Update app.py

app.py

@@ -1,25 +1,43 @@
+import os
+# Suppress TensorFlow info logs
+os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'  # Hide INFO & WARNING, only show errors
+# Optional: disable oneDNN logs if you want completely consistent float ops
+# os.environ['TF_ENABLE_ONEDNN_OPTS'] = '0'
+
 import gradio as gr
-from tensorflow.keras.models import load_model
-from tensorflow.keras.preprocessing import image as keras_image
-from PIL import Image
-import numpy as np
-import requests
-import io
+from TTS.api import TTS
+
+# Initialize Coqui TTS model (downloaded once and cached)
+# Replace with any model name from https://huggingface.co/coqui-ai
+# "tts_models/en/ljspeech/tacotron2-DDC" is small and works offline
+tts = TTS(model_name="tts_models/en/ljspeech/tacotron2-DDC")
 
-model = load_model("plant_disease_model.h5")
+def generate_speech(text):
+    """
+    Takes text input and generates speech audio using Coqui TTS.
+    Returns the path to the audio file for Gradio playback.
+    """
+    if not text.strip():
+        return None
+    output_path = "output.wav"
+    # Generate speech and save to file
+    tts.tts_to_file(text=text, file_path=output_path)
+    return output_path
 
-disease_map = {0: "Powdery Mildew", 1: "Leaf Spot", 2: "Healthy"}
-rec_map = {0: "Apply fungicide", 1: "Remove affected leaves", 2: "No action needed"}
-IMG_SIZE = (224, 224)
+# Simple Gradio UI
+with gr.Blocks() as demo:
+    gr.Markdown("# Offline Coqui TTS (Hugging Face Space)")
+    gr.Markdown("Enter text below and hear it synthesized offline using Coqui TTS.")
+    
+    with gr.Row():
+        text_input = gr.Textbox(label="Enter Text", placeholder="Type something...", lines=2)
+    with gr.Row():
+        speak_button = gr.Button("Generate Speech")
+    with gr.Row():
+        audio_output = gr.Audio(label="Generated Audio", type="filepath")
 
-def predict_leaf_disease(img):
-    img = img.convert("RGB")
-    img = img.resize(IMG_SIZE)
-    img_array = keras_image.img_to_array(img)
-    img_array = np.expand_dims(img_array, axis=0) / 255.0
-    preds = model.predict(img_array)
-    idx = np.argmax(preds)
-    severity = float(np.max(preds))
-    return f"Disease: {disease_map[idx]}\nSeverity: {severity:.2f}\nRecommendation: {rec_map[idx]}"
+    speak_button.click(fn=generate_speech, inputs=text_input, outputs=audio_output)
 
-iface = gr.Interface(fn=predict_leaf_disease, inputs=gr.Image(type="pil"), outputs="text")
+if __name__ == "__main__":
+    # Launch Gradio app (port/host auto-managed by HF Spaces)
+    demo.launch()