Spaces:
Running
Running
File size: 7,421 Bytes
1a86feb 6de2b81 1a86feb 6de2b81 1a86feb 6de2b81 1a86feb 6de2b81 1a86feb 6de2b81 1a86feb 6de2b81 1a86feb 6de2b81 1a86feb 6de2b81 1a86feb 6de2b81 1a86feb 6de2b81 1a86feb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 |
import json
import gradio as gr
import pandas as pd
from css_html import custom_css
from text_content import ABOUT_TEXT, CITATION_BUTTON_TEXT, CITATION_BUTTON_LABEL, ACKNOWLEDGEMENT_TEXT, NOTES_TEXT, HEAD_TEXT
from utils import (
AutoEvalColumn,
fields,
lang_map,
)
result_path = './results.json'
task_type = ["input reasoning", "output reasoning"]
cur_task = "input"
next_task = "output"
with open(result_path, 'r') as f:
data = json.load(f)
rows = []
for model_name, sub_col in data.items():
row = {}
for lang in sub_col["pass@1"]:
if cur_task in lang:
row[lang_map[lang.replace(f"_{cur_task}", "")]] = sub_col["pass@1"][lang]
row['Average'] = sum(row.values()) / len(row.values())
row['Average'] = round(row['Average'], 1)
row['Model'] = model_name
row['Size'] = sub_col['size']
rows.append(row)
df = pd.DataFrame(rows)
df = df.sort_values(by='Average', ascending=False)
rows = []
for model_name, sub_col in data.items():
row = {}
for lang in sub_col["pass@1"]:
if next_task in lang:
row[lang_map[lang.replace(f"_{next_task}", "")]] = sub_col["pass@1"][lang]
row['Average'] = sum(row.values()) / len(row.values())
row['Average'] = round(row['Average'], 1)
row['Model'] = model_name
row['Size'] = sub_col['size']
rows.append(row)
df_next = pd.DataFrame(rows)
df_next = df_next.sort_values(by='Average', ascending=False)
COLS = [c.name for c in fields(AutoEvalColumn) if not c.hidden]
TYPES = [c.type for c in fields(AutoEvalColumn) if not c.hidden]
COLS_LITE = [
c.name for c in fields(AutoEvalColumn) if c.displayed_by_default and not c.hidden
]
TYPES_LITE = [
c.type for c in fields(AutoEvalColumn) if c.displayed_by_default and not c.hidden
]
def select_columns(df, columns):
always_here_cols = [
AutoEvalColumn.model.name,
AutoEvalColumn.size.name,
]
# We use COLS to maintain sorting
filtered_df = df[
always_here_cols + [c for c in COLS if c in df.columns and c in columns]
]
return filtered_df
def select_tasks(df, columns, df_next):
always_here_cols = [
AutoEvalColumn.model.name,
AutoEvalColumn.size.name,
]
df,df_next = df_next,df
filtered_df = df[
always_here_cols + [c for c in COLS if c in df.columns and c in columns]
]
return df,filtered_df,df_next
demo = gr.Blocks(css=custom_css)
with demo:
with gr.Column():
gr.Markdown(
"""<div style="text-align: center;"><h1>CRUXEVAL-X Leaderboard</h1></div>\
<br>\
""",
elem_classes="markdown-text",
)
gr.Markdown(HEAD_TEXT, elem_classes="markdown-text")
with gr.Tabs(elem_classes="tab-buttons") as tabs:
with gr.Column():
with gr.Tabs(elem_classes="A100-tabs") as A100_tabs:
with gr.TabItem("🔍 Evaluation Table", id=0):
with gr.Column():
with gr.Accordion("⏬ Tasks", open=True):
shown_tasks = gr.Radio(
choices=[
c
for c in task_type
],
value=[
c
for c in task_type
if cur_task in c
][0] if any(cur_task in c for c in task_type) else None,
label="",
elem_id="task-select",
interactive=True,
)
with gr.Accordion("⏬ Languages", open=True):
shown_languages = gr.CheckboxGroup(
choices=[
c
for c in COLS
if c
not in [
AutoEvalColumn.model.name,
AutoEvalColumn.size.name
]
],
value=[
c
for c in COLS_LITE
if c
not in [
AutoEvalColumn.model.name,
AutoEvalColumn.size.name
]
],
label="",
elem_id="column-select",
interactive=True,
)
leaderboard_df = gr.components.Dataframe(
value=df[
[
AutoEvalColumn.model.name,
AutoEvalColumn.size.name,
]
+ shown_languages.value
],
headers=COLS,
datatype=TYPES,
elem_id="leaderboard-table",
interactive=False,
)
hidden_leaderboard_df = gr.components.Dataframe(
value=df,
headers=COLS,
datatype=["str" for _ in range(len(COLS))],
visible=False,
)
leaderboard_next = gr.components.Dataframe(
value=df_next,
headers=COLS,
datatype=["str" for _ in range(len(COLS))],
visible=False,
)
shown_languages.change(
select_columns,
[hidden_leaderboard_df, shown_languages],
leaderboard_df,
)
shown_tasks.change(
select_tasks,
[hidden_leaderboard_df, shown_languages, leaderboard_next],
[hidden_leaderboard_df, leaderboard_df, leaderboard_next],
)
gr.Markdown(NOTES_TEXT, elem_classes="markdown-text")
with gr.TabItem("📝 About", id=1):
gr.Markdown(ABOUT_TEXT, elem_classes="markdown-text")
with gr.Row():
with gr.Accordion("📙 Citation", open=False):
citation_button = gr.Textbox(
value=CITATION_BUTTON_TEXT,
label=CITATION_BUTTON_LABEL,
lines=10,
elem_id="citation-button",
show_copy_button=True,
)
with gr.Row():
with gr.Accordion("🙏 Acknowledgement", open=False):
gr.Markdown(ACKNOWLEDGEMENT_TEXT)
demo.launch() |