Spaces:
Running
Running
Yotam-Perlitz
commited on
Commit
β’
386e6e6
1
Parent(s):
ad12749
fix location of holistic benchmarks list
Browse filesSigned-off-by: Yotam-Perlitz <y.perlitz@ibm.com>
app.py
CHANGED
@@ -5,7 +5,28 @@ import pandas as pd
|
|
5 |
import plotly.express as px
|
6 |
import streamlit as st
|
7 |
from bat import Benchmark, Config, Reporter, Tester
|
8 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
9 |
|
10 |
|
11 |
def get_nice_benchmark_name(bench_name):
|
@@ -49,7 +70,7 @@ st.markdown(
|
|
49 |
)
|
50 |
|
51 |
|
52 |
-
all_scenarios_for_aggragate =
|
53 |
|
54 |
st.subheader("The Leaderboard", divider=True)
|
55 |
# st.subheader("ποΈββοΈ BenchBench Leaderboard π", divider=True)
|
@@ -157,27 +178,6 @@ def run_load(
|
|
157 |
n_exps=n_exps if n_models_taken_list != [0] else 1,
|
158 |
)
|
159 |
|
160 |
-
holistic_scenarios = [
|
161 |
-
"arena_hard",
|
162 |
-
"mixeval",
|
163 |
-
"agieval",
|
164 |
-
"arc_c",
|
165 |
-
"alpacav1",
|
166 |
-
"alpacav2",
|
167 |
-
"alpacaeval2_lc",
|
168 |
-
"arena_elo",
|
169 |
-
"bbh",
|
170 |
-
"eq_benchv2",
|
171 |
-
"gpt4all",
|
172 |
-
"hugging_6",
|
173 |
-
"llmonitor",
|
174 |
-
"magi",
|
175 |
-
"mmlu",
|
176 |
-
"mt_bench",
|
177 |
-
"biggen_mwr",
|
178 |
-
"olmes_average",
|
179 |
-
"mmlu_pro",
|
180 |
-
]
|
181 |
holistic = Benchmark()
|
182 |
holistic.load_local_catalog()
|
183 |
holistic.df = holistic.df.query("scenario in @holistic_scenarios")
|
|
|
5 |
import plotly.express as px
|
6 |
import streamlit as st
|
7 |
from bat import Benchmark, Config, Reporter, Tester
|
8 |
+
|
9 |
+
holistic_scenarios = [
|
10 |
+
"arena_hard",
|
11 |
+
"mixeval",
|
12 |
+
"agieval",
|
13 |
+
"arc_c",
|
14 |
+
"alpacav1",
|
15 |
+
"alpacav2",
|
16 |
+
"alpacaeval2_lc",
|
17 |
+
"arena_elo",
|
18 |
+
"bbh",
|
19 |
+
"eq_benchv2",
|
20 |
+
"gpt4all",
|
21 |
+
"hugging_6",
|
22 |
+
"llmonitor",
|
23 |
+
"magi",
|
24 |
+
"mmlu",
|
25 |
+
"mt_bench",
|
26 |
+
"biggen_mwr",
|
27 |
+
"olmes_average",
|
28 |
+
"mmlu_pro",
|
29 |
+
]
|
30 |
|
31 |
|
32 |
def get_nice_benchmark_name(bench_name):
|
|
|
70 |
)
|
71 |
|
72 |
|
73 |
+
all_scenarios_for_aggragate = holistic_scenarios
|
74 |
|
75 |
st.subheader("The Leaderboard", divider=True)
|
76 |
# st.subheader("ποΈββοΈ BenchBench Leaderboard π", divider=True)
|
|
|
178 |
n_exps=n_exps if n_models_taken_list != [0] else 1,
|
179 |
)
|
180 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
181 |
holistic = Benchmark()
|
182 |
holistic.load_local_catalog()
|
183 |
holistic.df = holistic.df.query("scenario in @holistic_scenarios")
|