Same thing but for the other places where it's used.

comfy/ldm/modules/attention.py

@@ -20,6 +20,11 @@ except:
 import os
 _ATTN_PRECISION = os.environ.get("ATTN_PRECISION", "fp32")
 
+try:
+    OOM_EXCEPTION = torch.cuda.OutOfMemoryError
+except:
+    OOM_EXCEPTION = Exception
+
 def exists(val):
     return val is not None
 
@@ -316,7 +321,7 @@ class CrossAttentionDoggettx(nn.Module):
                     r1[:, i:end] = einsum('b i j, b j d -> b i d', s2, v)
                     del s2
                 break
-            except torch.cuda.OutOfMemoryError as e:
+            except OOM_EXCEPTION as e:
                 if first_op_done == False:
                     torch.cuda.empty_cache()
                     torch.cuda.ipc_collect()

comfy/ldm/modules/diffusionmodules/model.py

@@ -16,6 +16,10 @@ except:
     XFORMERS_IS_AVAILBLE = False
     print("No module 'xformers'. Proceeding without it.")
 
+try:
+    OOM_EXCEPTION = torch.cuda.OutOfMemoryError
+except:
+    OOM_EXCEPTION = Exception
 
 def get_timestep_embedding(timesteps, embedding_dim):
     """
@@ -229,7 +233,7 @@ class AttnBlock(nn.Module):
                     r1[:, :, i:end] = torch.bmm(v, s2)
                     del s2
                 break
-            except torch.cuda.OutOfMemoryError as e:
+            except OOM_EXCEPTION as e:
                 if first_op_done == False:
                     steps *= 2
                     if steps > 128: