from examples.sentence_level.wmt_2018.common.util.postprocess import format_submission
from examples.sentence_level.wmt_2018.common.util.reader import read_annotated_file, read_test_file
from examples.sentence_level.wmt_2018.en_lv.nmt.monotransquest_config import TEMP_DIRECTORY, MODEL_NAME, \
monotransquest_config, MODEL_TYPE, SEED, RESULT_FILE, SUBMISSION_FILE, RESULT_IMAGE
from transquest.algo.sentence_level.monotransquest.evaluation import pearson_corr, spearman_corr
from transquest.algo.sentence_level.monotransquest.run_model import MonoTransQuestModel
if not os.path.exists(TEMP_DIRECTORY):
os.makedirs(TEMP_DIRECTORY)
TRAIN_FOLDER = "examples/sentence_level/wmt_2018/en_lv/data/en_lv/"
DEV_FOLDER = "examples/sentence_level/wmt_2018/en_lv/data/en_lv/"
TEST_FOLDER = "examples/sentence_level/wmt_2018/en_lv/data/en_lv/"
train = read_annotated_file(path=TRAIN_FOLDER,
original_file="train.nmt.src",
translation_file="train.nmt.mt",
hter_file="train.nmt.hter")
dev = read_annotated_file(path=DEV_FOLDER,
original_file="dev.nmt.src",
translation_file="dev.nmt.mt",
hter_file="dev.nmt.hter")
test = read_test_file(path=TEST_FOLDER,
original_file="test.nmt.src",
translation_file="test.nmt.mt")
train = train[['original', 'translation', 'hter']]
dev = dev[['original', 'translation', 'hter']]
test = test[['index', 'original', 'translation']]
index = test['index'].to_list()
train = train.rename(columns={
"examples/sentence_level/wmt_2018/en_lv/data/en_lv",
"examples/sentence_level/wmt_2018/en_lv/data/en_lv", "smt"
],
}
train_list = []
dev_list = []
test_list = []
index_list = []
test_sentence_pairs_list = []
for key, value in languages.items():
if value[3] == "nmt":
train_temp = read_annotated_file(path=value[0],
original_file="train.nmt.src",
translation_file="train.nmt.mt",
hter_file="train.nmt.hter")
dev_temp = read_annotated_file(path=value[1],
original_file="dev.nmt.src",
translation_file="dev.nmt.mt",
hter_file="dev.nmt.hter")
test_temp = read_test_file(path=value[2],
original_file="test.nmt.src",
translation_file="test.nmt.mt")
elif value[3] == "smt":
train_temp = read_annotated_file(path=value[0],
original_file="train.smt.src",
translation_file="train.smt.mt",
hter_file="train.smt.hter")
dev_temp = read_annotated_file(path=value[1],