Update app.py

app.py

@@ -26,7 +26,7 @@ def predict(text, speaker):
 
     inputs = processor(text=text, return_tensors="pt")
 
-    speaker_embedding = np.load(speaker_embeddings[speaker[:3]])
+    speaker_embedding = np.random.default_rng().uniform(-0.20,0.20,512)
     speaker_embedding = torch.tensor(speaker_embedding).unsqueeze(0)
 
     speech = model.generate_speech(inputs["input_ids"], speaker_embedding, vocoder=vocoder)
@@ -74,19 +74,18 @@ article = """
 """
 
 examples = [
-    ["It is not in the stars to hold our destiny but in ourselves.", "BDL (male)"],
-    ["The octopus and Oliver went to the opera in October.", "CLB (female)"],
-    ["She sells seashells by the seashore. I saw a kitten eating chicken in the kitchen.", "RMS (male)"],
-    ["Brisk brave brigadiers brandished broad bright blades, blunderbusses, and bludgeons—balancing them badly.", "SLT (female)"],
-    ["A synonym for cinnamon is a cinnamon synonym.", "BDL (male)"],
-    ["How much wood would a woodchuck chuck if a woodchuck could chuck wood? He would chuck, he would, as much as he could, and chuck as much wood as a woodchuck would if a woodchuck could chuck wood.", "CLB (female)"],
+    ["It is not in the stars to hold our destiny but in ourselves."],
+    ["The octopus and Oliver went to the opera in October."],
+    ["She sells seashells by the seashore. I saw a kitten eating chicken in the kitchen."],
+    ["Brisk brave brigadiers brandished broad bright blades, blunderbusses, and bludgeons—balancing them badly."],
+    ["A synonym for cinnamon is a cinnamon synonym."],
+    ["How much wood would a woodchuck chuck if a woodchuck could chuck wood? He would chuck, he would, as much as he could, and chuck as much wood as a woodchuck would if a woodchuck could chuck wood."],
 ]
 
 gr.Interface(
     fn=predict,
     inputs=[
         gr.Text(label="Input Text"),
-        gr.Radio(label="Speaker", choices=["BDL (male)", "CLB (female)", "RMS (male)", "SLT (female)"], value="BDL (male)"),
     ],
     outputs=[
         gr.Audio(label="Generated Speech", type="numpy"),