Update app.py

app.py

@@ -127,45 +127,27 @@ def extract_full_paper_with_labels(pdf_path, progress=None):
 
 def upload_with_progress(file_path, repo_id, token, progress):
     """
-    Upload file to Hugging Face Dataset with progress tracking.
+    Upload file to Hugging Face Dataset using the official SDK with progress tracking.
     """
-    import requests  # Ensure this is imported if not already
+    from huggingface_hub import HfApi
+    from huggingface_hub.utils import HfHubHTTPError
+    import os
+
     print(f"📤 Starting upload of Parquet: {file_path}")
     file_size = os.path.getsize(file_path)
 
     api = HfApi()
 
     try:
-        # Open the file in binary read mode
-        with open(file_path, 'rb') as f:
-            chunk_size = 1024 * 1024  # 1 MB chunks
-            uploaded = 0
-
-            # Prepare headers
-            headers = {
-                "Authorization": f"Bearer {token}"
-            }
-
-            # Construct upload URL
-            upload_url = f"https://huggingface.co/api/datasets/{repo_id}/upload"
-
-            while True:
-                chunk = f.read(chunk_size)
-                if not chunk:
-                    break  # Finished reading file
-
-                # Upload chunk
-                response = requests.put(upload_url, headers=headers, data=chunk)
-
-                if response.status_code != 200:
-                    raise Exception(f"Upload failed: {response.text}")
-
-                # Update progress
-                uploaded += len(chunk)
-                if progress is not None:
-                    progress(uploaded / file_size, desc=f"Uploading... {uploaded // (1024 * 1024)}MB/{file_size // (1024 * 1024)}MB")
+        # ✅ Official upload method from huggingface_hub
+        api.upload_file(
+            path_or_fileobj=file_path,
+            path_in_repo=os.path.basename(file_path),
+            repo_id=repo_id,
+            repo_type="dataset",  # Important to specify it's a dataset repo
+            token=token
+        )
 
-        # Final progress update
         if progress is not None:
             progress(1, desc="✅ Upload Complete")