Update app.py

app.py

@@ -1,41 +1,20 @@
-import gradio as gr 
-import numpy as np 
-from elevenlabs import voices, generate, set_api_key, UnauthenticatedRateLimitError
+import gradio as gr
 
-def pad_buffer(audio):
-    buffer_size = len(audio)
-    element_size = np.dtype(np.int16).itemsize
-    if buffer_size % element_size != 0:
-        audio = audio + b'\0' * (element_size - (buffer_size % element_size))
-    return audio 
-
-def generate_voice(text, voice_name):
-    model_name = "eleven_multilingual_v1"
-    try:
-        audio = generate(
-            text[:250],
-            voice=voice_name, 
-            model=model_name
-        )
-        return (44100, np.frombuffer(pad_buffer(audio), dtype=np.int16))
-    except UnauthenticatedRateLimitError as e:
-        raise gr.Error("Thanks for trying out ElevenLabs TTS! You've reached the free tier limit. Please provide an API key to continue.") 
-    except Exception as e:
-        raise gr.Error(str(e))
-
-all_voices = voices()
-desired_voices = ["Antonio"]
-filtered_voices = [voice.name for voice in all_voices if voice.name in desired_voices]
-
-input_text = gr.Textbox(label="Input Text", lines=2)
-input_voice = gr.Dropdown(choices=filtered_voices, default="Antonio", label="Voice")
-out_audio = gr.Audio(label="Generated Voice", type="numpy")
+# Define the function to predict using the model
+def predict_image(img):
+    model = gr.load("models/google/vit-base-patch16-224")
+    return model.predict(img)
 
+# Create the interface
 iface = gr.Interface(
-    fn=generate_voice,
-    inputs=[input_text, input_voice],
-    outputs=out_audio,
-    live=True
+    fn=predict_image,
+    inputs=gr.Image(),
+    outputs="label",
+    live=True,
+    capture_session=True,
+    title="Image recognition",
+    description="Upload an image you want to categorize.",
+    theme="Monochrome"
 )
 
 iface.launch()