Initial commit

Dockerfile

@@ -5,6 +5,7 @@ WORKDIR /app
 RUN useradd -m -u 1000 user
 USER user
 ENV PATH="/home/user/.local/bin:$PATH"
+ENV PYTHONUNBUFFERED=1
 
 COPY --chown=user requirements.txt requirements.txt
 RUN pip install --no-cache-dir --upgrade pip && \
@@ -12,8 +13,4 @@ RUN pip install --no-cache-dir --upgrade pip && \
 
 COPY --chown=user . .
 
-ENV DATA_DIR=/data
-
-EXPOSE 7860
-
-CMD ["python", "app.py"]
+CMD ["python", "app.py"]

app.py

@@ -13,15 +13,12 @@ from huggingface_hub import HfApi, hf_hub_download
 
 DATASET_REPO_ID = os.environ.get("DATASET_REPO_ID", "fast-stager/property-labels") 
 HF_TOKEN = os.environ.get("HF_TOKEN") 
-
 CACHE_DIR = "/tmp"
 URL_FILE = "urls.txt" 
-
 LABEL_FILE = os.path.join(CACHE_DIR, "annotations.csv")
 VERIFY_FILE = os.path.join(CACHE_DIR, "verifications.csv")
 SKIP_FILE = os.path.join(CACHE_DIR, "skipped.csv")
 LOCK_FILE = os.path.join(CACHE_DIR, "data.lock")
-
 MAX_IMAGES = 6
 THUMB_SIZE = (350, 350)
 ROOM_CLASSES = ["living_room", "bedroom", "kitchen", "bathroom", "dining_room", "outdoor", "other"]
@@ -32,7 +29,6 @@ def sync_pull():
         return
 
     print(f"🔄 Syncing from {DATASET_REPO_ID}...")
-    
     for filename in ["annotations.csv", "verifications.csv", "skipped.csv"]:
         try:
             hf_hub_download(
@@ -44,7 +40,7 @@ def sync_pull():
             )
             print(f"✅ Loaded {filename}")
         except Exception:
-            print(f"ℹ️ {filename} not found on Hub. Creating empty in /tmp.")
+            print(f"ℹ️ {filename} missing on Hub. Starting fresh.")
 
 def sync_push_background(local_path, remote_filename):
     if not HF_TOKEN: return
@@ -68,7 +64,6 @@ def sync_push_background(local_path, remote_filename):
 
 def init_files():
     sync_pull()
-
     for f in [LABEL_FILE, VERIFY_FILE, SKIP_FILE]:
         if not os.path.exists(f):
             if f == LABEL_FILE: cols = ["timestamp", "user", "group_id", "url", "score", "label"]
@@ -77,18 +72,14 @@ def init_files():
             pd.DataFrame(columns=cols).to_csv(f, index=False)
             
     if not os.path.exists(URL_FILE):
-        print("⚠️ urls.txt not found in root directory! Please upload it to your Space.")
-    else:
-        print(f"✅ urls.txt found ({len(open(URL_FILE).readlines())} lines)")
+        print("⚠️ urls.txt not found! Please upload it to the repo.")
 
 init_files()
 
 def get_image_optimized(url):
     if not url: return None
     try:
-        headers = {
-            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36'
-        }
+        headers = {'User-Agent': 'Mozilla/5.0'}
         response = requests.get(url, headers=headers, timeout=3)
         if response.status_code == 200:
             img = Image.open(BytesIO(response.content))
@@ -174,7 +165,6 @@ def render_workspace(mode, history, specific_index=None, move_back=False):
         found = False
         for i, gid in enumerate(all_groups):
             if gid in s_done: continue
-            
             is_ready = False
             if mode == "label" and gid not in l_done: is_ready = True
             elif mode == "verify" and gid in l_done and gid not in v_done: is_ready = True
@@ -223,7 +213,6 @@ def render_workspace(mode, history, specific_index=None, move_back=False):
         if i < len(urls):
             u = urls[i]
             img_data = processed_images[i]
-            
             updates[img_c] = gr.update(value=img_data, visible=True)
             
             v_sc = saved_vals.get(u, {}).get('score', 5)
@@ -313,7 +302,7 @@ with gr.Blocks(theme=gr.themes.Soft(), title="Labeling Tool") as demo:
         with gr.Tab("Workspace", id=0):
             with gr.Group() as screen_menu:
                 gr.Markdown("# Welcome! 👋")
-                gr.Markdown("Connected to HF Dataset for free persistence.")
+                gr.Markdown("Connected to HF Dataset for persistence.")
                 with gr.Row():
                     b_start_l = gr.Button("Start Labeling", variant="primary")
                     b_start_v = gr.Button("Start Verification")

requirements.txt

@@ -1,6 +1,6 @@
-huggingface-hub==0.34.3
-gradio==5.0.0
+gradio==4.44.1
 pandas
 requests
 Pillow
-filelock
+filelock
+huggingface_hub>=0.25.0