Add MLU Support.

[a120092009]

What does this PR do?

Add Cambricon MLU support for diffusers. Context parallelism test will be error without mlu support due to " Supported strategies are: balanced, cpu".

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
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.

Test Code

import torch
from diffusers import AutoModel, QwenImagePipeline, ContextParallelConfig

try:
    torch.distributed.init_process_group("cncl")
    rank = torch.distributed.get_rank()
    device = torch.device("mlu", rank % torch.mlu.device_count())
    torch.mlu.set_device(device)

    transformer = AutoModel.from_pretrained("/data/sd/sd_models/hf_models/Qwen/Qwen-Image", subfolder="transformer", torch_dtype=torch.bfloat16, parallel_config=ContextParallelConfig(ulysses_degree=2), device_map='mlu')
    pipeline = QwenImagePipeline.from_pretrained("/data/sd/sd_models/hf_models/Qwen/Qwen-Image", transformer=transformer, torch_dtype=torch.bfloat16, device_map="mlu")

    prompt = """
    cinematic film still of a cat sipping a margarita in a pool in Palm Springs, California
    highly detailed, high budget hollywood movie, cinemascope, moody, epic, gorgeous, film grain
    """

    # Must specify generator so all ranks start with same latents (or pass your own)
    generator = torch.Generator().manual_seed(42)
    image = pipeline(prompt, num_inference_steps=50, generator=generator).images[0]

    if rank == 0:
        image.save("output.png")

except Exception as e:
    print(f"An error occurred: {e}")
    torch.distributed.breakpoint()
    raise

finally:
    if torch.distributed.is_initialized():
        torch.distributed.destroy_process_group()

Result:

[sayakpaul]

@a120092009 thanks for your PR!

Could you talk a bit about the performance of this new accelerator? Any advantages over regular NVIDIA GPUs, etc.?

[a120092009]

Cambricon is a well-known AI-chip company in China.

For the detailed product message, please see our homepage: https://www.cambricon.com/

[yiyixuxu]

thanks for the PR!

[yiyixuxu]

Suggested change

	elif torch.mlu.is_available():
	elif is_mlu_available:

also need to import this

[a120092009]

fixed

[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]

@bot /style

[github-actions]

Style bot fixed some files and pushed the changes.