Remove duplicates

script.py

@@ -24,7 +24,7 @@ class ImageDataset(Dataset):
     def __getitem__(self, idx):
         row = self.metadata.iloc[idx]
         image_path = Path(self.images_root_path) / row.filename
-        img = Image.open(image_path)
+        img = Image.open(image_path).convert("RGB")
         img = torch.from_numpy(np.array(img))
         return {"features": img, "observation_id": row.observation_id}
 
@@ -79,7 +79,7 @@ def make_submission(
         for observation_id, class_id in zip(observation_ids, class_ids):
             row = {"observation_id": int(observation_id), "class_id": int(class_id)}
             rows.append(row)
-    submission_df = pd.DataFrame(rows)
+    submission_df = pd.DataFrame(rows).drop_duplicates("observation_id", keep="first")
     submission_df.to_csv(output_csv_path, index=False)