fix E721 Do not compare types, use isinstance() (huggingface#4992)

kashif · web-flow · commit 73bf620dec56 · 2023-09-12T16:52:25.000+02:00
diff --git a/examples/community/lpw_stable_diffusion_xl.py b/examples/community/lpw_stable_diffusion_xl.py
@@ -1138,7 +1138,7 @@ def __call__(
         num_warmup_steps = max(len(timesteps) - num_inference_steps * self.scheduler.order, 0)
 
         # 7.1 Apply denoising_end
-        if denoising_end is not None and type(denoising_end) == float and denoising_end > 0 and denoising_end < 1:
+        if denoising_end is not None and isinstance(denoising_end, float) and denoising_end > 0 and denoising_end < 1:
             discrete_timestep_cutoff = int(
                 round(
                     self.scheduler.config.num_train_timesteps
diff --git a/examples/community/stable_diffusion_xl_reference.py b/examples/community/stable_diffusion_xl_reference.py
@@ -701,7 +701,7 @@ def hacked_UpBlock2D_forward(self, hidden_states, res_hidden_states_tuple, temb=
         num_warmup_steps = max(len(timesteps) - num_inference_steps * self.scheduler.order, 0)
 
         # 10.1 Apply denoising_end
-        if denoising_end is not None and type(denoising_end) == float and denoising_end > 0 and denoising_end < 1:
+        if denoising_end is not None and isinstance(denoising_end, float) and denoising_end > 0 and denoising_end < 1:
             discrete_timestep_cutoff = int(
                 round(
                     self.scheduler.config.num_train_timesteps
diff --git a/src/diffusers/experimental/rl/value_guided_sampling.py b/src/diffusers/experimental/rl/value_guided_sampling.py
@@ -76,7 +76,7 @@ def de_normalize(self, x_in, key):
         return x_in * self.stds[key] + self.means[key]
 
     def to_torch(self, x_in):
-        if type(x_in) is dict:
+        if isinstance(x_in, dict):
             return {k: self.to_torch(v) for k, v in x_in.items()}
         elif torch.is_tensor(x_in):
             return x_in.to(self.unet.device)
diff --git a/src/diffusers/pipelines/audio_diffusion/pipeline_audio_diffusion.py b/src/diffusers/pipelines/audio_diffusion/pipeline_audio_diffusion.py
@@ -178,7 +178,7 @@ def __call__(
         self.scheduler.set_timesteps(steps)
         step_generator = step_generator or generator
         # For backwards compatibility
-        if type(self.unet.config.sample_size) == int:
+        if isinstance(self.unet.config.sample_size, int):
             self.unet.config.sample_size = (self.unet.config.sample_size, self.unet.config.sample_size)
         if noise is None:
             noise = randn_tensor(
diff --git a/src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl.py b/src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl.py
@@ -810,7 +810,7 @@ def __call__(
         num_warmup_steps = max(len(timesteps) - num_inference_steps * self.scheduler.order, 0)
 
         # 7.1 Apply denoising_end
-        if denoising_end is not None and type(denoising_end) == float and denoising_end > 0 and denoising_end < 1:
+        if denoising_end is not None and isinstance(denoising_end, float) and denoising_end > 0 and denoising_end < 1:
             discrete_timestep_cutoff = int(
                 round(
                     self.scheduler.config.num_train_timesteps
diff --git a/src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl_img2img.py b/src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl_img2img.py
@@ -885,7 +885,7 @@ def __call__(
 
         # 5. Prepare timesteps
         def denoising_value_valid(dnv):
-            return type(denoising_end) == float and 0 < dnv < 1
+            return isinstance(denoising_end, float) and 0 < dnv < 1
 
         self.scheduler.set_timesteps(num_inference_steps, device=device)
         timesteps, num_inference_steps = self.get_timesteps(
diff --git a/src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl_inpaint.py b/src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl_inpaint.py
@@ -1120,7 +1120,7 @@ def __call__(
 
         # 4. set timesteps
         def denoising_value_valid(dnv):
-            return type(denoising_end) == float and 0 < dnv < 1
+            return isinstance(denoising_end, float) and 0 < dnv < 1
 
         self.scheduler.set_timesteps(num_inference_steps, device=device)
         timesteps, num_inference_steps = self.get_timesteps(
diff --git a/src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl_instruct_pix2pix.py b/src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl_instruct_pix2pix.py
@@ -837,7 +837,7 @@ def __call__(
 
         # 11. Denoising loop
         num_warmup_steps = max(len(timesteps) - num_inference_steps * self.scheduler.order, 0)
-        if denoising_end is not None and type(denoising_end) == float and denoising_end > 0 and denoising_end < 1:
+        if denoising_end is not None and isinstance(denoising_end, float) and denoising_end > 0 and denoising_end < 1:
             discrete_timestep_cutoff = int(
                 round(
                     self.scheduler.config.num_train_timesteps
diff --git a/src/diffusers/pipelines/t2i_adapter/pipeline_stable_diffusion_xl_adapter.py b/src/diffusers/pipelines/t2i_adapter/pipeline_stable_diffusion_xl_adapter.py
@@ -886,7 +886,7 @@ def __call__(
         num_warmup_steps = max(len(timesteps) - num_inference_steps * self.scheduler.order, 0)
 
         # 7.1 Apply denoising_end
-        if denoising_end is not None and type(denoising_end) == float and denoising_end > 0 and denoising_end < 1:
+        if denoising_end is not None and isinstance(denoising_end, float) and denoising_end > 0 and denoising_end < 1:
             discrete_timestep_cutoff = int(
                 round(
                     self.scheduler.config.num_train_timesteps
diff --git a/tests/pipelines/consistency_models/test_consistency_models.py b/tests/pipelines/consistency_models/test_consistency_models.py
@@ -193,7 +193,7 @@ def get_inputs(self, seed=0, get_fixed_latents=False, device="cpu", dtype=torch.
         return inputs
 
     def get_fixed_latents(self, seed=0, device="cpu", dtype=torch.float32, shape=(1, 3, 64, 64)):
-        if type(device) == str:
+        if isinstance(device, str):
             device = torch.device(device)
         generator = torch.Generator(device=device).manual_seed(seed)
         latents = randn_tensor(shape, generator=generator, device=device, dtype=dtype)
diff --git a/tests/pipelines/unidiffuser/test_unidiffuser.py b/tests/pipelines/unidiffuser/test_unidiffuser.py
@@ -109,7 +109,7 @@ def get_dummy_inputs(self, device, seed=0):
         return inputs
 
     def get_fixed_latents(self, device, seed=0):
-        if type(device) == str:
+        if isinstance(device, str):
             device = torch.device(device)
         generator = torch.Generator(device=device).manual_seed(seed)
         # Hardcode the shapes for now.
@@ -545,7 +545,7 @@ def get_inputs(self, device, seed=0, generate_latents=False):
         return inputs
 
     def get_fixed_latents(self, device, seed=0):
-        if type(device) == str:
+        if isinstance(device, str):
             device = torch.device(device)
         latent_device = torch.device("cpu")
         generator = torch.Generator(device=latent_device).manual_seed(seed)
@@ -648,7 +648,7 @@ def get_inputs(self, device, seed=0, generate_latents=False):
         return inputs
 
     def get_fixed_latents(self, device, seed=0):
-        if type(device) == str:
+        if isinstance(device, str):
             device = torch.device(device)
         latent_device = torch.device("cpu")
         generator = torch.Generator(device=latent_device).manual_seed(seed)
