def step3():
key_vec = {}
maxx = 12505807
size = 10000
for i in range(size, maxx, size):
print(i, maxx)
res = os.popen("head -n {i} ./dataset/bind.txt | tail -n {size} | ./fasttext print-sentence-vectors ./models/model.bin".format(i=i, size=size)).read()
for line in res.split("\n"):
if line == "":
continue
vec = list(map(float, line.split()[-100:]))
txt = line.split()[:-100]
key = " ".join(txt)
if key_vec.get(key) is None:
key_vec[key] = vec
open("key_vec.pkl", "wb").write(pickle.dumps(key_vec))
评论列表
文章目录