Update space
Browse files- app.py +20 -1
- src/leaderboard/read_evals.py +1 -0
app.py
CHANGED
@@ -97,6 +97,25 @@ def init_leaderboard(dataframe):
|
|
97 |
)
|
98 |
|
99 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
100 |
demo = gr.Blocks(css=custom_css)
|
101 |
with demo:
|
102 |
gr.HTML(TITLE)
|
@@ -104,7 +123,7 @@ with demo:
|
|
104 |
|
105 |
with gr.Tabs(elem_classes="tab-buttons") as tabs:
|
106 |
with gr.TabItem("🏅 Overall", elem_id="llm-benchmark-tab-table", id=0):
|
107 |
-
leaderboard =
|
108 |
|
109 |
with gr.TabItem("🔢 Math", elem_id="math-tab-table", id=1):
|
110 |
|
|
|
97 |
)
|
98 |
|
99 |
|
100 |
+
|
101 |
+
def overall_leaderboard(dataframe):
|
102 |
+
if dataframe is None or dataframe.empty:
|
103 |
+
raise ValueError("Leaderboard DataFrame is empty or None.")
|
104 |
+
|
105 |
+
return Leaderboard(
|
106 |
+
value=dataframe,
|
107 |
+
datatype=[c.type for c in fields(AutoEvalColumn)],
|
108 |
+
select_columns=None,
|
109 |
+
search_columns=SearchColumns(primary_column=AutoEvalColumn.model.name, secondary_columns=[],
|
110 |
+
placeholder="Search by the model name",
|
111 |
+
label="Searching"),
|
112 |
+
hide_columns=[c.name for c in fields(AutoEvalColumn) if c.hidden],
|
113 |
+
filter_columns=None,
|
114 |
+
interactive=False,
|
115 |
+
)
|
116 |
+
|
117 |
+
|
118 |
+
|
119 |
demo = gr.Blocks(css=custom_css)
|
120 |
with demo:
|
121 |
gr.HTML(TITLE)
|
|
|
123 |
|
124 |
with gr.Tabs(elem_classes="tab-buttons") as tabs:
|
125 |
with gr.TabItem("🏅 Overall", elem_id="llm-benchmark-tab-table", id=0):
|
126 |
+
leaderboard = overall_leaderboard(LEADERBOARD_DF)
|
127 |
|
128 |
with gr.TabItem("🔢 Math", elem_id="math-tab-table", id=1):
|
129 |
|
src/leaderboard/read_evals.py
CHANGED
@@ -194,3 +194,4 @@ def get_raw_eval_results(results_path: str, requests_path: str) -> list[EvalResu
|
|
194 |
continue
|
195 |
|
196 |
return results
|
|
|
|
194 |
continue
|
195 |
|
196 |
return results
|
197 |
+
|