Spaces:
Running
on
CPU Upgrade
Running
on
CPU Upgrade
simplify task names
Browse files- src/about.py +20 -18
src/about.py
CHANGED
|
@@ -13,23 +13,23 @@ class Task:
|
|
| 13 |
class Tasks(Enum):
|
| 14 |
# task_key in the json file, metric_key in the json file, name to display in the leaderboard
|
| 15 |
task2 = Task("belebele_pol_Latn", "acc,none", "belebele_pol_Latn")
|
| 16 |
-
task3 = Task("polemo2_in", "exact_match,score-first", "
|
| 17 |
-
task4 = Task("polemo2_in_multiple_choice", "acc,none", "
|
| 18 |
-
task5 = Task("polemo2_out", "exact_match,score-first", "
|
| 19 |
-
task6 = Task("polemo2_out_multiple_choice", "acc,none", "
|
| 20 |
-
task7 = Task("polish_8tags_multiple_choice", "acc,none", "
|
| 21 |
-
task8 = Task("polish_8tags_regex", "exact_match,score-first", "
|
| 22 |
-
task9 = Task("polish_belebele_regex", "exact_match,score-first", "
|
| 23 |
-
task10 = Task("polish_dyk_multiple_choice", "acc,none", "
|
| 24 |
-
task11 = Task("polish_dyk_regex", "exact_match,score-first", "
|
| 25 |
-
task12 = Task("polish_ppc_multiple_choice", "acc,none", "
|
| 26 |
-
task13 = Task("polish_ppc_regex", "exact_match,score-first", "
|
| 27 |
-
task14 = Task("polish_psc_multiple_choice", "acc,none", "
|
| 28 |
-
task15 = Task("polish_psc_regex", "exact_match,score-first", "
|
| 29 |
-
task16 = Task("polish_cbd_multiple_choice", "acc,none", "
|
| 30 |
-
task17 = Task("polish_cbd_regex", "exact_match,score-first", "
|
| 31 |
-
task18 = Task("polish_klej_ner_multiple_choice", "acc,none", "
|
| 32 |
-
task19 = Task("polish_klej_ner_regex", "exact_match,score-first", "
|
| 33 |
|
| 34 |
NUM_FEWSHOT = 0 # Change with your few shot
|
| 35 |
# ---------------------------------------------------
|
|
@@ -41,7 +41,9 @@ TITLE = """<h1 align="center" id="space-title">Open PL LLM Leaderboard</h1>"""
|
|
| 41 |
|
| 42 |
# What does your leaderboard evaluate?
|
| 43 |
INTRODUCTION_TEXT = """
|
| 44 |
-
|
|
|
|
|
|
|
| 45 |
"""
|
| 46 |
|
| 47 |
# Which evaluations are you running? how can people reproduce what you have?
|
|
|
|
| 13 |
class Tasks(Enum):
|
| 14 |
# task_key in the json file, metric_key in the json file, name to display in the leaderboard
|
| 15 |
task2 = Task("belebele_pol_Latn", "acc,none", "belebele_pol_Latn")
|
| 16 |
+
task3 = Task("polemo2_in", "exact_match,score-first", "polemo2-in_g")
|
| 17 |
+
task4 = Task("polemo2_in_multiple_choice", "acc,none", "polemo2_in_mc")
|
| 18 |
+
task5 = Task("polemo2_out", "exact_match,score-first", "polemo2_out_g")
|
| 19 |
+
task6 = Task("polemo2_out_multiple_choice", "acc,none", "polemo2_out_mc")
|
| 20 |
+
task7 = Task("polish_8tags_multiple_choice", "acc,none", "8tags_mc")
|
| 21 |
+
task8 = Task("polish_8tags_regex", "exact_match,score-first", "8tags_g")
|
| 22 |
+
task9 = Task("polish_belebele_regex", "exact_match,score-first", "belebele_g")
|
| 23 |
+
task10 = Task("polish_dyk_multiple_choice", "acc,none", "dyk_mc")
|
| 24 |
+
task11 = Task("polish_dyk_regex", "exact_match,score-first", "dyk_g")
|
| 25 |
+
task12 = Task("polish_ppc_multiple_choice", "acc,none", "ppc_mc")
|
| 26 |
+
task13 = Task("polish_ppc_regex", "exact_match,score-first", "ppc_g")
|
| 27 |
+
task14 = Task("polish_psc_multiple_choice", "acc,none", "psc_mc")
|
| 28 |
+
task15 = Task("polish_psc_regex", "exact_match,score-first", "psc_g")
|
| 29 |
+
task16 = Task("polish_cbd_multiple_choice", "acc,none", "cbd_mc")
|
| 30 |
+
task17 = Task("polish_cbd_regex", "exact_match,score-first", "cbd_g")
|
| 31 |
+
task18 = Task("polish_klej_ner_multiple_choice", "acc,none", "klej_ner_mc")
|
| 32 |
+
task19 = Task("polish_klej_ner_regex", "exact_match,score-first", "klej_ner_g")
|
| 33 |
|
| 34 |
NUM_FEWSHOT = 0 # Change with your few shot
|
| 35 |
# ---------------------------------------------------
|
|
|
|
| 41 |
|
| 42 |
# What does your leaderboard evaluate?
|
| 43 |
INTRODUCTION_TEXT = """
|
| 44 |
+
_g suffix means that a model needs to generate an answer (only suitable for instructions-based models)
|
| 45 |
+
|
| 46 |
+
_mc suffix means that a model is scored against every possible class (suitable also for base models)
|
| 47 |
"""
|
| 48 |
|
| 49 |
# Which evaluations are you running? how can people reproduce what you have?
|