[LDMTextToImagePipeline] make text model generic (#162)

make text model generic

src/diffusers/pipelines/latent_diffusion/pipeline_latent_diffusion.py

@@ -45,11 +45,11 @@ class LDMTextToImagePipeline(DiffusionPipeline):
         # get unconditional embeddings for classifier free guidance
         if guidance_scale != 1.0:
             uncond_input = self.tokenizer([""] * batch_size, padding="max_length", max_length=77, return_tensors="pt")
-            uncond_embeddings = self.bert(uncond_input.input_ids.to(torch_device))
+            uncond_embeddings = self.bert(uncond_input.input_ids.to(torch_device))[0]
 
         # get prompt text embeddings
         text_input = self.tokenizer(prompt, padding="max_length", max_length=77, return_tensors="pt")
-        text_embeddings = self.bert(text_input.input_ids.to(torch_device))
+        text_embeddings = self.bert(text_input.input_ids.to(torch_device))[0]
 
         latents = torch.randn(
             (batch_size, self.unet.in_channels, self.unet.sample_size, self.unet.sample_size),
@@ -618,5 +618,4 @@ class LDMBertModel(LDMBertPreTrainedModel):
             output_hidden_states=output_hidden_states,
             return_dict=return_dict,
         )
-        sequence_output = outputs[0]
-        return sequence_output
+        return outputs