Fix and enforce no trailing whitespace.

comfy/extra_samplers/uni_pc.py

@@ -226,7 +226,7 @@ def model_wrapper(
             The input `model` has the following format:
             ``
                 model(x, t_input, **model_kwargs) -> noise | x_start | v | score
-            `` 
+            ``
 
             The input `classifier_fn` has the following format:
             ``
@@ -240,7 +240,7 @@ def model_wrapper(
             The input `model` has the following format:
             ``
                 model(x, t_input, cond, **model_kwargs) -> noise | x_start | v | score
-            `` 
+            ``
             And if cond == `unconditional_condition`, the model output is the unconditional DPM output.
 
             [4] Ho, Jonathan, and Tim Salimans. "Classifier-free diffusion guidance."
@@ -254,7 +254,7 @@ def model_wrapper(
     ``
         def model_fn(x, t_continuous) -> noise:
             t_input = get_model_input_time(t_continuous)
-            return noise_pred(model, x, t_input, **model_kwargs)         
+            return noise_pred(model, x, t_input, **model_kwargs)
     ``
     where `t_continuous` is the continuous time labels (i.e. epsilon to T). And we use `model_fn` for DPM-Solver.
 
@@ -359,7 +359,7 @@ class UniPC:
         max_val=1.,
         variant='bh1',
     ):
-        """Construct a UniPC. 
+        """Construct a UniPC.
 
         We support both data_prediction and noise_prediction.
         """
@@ -372,7 +372,7 @@ class UniPC:
 
     def dynamic_thresholding_fn(self, x0, t=None):
         """
-        The dynamic thresholding method. 
+        The dynamic thresholding method.
         """
         dims = x0.dim()
         p = self.dynamic_thresholding_ratio
@@ -404,7 +404,7 @@ class UniPC:
 
     def model_fn(self, x, t):
         """
-        Convert the model to the noise prediction model or the data prediction model. 
+        Convert the model to the noise prediction model or the data prediction model.
         """
         if self.predict_x0:
             return self.data_prediction_fn(x, t)
@@ -461,7 +461,7 @@ class UniPC:
 
     def denoise_to_zero_fn(self, x, s):
         """
-        Denoise at the final step, which is equivalent to solve the ODE from lambda_s to infty by first-order discretization. 
+        Denoise at the final step, which is equivalent to solve the ODE from lambda_s to infty by first-order discretization.
         """
         return self.data_prediction_fn(x, s)
 
@@ -510,7 +510,7 @@ class UniPC:
         col = torch.ones_like(rks)
         for k in range(1, K + 1):
             C.append(col)
-            col = col * rks / (k + 1) 
+            col = col * rks / (k + 1)
         C = torch.stack(C, dim=1)
 
         if len(D1s) > 0:
@@ -626,7 +626,7 @@ class UniPC:
             R.append(torch.pow(rks, i - 1))
             b.append(h_phi_k * factorial_i / B_h)
             factorial_i *= (i + 1)
-            h_phi_k = h_phi_k / hh - 1 / factorial_i 
+            h_phi_k = h_phi_k / hh - 1 / factorial_i
 
         R = torch.stack(R)
         b = torch.tensor(b, device=x.device)

comfy/ldm/cascade/stage_b.py

@@ -138,7 +138,7 @@ class StageB(nn.Module):
     #     nn.init.normal_(self.pixels_mapper[2].weight, std=0.02)  # conditionings
     #     torch.nn.init.xavier_uniform_(self.embedding[1].weight, 0.02)  # inputs
     #     nn.init.constant_(self.clf[1].weight, 0)  # outputs
-    # 
+    #
     #     # blocks
     #     for level_block in self.down_blocks + self.up_blocks:
     #         for block in level_block:
@@ -148,7 +148,7 @@ class StageB(nn.Module):
     #                 for layer in block.modules():
     #                     if isinstance(layer, nn.Linear):
     #                         nn.init.constant_(layer.weight, 0)
-    # 
+    #
     # def _init_weights(self, m):
     #     if isinstance(m, (nn.Conv2d, nn.Linear)):
     #         torch.nn.init.xavier_uniform_(m.weight)

comfy/ldm/cascade/stage_c.py

@@ -142,7 +142,7 @@ class StageC(nn.Module):
     #     nn.init.normal_(self.clip_img_mapper.weight, std=0.02)  # conditionings
     #     torch.nn.init.xavier_uniform_(self.embedding[1].weight, 0.02)  # inputs
     #     nn.init.constant_(self.clf[1].weight, 0)  # outputs
-    # 
+    #
     #     # blocks
     #     for level_block in self.down_blocks + self.up_blocks:
     #         for block in level_block:
@@ -152,7 +152,7 @@ class StageC(nn.Module):
     #                 for layer in block.modules():
     #                     if isinstance(layer, nn.Linear):
     #                         nn.init.constant_(layer.weight, 0)
-    # 
+    #
     # def _init_weights(self, m):
     #     if isinstance(m, (nn.Conv2d, nn.Linear)):
     #         torch.nn.init.xavier_uniform_(m.weight)

comfy/ldm/flux/model.py

@@ -168,7 +168,7 @@ class Flux(nn.Module):
                 out = blocks_replace[("single_block", i)]({"img": img,
                                                            "vec": vec,
                                                            "pe": pe,
-                                                           "attn_mask": attn_mask}, 
+                                                           "attn_mask": attn_mask},
                                                           {"original_block": block_wrap})
                 img = out["img"]
             else:

comfy/ldm/hydit/attn_layers.py

@@ -159,7 +159,7 @@ class CrossAttention(nn.Module):
 
         q = q.transpose(-2, -3).contiguous()        # q ->  B, L1, H, C - B, H, L1, C
         k = k.transpose(-2, -3).contiguous()      # k ->  B, L2, H, C - B, H, C, L2
-        v = v.transpose(-2, -3).contiguous() 
+        v = v.transpose(-2, -3).contiguous()
 
         context = optimized_attention(q, k, v, self.num_heads, skip_reshape=True, attn_precision=self.attn_precision)
 

comfy/model_base.py

@@ -787,7 +787,7 @@ class Flux(BaseModel):
         cross_attn = kwargs.get("cross_attn", None)
         if cross_attn is not None:
             out['c_crossattn'] = comfy.conds.CONDRegular(cross_attn)
-        # upscale the attention mask, since now we 
+        # upscale the attention mask, since now we
         attention_mask = kwargs.get("attention_mask", None)
         if attention_mask is not None:
             shape = kwargs["noise"].shape

comfy/model_detection.py

@@ -576,7 +576,7 @@ def unet_config_from_diffusers_unet(state_dict, dtype=None):
             'dtype': dtype, 'in_channels': 9, 'model_channels': 320, 'num_res_blocks': [2, 2, 2, 2], 'transformer_depth': [1, 1, 1, 1, 1, 1, 0, 0],
             'channel_mult': [1, 2, 4, 4], 'transformer_depth_middle': 1, 'use_linear_in_transformer': False, 'context_dim': 768, 'num_heads': 8,
             'transformer_depth_output': [1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0],
-            'use_temporal_attention': False, 'use_temporal_resblock': False}  
+            'use_temporal_attention': False, 'use_temporal_resblock': False}
 
 
     supported_models = [SDXL, SDXL_refiner, SD21, SD15, SD21_uncliph, SD21_unclipl, SDXL_mid_cnet, SDXL_small_cnet, SDXL_diffusers_inpaint, SSD_1B, Segmind_Vega, KOALA_700M, KOALA_1B, SD09_XS, SD_XS, SDXL_diffusers_ip2p, SD15_diffusers_inpaint]

comfy/utils.py

@@ -727,7 +727,7 @@ def bislerp(samples, width, height):
         res *= (b1_norms * (1.0-r) + b2_norms * r).expand(-1,c)
 
         #edge cases for same or polar opposites
-        res[dot > 1 - 1e-5] = b1[dot > 1 - 1e-5] 
+        res[dot > 1 - 1e-5] = b1[dot > 1 - 1e-5]
         res[dot < 1e-5 - 1] = (b1 * (1.0-r) + b2 * r)[dot < 1e-5 - 1]
         return res