diff --git a/src/diffusers/pipelines/glide/pipeline_glide.py b/src/diffusers/pipelines/glide/pipeline_glide.py
index 8d7a75f542..5ba28d87db 100644
--- a/src/diffusers/pipelines/glide/pipeline_glide.py
+++ b/src/diffusers/pipelines/glide/pipeline_glide.py
@@ -817,7 +817,7 @@ class GlidePipeline(DiffusionPipeline):
         num_trained_timesteps = self.upscale_scheduler.timesteps
         inference_step_times = range(0, num_trained_timesteps, num_trained_timesteps // num_inference_steps_upscale)
 
-        for t in tqdm.tqdm(reversed(range(num_inference_steps_upscale)), total=num_inference_steps_upscale):
+        for t in tqdm(reversed(range(num_inference_steps_upscale)), total=num_inference_steps_upscale):
             # 1. predict noise residual
             with torch.no_grad():
                 time_input = torch.tensor([inference_step_times[t]] * image.shape[0], device=torch_device)
diff --git a/src/diffusers/pipelines/pndm/pipeline_pndm.py b/src/diffusers/pipelines/pndm/pipeline_pndm.py
index 27f4770490..29d24fdc82 100644
--- a/src/diffusers/pipelines/pndm/pipeline_pndm.py
+++ b/src/diffusers/pipelines/pndm/pipeline_pndm.py
@@ -53,7 +53,7 @@ class PNDMPipeline(DiffusionPipeline):
             image = self.scheduler.step_prk(model_output, t, image, num_inference_steps)["prev_sample"]
 
         timesteps = self.scheduler.get_time_steps(num_inference_steps)
-        for t in tqdm.tqdm(range(len(timesteps))):
+        for t in tqdm(range(len(timesteps))):
             t_orig = timesteps[t]
             model_output = self.unet(image, t_orig)