update2

src/display/utils.py

@@ -119,7 +119,7 @@ for task in Tasks:
     # auto_eval_column_dict.append([f"{task.name}_gpu_util", ColumnContent, ColumnContent(f"{task.value.col_name} {GPU_Util}", "number", True, hidden=True)])
     if task.value.benchmark in MULTIPLE_CHOICEs:
         continue
-    auto_eval_column_dict.append([f"{task.name}_prefilling_time", ColumnContent, ColumnContent(f"{task.value.col_name} {PREs}", "number", False, hidden=True)])
+    # auto_eval_column_dict.append([f"{task.name}_prefilling_time", ColumnContent, ColumnContent(f"{task.value.col_name} {PREs}", "number", False, hidden=True)])
     auto_eval_column_dict.append([f"{task.name}_decoding_throughput", ColumnContent, ColumnContent(f"{task.value.col_name} {TS}", "number", True, hidden=True)])
     auto_eval_column_dict.append([f"{task.name}_mbu", ColumnContent, ColumnContent(f"{task.value.col_name} {MBU}", "number", True, hidden=True)])
     auto_eval_column_dict.append([f"{task.name}_mfu", ColumnContent, ColumnContent(f"{task.value.col_name} {MFU}", "number", True, hidden=True)])

src/populate.py

@@ -75,7 +75,7 @@ def get_leaderboard_df(
             df[col] = np.nan
 
     if not df.empty:
-        df = df.map(lambda x: round(x, 2) if isinstance(x, (int, float)) else x)
+        df = df.round(decimals=2)
 
         # filter out if any of the benchmarks have not been produced
         # df = df[has_no_nan_values(df, benchmark_cols)]