Deploy maintained LiveHouse-TS application and public results

README.md

@@ -1,5 +1,5 @@
 ---
-title: TSFM Realworld Bench
+title: LiveHouse-TS
 emoji: 📈
 colorFrom: blue
 colorTo: green
@@ -11,11 +11,11 @@ license: apache-2.0
 python_version: 3.11.9
 ---
 
-# TSFM Realworld Bench Leaderboard
+# LiveHouse-TS Leaderboard
 
-Leaderboard for the **TSFM Realworld Benchmark** — evaluating time series foundation models on real-world univariate-target forecasting tasks.
+**LiveHouse-TS** evaluates time series foundation models on real-world univariate-target forecasting tasks.
 
-Results are submitted via pull request to the [TS-Live](https://github.com/zhouziyu02/TS-Live) repository.
+Results are submitted via pull request to the [TS-Live](https://github.com/Thinkcat-Lab/LiveHouse-TS) repository.
 
 The leaderboard includes GIFT-Eval-style result slices aggregated by actual prediction
 length, domain, and frequency, normalized per configuration against Seasonal-Naive.

app.py

@@ -6,6 +6,10 @@ from pathlib import Path
 
 import gradio as gr
 
+# Gradio checks its own HTTP endpoint at startup; keep loopback off outbound proxies.
+os.environ["NO_PROXY"] = ",".join(filter(None, [os.getenv("NO_PROXY", os.getenv("no_proxy", "")),
+                                               "localhost,127.0.0.1,0.0.0.0"]))
+
 SPACE_DIR = Path(__file__).resolve().parent
 if str(SPACE_DIR) not in sys.path:
     sys.path.insert(0, str(SPACE_DIR))
@@ -34,7 +38,7 @@ from src.utils import (
     prepare_values_table,
 )
 
-RESULTS_PATH = os.getenv("TSFM_RESULTS_PATH", "space/results" if Path("space/results").exists() else "results")
+RESULTS_PATH = os.getenv("TSFM_RESULTS_PATH", str(SPACE_DIR / "results"))
 REFRESH_SECONDS = int(os.getenv("TSFM_REFRESH_SECONDS", "300"))
 DOMAIN_OUTPUT_OFFSET = 14
 MODEL_COLUMN_WIDTH = "180px"
@@ -257,7 +261,9 @@ def _make_forecast_plotter(dataset_key: str):
 
 def refresh_leaderboard() -> tuple:
     grouped = get_grouped_dfs(RESULTS_PATH)
-    datasets = grouped["datasets"]
+    # Gradio outputs are fixed at startup; missing/new datasets must not change
+    # callback arity or shift another dataset's values into an existing table.
+    datasets = DATASETS
     one_day_rank, one_day_metrics = _fixed_window_outputs("1d")
     seven_day_rank, seven_day_metrics = _fixed_window_outputs("7d")
     thirty_day_rank, thirty_day_metrics = _fixed_window_outputs("30d")

constraints.txt

@@ -0,0 +1,72 @@
+# Tested deployment dependency constraints, generated 2026-09-05 on Python 3.11.
+# Regenerate in a clean environment after running CI and smoke tests.
+aiofiles==23.2.1
+annotated-doc==0.0.5
+annotated-types==0.8.0
+anyio==4.15.0
+certifi==2026.7.22
+charset-normalizer==3.5.1
+click==8.5.0
+contourpy==1.3.3; python_version >= "3.11"
+contourpy==1.3.2; python_version < "3.11"
+cycler==0.12.1
+fastapi==0.112.4
+ffmpy==1.0.0
+filelock==3.32.5
+fonttools==4.64.0
+fsspec==2026.7.0
+gluonts==0.16.3
+gradio==4.44.0
+gradio_client==1.3.0
+h11==0.16.0
+hf-xet==1.6.0
+httpcore==1.0.9
+httpx==0.28.1
+huggingface-hub==0.35.2
+idna==3.19
+importlib_resources==6.5.2
+iniconfig==2.3.0
+Jinja2==3.1.6
+kiwisolver==1.5.1
+markdown-it-py==4.2.0
+MarkupSafe==2.1.5
+matplotlib==3.9.4
+mdurl==0.1.2
+numpy==1.26.4
+orjson==3.12.0
+packaging==26.3
+pandas==2.2.3
+patsy==1.0.3
+pillow==10.4.0
+plotly==5.24.1
+pluggy==1.6.0
+pydantic==2.10.6
+pydantic_core==2.27.2
+pydub==0.25.1
+Pygments==2.21.0
+pyparsing==3.3.2
+pytest==8.4.2
+python-dateutil==2.9.0.post0
+python-dotenv==1.2.3
+python-multipart==0.0.32
+pytz==2026.3.post1
+PyYAML==6.0.3
+requests==2.34.2
+rich==15.0.0
+ruff==0.16.6
+scipy==1.11.4
+semantic-version==2.10.0
+shellingham==1.5.4
+six==1.17.0
+starlette==0.37.2
+statsmodels==0.14.6
+tenacity==9.1.4
+tomlkit==0.12.0
+toolz==0.12.1
+tqdm==4.70.0
+typer==0.27.2
+typing_extensions==4.16.0
+tzdata==2026.3
+urllib3==2.7.0
+uvicorn==0.30.6
+websockets==12.0

requirements.txt

@@ -1,10 +1,13 @@
-# gradio is installed by HF Space from README sdk_version — do NOT pin here
-huggingface-hub>=0.18.0,<0.35.3
+-c constraints.txt
+gradio==4.44.0
+huggingface-hub>=0.23,<0.35.3
+pydantic>=2.7,<2.11
 starlette==0.37.2
 fastapi==0.112.4
-jinja2==3.1.4
+jinja2>=3.1.6,<4
 uvicorn==0.30.6
-pandas
-numpy
-scipy
-plotly
+pandas>=2.2,<2.3
+numpy>=1.26,<2
+scipy>=1.11,<1.12
+plotly>=5,<6
+matplotlib>=3.9,<3.10

src/about.py

@@ -1,7 +1,7 @@
-TITLE = """ TSFM Realworld Bench Leaderboard """
+TITLE = """ LiveHouse-TS Leaderboard """
 
 INTRODUCTION_TEXT = """
-**TSFM Realworld Bench** evaluates time series foundation models on **live TS-Bench
+**LiveHouse-TS** evaluates time series foundation models on **live TS-Bench
 real-world data** with **zero-shot API inference** via [TSFM.ai](https://tsfm.ai/).
 Following [GIFT-Eval](https://huggingface.co/spaces/Salesforce/GIFT-Eval), the leaderboard
 reports **absolute metric values** and **per-dataset ranks**. The Overall tab also reports
@@ -55,9 +55,9 @@ on future live releases.
 CITATION_BUTTON_LABEL = "Copy citation"
 CITATION_BUTTON_TEXT = r"""
 @misc{tsfm_realworld_bench,
-  title={TSFM Realworld Bench: A Benchmark for Time Series Foundation Models},
-  author={TSFM Realworld Bench Team},
+  title={LiveHouse-TS: A Live Benchmark for Time Series Foundation Models},
+  author={LiveHouse-TS Team},
   year={2026},
-  howpublished={\url{https://huggingface.co/spaces/CityMindDev/TSFM-Realworld-Bench}}
+  howpublished={\url{https://huggingface.co/spaces/CityMindDev/LiveHouse-TS}}
 }
 """

src/display/css_html_js.py

@@ -1,4 +1,4 @@
-# Gradio custom styles for the TSFM Realworld Bench leaderboard.
+# Gradio custom styles for the LiveHouse-TS leaderboard.
 custom_css = """
 :root {
     --bench-ink: #18202f;

src/utils.py

@@ -261,7 +261,7 @@ def load_all_results(root_dir: str = "results") -> pd.DataFrame:
         df = df[df["model"].map(_model_key).isin(model_allowlist)].copy()
         if df.empty:
             return pd.DataFrame(columns=RESULT_COLUMNS)
-    
+
     parts = df["dataset"].astype(str).str.split("/", expand=True)
     if parts.shape[1] < 3:
         for idx in range(parts.shape[1], 3):