s470628/challenging-america-word-gap-prediction
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

test 5 version

Browse Source
This commit is contained in:
pietrzakkuba 2022-04-03 09:59:04 +02:00
parent fa9858ae9a
commit 4e91ffc077
3 changed files with 14412 additions and 14410 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16882
dev-0/out.tsv
View File

File diff suppressed because one or more lines are too long

12
model.py
View File

@ -21,14 +21,16 @@ def preprocess(text):
def predict(word_before, word_after): def predict(word_before, word_after):
prob_list = model[(word_before, word_after)].items() prob_list = dict(Counter(model[(word_before, word_after)]).most_common(5)).items()
predictions = [] predictions = []
total = 0.0 prob_sum = 0.0
for key, value in prob_list: for key, value in prob_list:
total += value prob_sum += value
predictions.append(f'{key}:{value}') predictions.append(f'{key}:{value}')
if total == 0.0: if prob_sum == 0.0:
return 'the:1.0' return 'the:0:2 be:0.2 to:0.2 of:0.15 and:0.15 :0.1'
elif prob_sum < 1.0:
predictions.append(f':{1.0 - prob_sum}')
return ' '.join(predictions) return ' '.join(predictions)

11928
test-A/out.tsv
View File

File diff suppressed because one or more lines are too long