Adding multiple models plotting and sidebar
Browse files- src/display.py +8 -6
- src/plot.py +14 -6
src/display.py
CHANGED
@@ -12,14 +12,14 @@ def display_app():
|
|
12 |
st.markdown("# Open LLM Leaderboard Viz")
|
13 |
st.markdown("This is a visualization of the results in [open-llm-leaderboard/results](https://huggingface.co/datasets/open-llm-leaderboard/results)")
|
14 |
st.markdown("To select a model, click on the checkbox beside its name.")
|
15 |
-
st.markdown("This displays the top 100 models by default, but you can change that using the number input
|
16 |
-
st.markdown("By
|
17 |
st.markdown("If your model doesn't show up, please search it by its name.")
|
18 |
|
19 |
dataframe = load_dataframe()
|
20 |
|
21 |
sort_selection = st.selectbox(label = "Sort by:", options = list(dataframe.columns), index = 7)
|
22 |
-
number_of_row = st.number_input("Number of top rows to display", min_value=100, max_value=500, value="min", step=100)
|
23 |
ascending = True
|
24 |
|
25 |
if sort_selection is None:
|
@@ -32,6 +32,8 @@ def display_app():
|
|
32 |
|
33 |
|
34 |
name = st.text_input(label = ":mag: Search by name")
|
|
|
|
|
35 |
len_name_input = len(name)
|
36 |
if len_name_input > 0:
|
37 |
dataframe_by_search = search_by_name(name)
|
@@ -55,7 +57,7 @@ def display_app():
|
|
55 |
|
56 |
#Infer basic colDefs from dataframe types
|
57 |
gb = GridOptionsBuilder.from_dataframe(dataframe_display)
|
58 |
-
gb.configure_selection(selection_mode =
|
59 |
gb.configure_grid_options(domLayout='normal')
|
60 |
gridOptions = gb.build()
|
61 |
|
@@ -77,9 +79,9 @@ def display_app():
|
|
77 |
|
78 |
with column2:
|
79 |
if grid_response['selected_rows'] is not None and len(grid_response['selected_rows']) > 0:
|
80 |
-
figure = plot_radar_chart_rows(rows=grid_response['selected_rows'])
|
81 |
#figure = plot_radar_chart_name(dataframe= dataframe, model_name=grid_response['selected_rows'][0]["model_name"])
|
82 |
-
st.plotly_chart(figure, use_container_width=
|
83 |
else:
|
84 |
if len(subdata)>0:
|
85 |
figure = plot_radar_chart_name(dataframe=subdata, model_name=model_name)
|
|
|
12 |
st.markdown("# Open LLM Leaderboard Viz")
|
13 |
st.markdown("This is a visualization of the results in [open-llm-leaderboard/results](https://huggingface.co/datasets/open-llm-leaderboard/results)")
|
14 |
st.markdown("To select a model, click on the checkbox beside its name.")
|
15 |
+
st.markdown("This displays the top 100 models by default, but you can change that using the number input in the sidebar.")
|
16 |
+
st.markdown("By default as well, the maximum number of row you can display is 500, it is due to the problem with st_aggrid component loading.")
|
17 |
st.markdown("If your model doesn't show up, please search it by its name.")
|
18 |
|
19 |
dataframe = load_dataframe()
|
20 |
|
21 |
sort_selection = st.selectbox(label = "Sort by:", options = list(dataframe.columns), index = 7)
|
22 |
+
number_of_row = st.sidebar.number_input("Number of top rows to display", min_value=100, max_value=500, value="min", step=100)
|
23 |
ascending = True
|
24 |
|
25 |
if sort_selection is None:
|
|
|
32 |
|
33 |
|
34 |
name = st.text_input(label = ":mag: Search by name")
|
35 |
+
selection_mode = st.sidebar.radio(label= "Selection mode for the rows", options = ["single", "multiple"], index=0)
|
36 |
+
st.sidebar.write("In multiple mode, you can select up to three models. If you select more than three models, only the first three will be displayed and plotted.")
|
37 |
len_name_input = len(name)
|
38 |
if len_name_input > 0:
|
39 |
dataframe_by_search = search_by_name(name)
|
|
|
57 |
|
58 |
#Infer basic colDefs from dataframe types
|
59 |
gb = GridOptionsBuilder.from_dataframe(dataframe_display)
|
60 |
+
gb.configure_selection(selection_mode = selection_mode, use_checkbox=True)
|
61 |
gb.configure_grid_options(domLayout='normal')
|
62 |
gridOptions = gb.build()
|
63 |
|
|
|
79 |
|
80 |
with column2:
|
81 |
if grid_response['selected_rows'] is not None and len(grid_response['selected_rows']) > 0:
|
82 |
+
figure = plot_radar_chart_rows(rows=grid_response['selected_rows'][:3])
|
83 |
#figure = plot_radar_chart_name(dataframe= dataframe, model_name=grid_response['selected_rows'][0]["model_name"])
|
84 |
+
st.plotly_chart(figure, use_container_width=False)
|
85 |
else:
|
86 |
if len(subdata)>0:
|
87 |
figure = plot_radar_chart_name(dataframe=subdata, model_name=model_name)
|
src/plot.py
CHANGED
@@ -7,6 +7,8 @@ import pandas as pd
|
|
7 |
fillcolor = "#FFD21E"
|
8 |
line_color = "#FF9D00"
|
9 |
|
|
|
|
|
10 |
# opacity of the plot
|
11 |
opacity = 0.75
|
12 |
|
@@ -109,7 +111,7 @@ def plot_radar_chart_name(dataframe: pd.DataFrame, model_name: str, categories:
|
|
109 |
|
110 |
|
111 |
#@st.cache_data
|
112 |
-
def plot_radar_chart_rows(rows: object, columns:list = columns, categories: list = categories,
|
113 |
"""
|
114 |
plot the results of the model selected by the checkbox
|
115 |
|
@@ -124,22 +126,27 @@ def plot_radar_chart_rows(rows: object, columns:list = columns, categories: list
|
|
124 |
dataset = pd.DataFrame(rows, columns=columns)
|
125 |
data = dataset[categories].to_numpy()
|
126 |
data = data.astype(float)
|
|
|
|
|
|
|
|
|
127 |
|
128 |
# add data to close the area of the radar chart
|
129 |
data = np.append(data, data[:,0].reshape((-1,1)), axis=1)
|
130 |
categories_theta = categories.copy()
|
131 |
categories_theta.append(categories[0])
|
132 |
|
133 |
-
|
134 |
for i in range(len(dataset)):
|
135 |
-
|
|
|
136 |
fig.add_trace(go.Scatterpolar(
|
137 |
r=data[i,:],
|
138 |
theta=categories_theta,
|
139 |
fill='toself',
|
140 |
-
fillcolor =
|
141 |
opacity = opacity,
|
142 |
-
line=dict(color =
|
143 |
name= dataset.loc[i,"model_name"]
|
144 |
))
|
145 |
fig.update_layout(
|
@@ -148,7 +155,8 @@ def plot_radar_chart_rows(rows: object, columns:list = columns, categories: list
|
|
148 |
visible=True,
|
149 |
range=[0, 100.]
|
150 |
)),
|
151 |
-
showlegend=
|
152 |
)
|
|
|
153 |
|
154 |
return fig
|
|
|
7 |
fillcolor = "#FFD21E"
|
8 |
line_color = "#FF9D00"
|
9 |
|
10 |
+
fill_color_list = [fillcolor, "#F05998", "#40BAF0"]
|
11 |
+
line_color_list = [line_color, "#5E233C", "#194A5E"]
|
12 |
# opacity of the plot
|
13 |
opacity = 0.75
|
14 |
|
|
|
111 |
|
112 |
|
113 |
#@st.cache_data
|
114 |
+
def plot_radar_chart_rows(rows: object, columns:list = columns, categories: list = categories, fillcolor_list: str = fill_color_list, line_color_list:str = line_color_list):
|
115 |
"""
|
116 |
plot the results of the model selected by the checkbox
|
117 |
|
|
|
126 |
dataset = pd.DataFrame(rows, columns=columns)
|
127 |
data = dataset[categories].to_numpy()
|
128 |
data = data.astype(float)
|
129 |
+
showLegend = False
|
130 |
+
if len(rows) > 1:
|
131 |
+
showLegend = True
|
132 |
+
|
133 |
|
134 |
# add data to close the area of the radar chart
|
135 |
data = np.append(data, data[:,0].reshape((-1,1)), axis=1)
|
136 |
categories_theta = categories.copy()
|
137 |
categories_theta.append(categories[0])
|
138 |
|
139 |
+
opacity = 0.75
|
140 |
for i in range(len(dataset)):
|
141 |
+
colors = fillcolor_list[i]
|
142 |
+
|
143 |
fig.add_trace(go.Scatterpolar(
|
144 |
r=data[i,:],
|
145 |
theta=categories_theta,
|
146 |
fill='toself',
|
147 |
+
fillcolor = colors,
|
148 |
opacity = opacity,
|
149 |
+
line=dict(color = line_color_list[i]),
|
150 |
name= dataset.loc[i,"model_name"]
|
151 |
))
|
152 |
fig.update_layout(
|
|
|
155 |
visible=True,
|
156 |
range=[0, 100.]
|
157 |
)),
|
158 |
+
showlegend=showLegend
|
159 |
)
|
160 |
+
opacity -= .2
|
161 |
|
162 |
return fig
|