Add MLU Support.

src/diffusers/pipelines/pipeline_utils.py

@@ -62,6 +62,7 @@
     is_accelerate_version,
     is_hpu_available,
     is_torch_npu_available,
+    is_mlu_available,
     is_torch_version,
     is_transformers_version,
     logging,

src/diffusers/utils/__init__.py

@@ -81,6 +81,7 @@
     is_google_colab,
     is_hf_hub_version,
     is_hpu_available,
+    is_mlu_available,
     is_inflect_available,
     is_invisible_watermark_available,
     is_k_diffusion_available,

src/diffusers/utils/import_utils.py

@@ -192,6 +192,7 @@ def _is_package_available(pkg_name: str, get_dist_name: bool = False) -> Tuple[b
 
 _torch_xla_available, _torch_xla_version = _is_package_available("torch_xla")
 _torch_npu_available, _torch_npu_version = _is_package_available("torch_npu")
+_torch_mlu_available, _torch_mlu_version = _is_package_available("torch_mlu")
 _transformers_available, _transformers_version = _is_package_available("transformers")
 _hf_hub_available, _hf_hub_version = _is_package_available("huggingface_hub")
 _kernels_available, _kernels_version = _is_package_available("kernels")
@@ -243,6 +244,10 @@ def is_torch_npu_available():
     return _torch_npu_available
 
 
+def is_mlu_available():
+    return _torch_mlu_available
+
+
 def is_flax_available():
     return _flax_available
 

src/diffusers/utils/torch_utils.py

@@ -286,6 +286,8 @@ def get_device():
         return "xpu"
     elif torch.backends.mps.is_available():
         return "mps"
+    elif torch.mlu.is_available():
+        return "mlu"
     else:
         return "cpu"