Default values in SD3 pipelines when submodules are not loaded

[hlky]

What does this PR do?

Values like vae_scale_factor in pipelines are derived from configurations of component modules e.g. self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1), some pipelines are missing default values when the module is not loaded

self.vae_scale_factor = (
    2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8
)

This PR adds default values in SD3 pipelines.

Reproduction

from diffusers import StableDiffusion3Img2ImgPipeline
import torch

pipe = StableDiffusion3Img2ImgPipeline.from_pretrained(
    "stabilityai/stable-diffusion-3.5-large",
    vae=None, # or `tokenizer` or `transformer`
    torch_dtype=torch.bfloat16,
)

Who can review?

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.

[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.

[yiyixuxu]

thank you!