Update app.py

app.py

@@ -3,43 +3,52 @@ import requests
 from PIL import Image
 from io import BytesIO
 
-st.title("Cucumber Disease Prediction")
-uploaded_file = st.file_uploader("Upload a photo of a cucumber plant or cucumber", type=["jpg", "jpeg", "png"])
+# Configuración de la API
+API_URL = "https://api-inference.huggingface.co/models/riabayonaor/modelo_prediccion_enfermedades_pepinos"
+META_LLAMA_API_URL = "https://api-inference.huggingface.co/models/meta-llama/Meta-Llama-3-8B-Instruct"
+headers = {"Authorization": "Bearer YOUR_HUGGINGFACE_API_KEY"}  # Reemplaza con tu API Key
+
+def query(image_bytes):
+    response = requests.post(API_URL, headers=headers, data=image_bytes)
+    return response.json()
+
+def llama_query(prompt):
+    response = requests.post(META_LLAMA_API_URL, headers=headers, json={"inputs": prompt})
+    return response.json()
+
+# Interfaz de Streamlit
+st.title("Predicción de Enfermedades en Pepinos")
+uploaded_file = st.file_uploader("Sube una foto de una planta de pepino o un pepino", type=["jpg", "jpeg", "png"])
 
 if uploaded_file is not None:
     image = Image.open(uploaded_file)
-    st.image(image, caption='Uploaded Image.', use_column_width=True)
-    st.write("")
-    st.write("Classifying...")
+    st.image(image, caption='Imagen subida.', use_column_width=True)
+    st.write("Clasificando...")
 
-    # Convert image to bytes
+    # Convertir la imagen a bytes
     img_byte_arr = BytesIO()
     image.save(img_byte_arr, format='PNG')
     img_byte_arr = img_byte_arr.getvalue()
 
-    # Send image to Hugging Face model
-    headers = {"Authorization": "Bearer YOUR_HUGGINGFACE_API_KEY"}
-    api_url = "https://api-inference.huggingface.co/models/riabayonaor/modelo_prediccion_enfermedades_pepinos"
-    response = requests.post(api_url, headers=headers, files={"file": img_byte_arr})
-    
-    if response.status_code == 200:
-        predictions = response.json()
-        # Assuming the predictions are in the format [{label: "label1", score: 0.95}, {label: "label2", score: 0.05}]
+    # Enviar la imagen al modelo de Hugging Face
+    predictions = query(img_byte_arr)
+
+    if "error" not in predictions:
+        # Suponiendo que las predicciones están en el formato [{label: "label1", score: 0.95}, {label: "label2", score: 0.05}]
         top_prediction = max(predictions, key=lambda x: x["score"])
-        st.write(f"Top prediction: {top_prediction['label']} with confidence {top_prediction['score']:.2f}")
+        st.write(f"Predicción principal: {top_prediction['label']} con confianza {top_prediction['score']:.2f}")
 
-        # Use top label for Meta Llama model
-        prompt = f"This disease is {top_prediction['label']}. Explain what it is and suggest possible insecticides or solutions."
-        
-        # Call Meta Llama model
-        meta_llama_url = "https://api-inference.huggingface.co/models/meta-llama/Meta-Llama-3-8B-Instruct"
-        llama_response = requests.post(meta_llama_url, headers=headers, json={"inputs": prompt})
+        # Usar la etiqueta principal para el modelo de Meta Llama
+        prompt = f"Esta enfermedad es {top_prediction['label']}. Explica qué es y sugiere posibles insecticidas o soluciones."
+
+        # Llamar al modelo Meta Llama
+        llama_response = llama_query(prompt)
         
-        if llama_response.status_code == 200:
-            explanation = llama_response.json()[0]["generated_text"]
-            st.write("Explanation and Possible Solutions:")
+        if "error" not in llama_response:
+            explanation = llama_response[0]["generated_text"]
+            st.write("Explicación y posibles soluciones:")
             st.write(explanation)
         else:
-            st.write("Failed to get a response from Meta Llama model.")
+            st.write("No se pudo obtener una respuesta del modelo Meta Llama.")
     else:
-        st.write("Failed to classify the image.")
+        st.write("No se pudo clasificar la imagen.")