deploy app, storage, readme

app.py

@@ -7,7 +7,7 @@ import html
 from pathlib import Path
 
 import pandas as pd
-import requests
+from huggingface_hub import hf_hub_download
 
 from storage import VoteStorage
 
@@ -17,37 +17,17 @@ RATINGS_APP_TOKEN = os.getenv("RATINGS_APP_TOKEN")
 POOL_REPO_ID = "taigasan/e6-visual-ratings"
 VOTE_STORAGE = VoteStorage(mode="local" if DEBUG_MODE else "hf", local_dir=LOCAL_DATA_DIR, token=RATINGS_APP_TOKEN)
 
-
-def _download_pool_parquet(repo_id: str, token: str | None) -> str:
-    root = Path(__file__).resolve().parent / LOCAL_DATA_DIR
-    root.mkdir(parents=True, exist_ok=True)
-    out_path = root / "pool.parquet"
-    tmp_path = out_path.with_suffix(".parquet.tmp")
-    url = f"https://huggingface.co/datasets/{repo_id}/resolve/main/pool.parquet"
-    headers = {"Authorization": f"Bearer {token}"} if token else {}
-    last_err = None
-    for _ in range(3):
-        try:
-            with requests.get(url, headers=headers, stream=True, timeout=(10, 60)) as r:
-                r.raise_for_status()
-                with open(tmp_path, "wb") as f:
-                    for chunk in r.iter_content(chunk_size=1 << 20):
-                        if chunk:
-                            f.write(chunk)
-            os.replace(tmp_path, out_path)
-            return str(out_path)
-        except Exception as e:
-            last_err = e
-    if out_path.exists():
-        return str(out_path)
-    raise RuntimeError(f"Failed to download pool.parquet from {repo_id}: {last_err}")
-
 # -- Pool dataset -----------------------------------------------------------
 if DEBUG_MODE:
     _pool_path = str(Path(__file__).resolve().parent / LOCAL_DATA_DIR / "pool.parquet")
     assert Path(_pool_path).exists(), f"Missing local debug pool file: {_pool_path}"
 else:
-    _pool_path = _download_pool_parquet(POOL_REPO_ID, RATINGS_APP_TOKEN)
+    _pool_path = hf_hub_download(
+        repo_id=POOL_REPO_ID,
+        filename="pool.parquet",
+        repo_type="dataset",
+        token=RATINGS_APP_TOKEN
+    )
 _pool_df = pd.read_parquet(_pool_path)
 _pool_group_dfs = {g: gdf for g, gdf in _pool_df.groupby("group")}