan <aryan@huggingface.co>
2.5 KiB
2.5 KiB
Wan
Wan 2.1 by the Alibaba Wan Team.
Make sure to check out the Schedulers guide to learn how to explore the tradeoff between scheduler speed and quality, and see the reuse components across pipelines section to learn how to efficiently load the same components into multiple pipelines.
Recommendations for inference:
- VAE in
torch.float32for better decoding quality. num_framesshould be of the form4 * k + 1, for example49or81.- For smaller resolution videos, try lower values of
shift(between2.0to5.0) in the Scheduler. For larger resolution videos, try higher values (between7.0and12.0). The default value is3.0for Wan.
Using a custom scheduler
Wan can be used with many different schedulers, each with their own benefits regarding speed and generation quality. By default, Wan uses the UniPCMultistepScheduler(prediction_type="flow_prediction", use_flow_sigmas=True, flow_shift=3.0) scheduler. You can use a different scheduler as follows:
from diffusers import FlowMatchEulerDiscreteScheduler, UniPCMultistepScheduler, WanPipeline
scheduler_a = FlowMatchEulerDiscreteScheduler(shift=5.0)
scheduler_b = UniPCMultistepScheduler(prediction_type="flow_prediction", use_flow_sigmas=True, flow_shift=4.0)
pipe = WanPipeline.from_pretrained("Wan-AI/Wan2.1-T2V-1.3B-Diffusers", scheduler=<CUSTOM_SCHEDULER_HERE>)
# or,
pipe.scheduler = <CUSTOM_SCHEDULER_HERE>
WanPipeline
autodoc WanPipeline
- all
- call
WanImageToVideoPipeline
autodoc WanImageToVideoPipeline
- all
- call
WanPipelineOutput
autodoc pipelines.wan.pipeline_output.WanPipelineOutput