fix more tests

Author: a-r-r-o-w

src/diffusers/hooks/pyramid_attention_broadcast.py

@@ -18,6 +18,7 @@
 
 import torch
 
+from ..models.attention import AttentionModuleMixin
 from ..models.attention_processor import Attention, MochiAttention
 from ..utils import logging
 from .hooks import HookRegistry, ModelHook
@@ -227,7 +228,7 @@ def apply_pyramid_attention_broadcast(module: torch.nn.Module, config: PyramidAt
         config.spatial_attention_block_skip_range = 2
 
     for name, submodule in module.named_modules():
-        if not isinstance(submodule, _ATTENTION_CLASSES):
+        if not isinstance(submodule, (*_ATTENTION_CLASSES, AttentionModuleMixin)):
             # PAB has been implemented specific to Diffusers' Attention classes. However, this does not mean that PAB
             # cannot be applied to this layer. For custom layers, users can extend this functionality and implement
             # their own PAB logic similar to `_apply_pyramid_attention_broadcast_on_attention_class`.

tests/pipelines/chroma/test_pipeline_chroma_img2img.py

@@ -8,12 +8,7 @@
 from diffusers import AutoencoderKL, ChromaImg2ImgPipeline, ChromaTransformer2DModel, FlowMatchEulerDiscreteScheduler
 from diffusers.utils.testing_utils import floats_tensor, torch_device
 
-from ..test_pipelines_common import (
-    FluxIPAdapterTesterMixin,
-    PipelineTesterMixin,
-    check_qkv_fusion_matches_attn_procs_length,
-    check_qkv_fusion_processors_exist,
-)
+from ..test_pipelines_common import FluxIPAdapterTesterMixin, PipelineTesterMixin, check_qkv_fused_layers_exist
 
 
 class ChromaImg2ImgPipelineFastTests(
@@ -129,12 +124,10 @@ def test_fused_qkv_projections(self):
         # TODO (sayakpaul): will refactor this once `fuse_qkv_projections()` has been added
         # to the pipeline level.
         pipe.transformer.fuse_qkv_projections()
-        assert check_qkv_fusion_processors_exist(pipe.transformer), (
-            "Something wrong with the fused attention processors. Expected all the attention processors to be fused."
+        self.assertTrue(
+            check_qkv_fused_layers_exist(pipe.transformer, ["to_qkv"]),
+            ("Something wrong with the fused attention layers. Expected all the attention projections to be fused."),
         )
-        assert check_qkv_fusion_matches_attn_procs_length(
-            pipe.transformer, pipe.transformer.original_attn_processors
-        ), "Something wrong with the attention processors concerning the fused QKV projections."
 
         inputs = self.get_dummy_inputs(device)
         image = pipe(**inputs).images