simplify datastream to deal with load

app.py

@@ -6,7 +6,6 @@ import tempfile
 
 import gradio as gr
 from PIL import Image, ImageDraw, ImageFont
-from huggingface_hub import whoami, upload_file
 
 from criteria import check_certification as check_certification_criteria
 from org import join_finishers_org
@@ -53,23 +52,6 @@ def generate_certificate(
 
     return im, "certificate.pdf"
 
-
-def get_user_info(oauth_token):
-    """Get user info from HF token."""
-    if oauth_token is None:
-        return None, None, None, None
-    try:
-        user_info = whoami(oauth_token.token)
-        username = user_info["name"]
-        name_parts = user_info["fullname"].split(" ", 1)
-        first_name = name_parts[0]
-        last_name = name_parts[1] if len(name_parts) > 1 else ""
-        profile_url = user_info["avatarUrl"]
-        return username, first_name, last_name, profile_url
-    except:
-        return None, None, None, None
-
-
 def create_linkedin_button(username: str, cert_url: str | None) -> str:
     """Create LinkedIn 'Add to Profile' button HTML."""
     current_year = date.today().year
@@ -149,13 +131,20 @@ def upload_certificate_to_hub(username: str, certificate_img) -> str:
             return None
 
 
-def check_certification(token: gr.OAuthToken | None):
+def check_certification(token: gr.OAuthToken | None, profile: gr.OAuthProfile | None):
     """Check certification status for logged-in user."""
-    if token is None:
+    if token is None or profile is None:
         gr.Warning("Please log in to Hugging Face before checking certification!")
         return None, None, None, gr.Row.update(visible=False)
 
-    username, first_name, last_name, profile_url = get_user_info(token)
+    username = profile.username
+    first_name = profile.name.split(sep=" ")[0]
+    try:
+        last_name = " ".join(profile.name.split(sep=" ")[1:])
+    except IndexError:
+        last_name = ""
+    profile_url = profile.picture
+    
     if not username:
         return (
             "Please login with your Hugging Face account to check certification status",
@@ -194,39 +183,35 @@ def check_certification(token: gr.OAuthToken | None):
     )
 
 
-def create_gradio_interface():
-    """Create Gradio web interface with OAuth login."""
-    with gr.Blocks() as demo:
-        gr.Markdown("""
-        # Get your Hugging Face Course Certificate 🎓
-        The certification process is completely free.
-        
-        To receive your certificate, you need to **pass 80% of the quiz**.
-
-        There's **no deadlines, the course is self-paced**.
+with gr.Blocks() as demo:
+    gr.Markdown("""
+    # Get your Hugging Face Course Certificate 🎓
+    The certification process is completely free.
+    
+    To receive your certificate, you need to **pass 80% of the quiz**.
 
-        Don't hesitate to share your certificate on Twitter 
-        (tag @huggingface) and on Linkedin.
-        """)
+    There's **no deadlines, the course is self-paced**.
 
-        # Add login button
-        gr.LoginButton()
+    Don't hesitate to share your certificate on Twitter 
+    (tag @huggingface) and on Linkedin.
+    """)
 
-        check_progress_button = gr.Button(value="Check My Progress")
+    # Add login button
+    gr.LoginButton()
 
-        output_text = gr.Markdown(visible=False, sanitize_html=False)
-        output_img = gr.Image(type="pil", visible=False)
+    check_progress_button = gr.Button(value="Check My Progress")
 
-        check_progress_button.click(
-            fn=check_certification,
-            outputs=[output_text, output_img],
-        ).then(
-            fn=join_finishers_org,
-        )
+    output_text = gr.Markdown(visible=False, sanitize_html=False)
+    output_img = gr.Image(type="pil", visible=False)
+    
+    check_progress_button.click(
+        fn=check_certification,
+        outputs=[output_text, output_img],
+    ).then(
+        fn=join_finishers_org,
+    )
 
-    return demo
 
 
 if __name__ == "__main__":
-    demo = create_gradio_interface()
     demo.launch(debug=True)

criteria.py

@@ -3,7 +3,6 @@ from typing import Dict, List, Optional, NamedTuple, Tuple
 from datetime import datetime
 
 from datasets import load_dataset
-from huggingface_hub import HfApi
 
 # Environment variables and constants
 EXAM_DATASET_ID = os.getenv(
@@ -37,9 +36,12 @@ def get_user_results(username: str) -> List[Dict]:
         List of user's quiz results
     """
     try:
-        user_results = load_dataset(path=EXAM_DATASET_ID, split=username)
-
-        results = user_results.to_list()
+        user_results = load_dataset(
+            path=EXAM_DATASET_ID,
+            streaming=True,
+            data_files=f"data/{username}-00000-of-00001.parquet",
+        )
+        results = list(user_results["train"])
         print(f"Found {len(results)} results for user {username}")
         return results
 
@@ -85,9 +87,11 @@ def calculate_pass_percentage(results: List[Dict]) -> Tuple[float, int]:
         # Find the best attempt
         best_attempt = max(
             attempts.values(),
-            key=lambda x: x["grade"]
-            if x["grade"] is not None
-            else (x["correct"] / x["total"] if x["total"] > 0 else 0),
+            key=lambda x: (
+                x["grade"]
+                if x["grade"] is not None
+                else (x["correct"] / x["total"] if x["total"] > 0 else 0)
+            ),
         )
 
         # If grade is available, use it; otherwise calculate from correct/total