Update app/main.py

app/main.py

@@ -1,21 +1,30 @@
-from fastapi import FastAPI
-from pydantic import BaseModel
-import numpy as np
-from tensorflow import keras
-
-app = FastAPI()
-
-# Load model
-model = keras.models.load_model("best_model.h5")
-
-# Input schema
-class InputData(BaseModel):
-    pixels: list  # flattened 28x28 = 784 values
-
-@app.post("/predict")
-def predict(data: InputData):
-    # Convert list → NumPy
-    X = np.array(data.pixels).reshape(1, 28, 28, 1) / 255.0  # normalize if trained that way
-    y_pred = model.predict(X)
-    predicted_class = int(np.argmax(y_pred, axis=1)[0])
-    return {"prediction": predicted_class}
+from fastapi import FastAPI
+from pydantic import BaseModel
+import numpy as np
+from tensorflow import keras
+import os
+
+app = FastAPI()
+
+# Debug: print current directory and files
+print("Current working directory:", os.getcwd())
+print("Files in app folder:", os.listdir("."))
+
+# Load model (must match filename)
+model_path = "best_model.h5"
+model = keras.models.load_model(model_path)
+
+# Input schema
+class InputData(BaseModel):
+    pixels: list  # 784 flattened pixels (28x28)
+
+@app.get("/")
+def root():
+    return {"message": "MNIST API running"}
+
+@app.post("/predict")
+def predict(data: InputData):
+    X = np.array(data.pixels).reshape(1, 28, 28, 1) / 255.0
+    y_pred = model.predict(X)
+    predicted_class = int(np.argmax(y_pred, axis=1)[0])
+    return {"prediction": predicted_class}