資源簡介
計劃趕不上變化快的fasttext來啦~ 包括代碼和分好詞的訓練集、測試集、停詞表,測試集F1值0.941,雖然沒比TextCNN高但是快啊!fasttext模塊安裝麻煩一點,具體看博客內(nèi)容吧。
代碼片段和文件信息
#?-*-?coding:?utf-8?-*-
“““
Created?on?Mon?Jul?27?21:12:20?2020
@author:?29811
“““
import?logging
import?fasttext
import?time
from?datetime?import?timedelta
basedir?=?‘D:/S/Learn/CNew/Cnews_Fasttext/‘
logging.basicConfig(format=‘%(asctime)s?:?%(levelname)s?:?%(message)s‘?level=logging.INFO)
start_time?=?time.time()
#?訓練
model?=?fasttext.train_supervised(basedir?+?‘cnews.seg.train‘?label_prefix?=?‘__label__‘?word_ngrams=3?bucket=2000000)
end_time?=?time.time()
time_dif?=?end_time?-?start_time
print(‘time?of?train:‘??timedelta(seconds=int(round(time_dif))))
#?保存
model.save_model(basedir?+?‘cnews.fasttext‘)
#?加載
model?=?fasttext.load_model(basedir?+?‘cnews.fasttext‘)
#測試
num??precision??recall?=?model.test(basedir?+?‘cnews.seg.test‘)
print(precision?*?recall?*?2?/?(recall?+?precision))
?屬性????????????大小?????日期????時間???名稱
-----------?---------??----------?-----??----
?????文件???????2867??2020-07-28?22:03??data_process.py
?????文件??112538995??2020-07-28?22:00??cnews.seg.train
?????文件???23865365??2020-07-28?22:03??cnews.seg.test
?????文件????????869??2020-07-29?12:00??CNEW_Fasttext.py
?????文件??????19782??2020-03-01?19:48??cnews.vocab.txt
-----------?---------??----------?-----??----
????????????136427878????????????????????5
評論
共有 條評論