Unverified Commit 56a4e794 authored by Lintang Sutawika's avatar Lintang Sutawika Committed by GitHub
Browse files

formatting (#2104)

parent 9884ad6e
......@@ -2,51 +2,74 @@ import argparse
import yaml
languages = ['eng', 'amh', 'ibo', 'fra', 'sna', 'lin', 'wol', 'ewe', 'lug', 'xho', 'kin', 'twi', 'zul', 'orm', 'yor',
'hau', 'sot', 'swa']
languages_REGEX = {"eng": "The answer is (\\-?[0-9\\.\\,]+)",
languages = [
"eng",
"amh",
"ibo",
"fra",
"sna",
"lin",
"wol",
"ewe",
"lug",
"xho",
"kin",
"twi",
"zul",
"orm",
"yor",
"hau",
"sot",
"swa",
]
languages_REGEX = {
"eng": "The answer is (\\-?[0-9\\.\\,]+)",
"amh": "መልሱ (\\-?[0-9\\.\\,]+)",
"ibo": "Azịza ya bụ (\\-?[0-9\\.\\,]+)",
'fra': "La réponse est(\\-?[0-9\\.\\,]+)",
'sna': "Mhinduro kumubvunzo ndi (\\-?[0-9\\.\\,]+)",
'lin': "Eyano ezali (\\-?[0-9\\.\\,]+)",
'wol': "Tontu li (\\-?[0-9\\.\\,]+)",
'ewe': "ŋuɖoɖoae nye (\\-?[0-9\\.\\,]+)",
'lug': "Ansa eri (\\-?[0-9\\.\\,]+)",
'xho': "Impendulo ngu (\\-?[0-9\\.\\,]+)",
'kin': "Igisubizo ni (\\-?[0-9\\.\\,]+)",
'twi': "Ne nnyiano yɛ (\\-?[0-9\\.\\,]+)",
'zul': "Impendulo ithi (\\-?[0-9\\.\\,]+)",
'orm': "Deebiin isaa (\\-?[0-9\\.\\,]+)",
'yor': "Ìdáhùn náà ni (\\-?[0-9\\.\\,]+)",
'hau': "Amsar ita ce (\\-?[0-9\\.\\,]+)",
'sot': "Karabo ke (\\-?[0-9\\.\\,]+)",
'swa': "Jibu ni (\\-?[0-9\\.\\,]+)",
}
"fra": "La réponse est(\\-?[0-9\\.\\,]+)",
"sna": "Mhinduro kumubvunzo ndi (\\-?[0-9\\.\\,]+)",
"lin": "Eyano ezali (\\-?[0-9\\.\\,]+)",
"wol": "Tontu li (\\-?[0-9\\.\\,]+)",
"ewe": "ŋuɖoɖoae nye (\\-?[0-9\\.\\,]+)",
"lug": "Ansa eri (\\-?[0-9\\.\\,]+)",
"xho": "Impendulo ngu (\\-?[0-9\\.\\,]+)",
"kin": "Igisubizo ni (\\-?[0-9\\.\\,]+)",
"twi": "Ne nnyiano yɛ (\\-?[0-9\\.\\,]+)",
"zul": "Impendulo ithi (\\-?[0-9\\.\\,]+)",
"orm": "Deebiin isaa (\\-?[0-9\\.\\,]+)",
"yor": "Ìdáhùn náà ni (\\-?[0-9\\.\\,]+)",
"hau": "Amsar ita ce (\\-?[0-9\\.\\,]+)",
"sot": "Karabo ke (\\-?[0-9\\.\\,]+)",
"swa": "Jibu ni (\\-?[0-9\\.\\,]+)",
}
LANGUAGES = {}
for lang in languages:
if lang == 'amh':
if lang == "amh":
LANGUAGES[lang] = { # English
"QUESTION": "ጥያቄ:",
"ANSWER": "በቅደም ተከተል መልስ:",
"DIRECT": "Answer:",
"REGEX": languages_REGEX[lang]}
elif lang == 'yor':
"REGEX": languages_REGEX[lang],
}
elif lang == "yor":
LANGUAGES[lang] = { # English
"QUESTION": "Ìbéèrè:",
"ANSWER": "Ìdáhùn lẹ́sẹsẹ:",
"DIRECT": "Answer:",
"REGEX": languages_REGEX[lang]}
"REGEX": languages_REGEX[lang],
}
else:
LANGUAGES[lang] = { # English
"QUESTION": "Question:",
"ANSWER": "Step-by-Step Answer:",
"DIRECT": "Answer:",
"REGEX": languages_REGEX[lang]}
"REGEX": languages_REGEX[lang],
}
def add_regex_pattern(regex_pattern):
......@@ -93,13 +116,12 @@ def gen_lang_yamls(output_dir: str, overwrite: bool, mode: str) -> None:
err = []
for lang in LANGUAGES.keys():
try:
yaml_template = "cot_yaml"
filter_list = {}
DELIMITER = None
if mode == "direct":
ANSWER = LANGUAGES['eng']["DIRECT"]
QUESTION = LANGUAGES['eng']["QUESTION"]
ANSWER = LANGUAGES["eng"]["DIRECT"]
QUESTION = LANGUAGES["eng"]["QUESTION"]
REGEX = None
task_name = f"afrimgsm_direct_{lang}"
yaml_template = "direct_yaml"
......@@ -122,8 +144,8 @@ def gen_lang_yamls(output_dir: str, overwrite: bool, mode: str) -> None:
QUESTION = LANGUAGES["eng"]["QUESTION"]
task_name = f"afrimgsm_en_cot_{lang}"
elif mode == "translate-direct":
ANSWER = LANGUAGES['eng']["DIRECT"]
QUESTION = LANGUAGES['eng']["QUESTION"]
ANSWER = LANGUAGES["eng"]["DIRECT"]
QUESTION = LANGUAGES["eng"]["QUESTION"]
REGEX = None
task_name = f"afrimgsm_translate_direct_{lang}"
yaml_template = "translate_direct_yaml"
......
dataset_name: eng
include: afrimmlu_common_yaml
task: afrimmlu_direct_eng
dataset_name: ewe
include: afrimmlu_common_yaml
task: afrimmlu_direct_ewe
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment