sdxl: configure edm inference and dynamic thresholding

[Fzilan]

What does this PR do?

Adds : args of EDMDiscretization & DynamicThresholding

1 Support args setting of EDMDiscretization in SDXL inference.

Usage of EDMDiscretization:

=======Example ① EDMEulerScheduler======= directly set the following args in sampling_without_streamlit.py:

--sampler EulerEDMSampler
--discretization EDMDiscretization 
--sigma_min 0.002
--sigma_max 80.0
--rho 7.0

or

set --sampler EulerEDMSampler in sampling_without_streamlit.py and set discretization_config as follows in denoiser_config of examples/stable_diffusion_xl/configs/inference/sd_xl_base.yaml

...
        discretization_config:
          target: gm.modules.diffusionmodules.discretizer.EDMDiscretization
          params:
            sigma_min: 0.02
            sigma_max: 80.0
            rho: 7.0

=======Example ② EDMDPMSolverMultistepScheduler======= directly set the following args in sampling_without_streamlit.py:

--sampler DPMPP2MSampler  #dpm++ 2nd order solver, multistep
--discretization EDMDiscretization 
--sigma_min 0.002
--sigma_max 80.0
--rho 7.0

or

set --sampler DPMPP2MSampler in sampling_without_streamlit.py and set discretization_config as follows in denoiser_config of examples/stable_diffusion_xl/configs/inference/sd_xl_base.yaml

...
        discretization_config:
          target: gm.modules.diffusionmodules.discretizer.EDMDiscretization
          params:
            sigma_min: 0.02
            sigma_max: 80.0
            rho: 7.0

2 Support dynamic thresholding method for CFG (classifier-free guidance)

Useage:

python sampling_without_streamlit.py \
...
--thresholding True 
--dynamic_thresholding_ratio 0.995 
--sample_max_value 1.0
... 

fixs

1. dreambooth sdxl config updates
2. timestepbias bug

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 make sure to update the documentation with your changes? E.g. record bug fixes or new features in What's New. Here are the documentation guidelines
• [ ] Did you build and run the code without any errors?
• [ ] Did you report the running environment (NPU type/MS version) and performance in the doc? (better record it for data loading, model inference, or training tasks)
• [ ] Did you write any new necessary tests?

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.

@xxx