added table

app.py

@@ -0,0 +1,43 @@
+import gradio as gr
+import pandas as pd
+
+mase = pd.read_csv("results/results_mase.csv")
+datasets = mase.dataset.unique()
+frameworks = mase.framework.unique()
+mase.set_index(["dataset", "framework"], inplace=True)
+
+data = {"Dataset": datasets}
+
+
+def mean(data, framework):
+    try:
+        return f"{round(mase.loc[data, framework].metric_error.mean(),3)} +/- {round(mase.loc[data, framework].metric_error.std(),3)}"
+    except KeyError as e:
+        return "n/a"
+
+
+for framework in frameworks:
+    data.update({framework: mean(data, framework) for data in datasets})
+
+df = pd.DataFrame(data=data)
+table = df.to_markdown()
+
+
+with gr.Blocks() as demo:
+    gr.Markdown(
+        f"""
+    # Time Series Forecasting Leaderboard
+    
+    This is a leaderboard of the MASE metric for time series forecasting problem on the different open datasets and models.
+
+    The table is generated from the paper [AutoGluon–TimeSeries: AutoML for Probabilistic Time Series Forecasting](https://github.com/autogluon/autogluon) by 
+Oleksandr Shchur, Caner Turkmen, Nick Erickson, Huibin Shen, Alexander Shirkov, Tony Hu, and Bernie Wang.
+
+    ## MASE Metric
+
+    {table}
+    """
+    )
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,2 @@
+tabulate
+pandas