diff --git a/Python/convert_onnx.py b/Python/convert_onnx.py
index 29f305027091e6d6a990151ff15a79adb04c1e5e..d59ebffea8b581a0a1419592af7dd12566132a9b 100644
--- a/Python/convert_onnx.py
+++ b/Python/convert_onnx.py
@@ -1,8 +1,18 @@
 import tensorflow as tf
 import tf2onnx
+import argparse
+
+# WHAT ?
+parser = argparse.ArgumentParser(description="WHAT ?!")
+parser.add_argument("--model", choices=["1", "2"], default="1",
+                    help="1 = ResNet50, 2 = Xception")
+args = parser.parse_args()
 
 # Load Sequential model
-seq_model = tf.keras.models.load_model("../models/resnet50.h5", compile=False)
+if args.model == "1":
+    seq_model = tf.keras.models.load_model("../models/ResNet50/pokemon_resnet50.h5", compile=False)
+elif args.model == "2":
+    seq_model = tf.keras.models.load_model("../models/ResNet50/pokemon_xception.h5", compile=False)
 
 # Create input layer with same shape
 inputs = tf.keras.Input(shape=(224, 224, 3), name="input")
@@ -15,9 +25,18 @@ model = tf.keras.Model(inputs=inputs, outputs=outputs)
 
 # Convert to ONNX
 spec = (tf.TensorSpec((1, 224, 224, 3), tf.float32, name="input"),)
-onnx_model, _ = tf2onnx.convert.from_keras(
-    model,
-    input_signature=spec,
-    opset=13,
-    output_path="../models/resnet50.onnx"
-)
\ No newline at end of file
+
+if args.model == "1":
+    onnx_model, _ = tf2onnx.convert.from_keras(
+        model,
+        input_signature=spec,
+        opset=13,
+        output_path="../models/ResNet50/pokemon_resnet50.onnx"
+    )
+elif args.model == "2":
+    onnx_model, _ = tf2onnx.convert.from_keras(
+        model,
+        input_signature=spec,
+        opset=13,
+        output_path="../models/ResNet50/pokemon_xception.onnx"
+    )
\ No newline at end of file