Update app.py

app.py

@@ -1,74 +1,53 @@
-Hugging Face's logo
-Hugging Face
- 
-Models
-Datasets
-Spaces
-Docs
-Enterprise
-Pricing
-
-
-
- 
-Spaces:
-
-
-willco-afk
-/
-tree-tests 
-
-like
-0
- 
-App
-
-Files
-Community
-Settings
-tree-tests
-/
-app.py
-
-willco-afk's picture
-willco-afk
-Update app.py
-8ef1729
-VERIFIED
-4 minutes ago
-raw
-
-Copy download link
-history
-blame
-edit
-delete
-
-1 kB
-import streamlit as st
-import numpy as np
-import tensorflow as tf
-from PIL import Image
 import os
-
-# Load the model directly from the path (no need to unzip)
-model_path = 'your_trained_model.keras'  # Path to your model
-model = tf.keras.models.load_model(model_path)  # Load the model
-
-# Streamlit UI for uploading an image
-st.title("Tree Decoration Prediction")
-uploaded_image = st.file_uploader("Upload an image", type=["jpg", "png"])
-
-if uploaded_image:
-    # Process the image for prediction
-    img = Image.open(uploaded_image)
-    img = img.resize((224, 224))  # Resize to match the input size of your model
-    img_array = np.array(img) / 255.0  # Normalize the image
-    img_array = np.expand_dims(img_array, axis=0)  # Add batch dimension
-
-    # Make prediction
-    prediction = model.predict(img_array)
-
-    # Display the result
-    st.image(uploaded_image, caption="Uploaded Image.", use_column_width=True)
-    st.write(f"Prediction: {'Decorated' if prediction[0] > 0.5 else 'Undecorated'}")
+from fastapi import FastAPI
+from google.cloud import storage
+from keras.models import load_model
+import tempfile
+import numpy as np
+from pydantic import BaseModel
+
+app = FastAPI()
+
+# Function to load the model from Google Cloud Storage
+def load_model_from_gcs(model_path):
+    client = storage.Client()
+    bucket = client.get_bucket('tree-decorator-model')  # Your bucket name
+    blob = bucket.blob(model_path)  # Path to your model in the bucket
+
+    # Save the model file locally in a temporary file
+    with tempfile.NamedTemporaryFile(delete=False) as temp_file:
+        blob.download_to_filename(temp_file.name)  # Download model to temporary file
+        model = load_model(temp_file.name)  # Load model from the temporary file
+    
+    return model
+
+# Load the model from Google Cloud Storage (provide the path to your model in the bucket)
+model = load_model_from_gcs('models/your_trained_model.keras')  # Path in GCS
+
+# Pydantic model for the incoming prediction request (adjust as needed)
+class ImageData(BaseModel):
+    image: str  # Base64-encoded image or URL of the image (you can adjust this)
+
+@app.get("/")
+def read_root():
+    return {"message": "Welcome to the Tree Decorator API!"}
+
+@app.post("/predict/")
+async def predict(data: ImageData):
+    # Example: Decode the image, preprocess it, and use the model for prediction
+    # Decode and preprocess the image data as required (e.g., using Pillow, OpenCV, etc.)
+    
+    # For simplicity, we'll assume 'data.image' is already preprocessed or passed in an acceptable format
+
+    # Example prediction (replace with actual image processing and prediction logic)
+    # prediction = model.predict(processed_image) 
+    
+    # Dummy response for demonstration
+    prediction = {"prediction": "decorated" if np.random.random() > 0.5 else "not decorated"}
+    
+    return prediction
+
+# Run the FastAPI app
+if __name__ == "__main__":
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=int(os.getenv("PORT", 8080)))