This commit is contained in:
AdamOsiowy123 2022-04-26 23:03:28 +02:00
parent 57059995fb
commit 56f44d61fc

2
run.py
View File

@ -49,9 +49,7 @@ def evaluate_and_save(path, file, model, vectorizer):
def main(): def main():
train_x = read_xz_file('train', 'in.tsv.xz', 500000) train_x = read_xz_file('train', 'in.tsv.xz', 500000)
# train_x = read_xz_file('train', 'in.tsv.xz')
train_y = read_tsv('train', 'expected.tsv', 500000) train_y = read_tsv('train', 'expected.tsv', 500000)
# train_y = read_tsv('train', 'expected.tsv')
tfidf_vectorizer = TfidfVectorizer() tfidf_vectorizer = TfidfVectorizer()
train_x_vectorized = tfidf_vectorizer.fit_transform(train_x['col_name'].values) train_x_vectorized = tfidf_vectorizer.fit_transform(train_x['col_name'].values)