將tensorflow訓練好的模型部署成sercice服務,並做預測
阿新 • • 發佈：2019-01-07
web框架用的是tornado，模型載入在上一篇部落格已經提到，下面寫了一個service.py檔案:
import tornado.webimport loggingimport stringimport osimport sysfrom predict import predictlogger = logging.getLogger('mylogger')logger.setLevel(logging.DEBUG)# 再建立一個handler，用於輸出到控制檯ch = logging.StreamHandler()ch.setLevel(logging.DEBUG)# 定義handler的輸出格式formatter = logging.Formatter('%(asctime)s - %(message)s')ch.setFormatter(formatter)# 給logger新增handlerlogger.addHandler(ch)def work(text):if text isNone:return0.else:return predict(text)classMainHandler(tornado.web.RequestHandler):def post(self):text = self.get_argument('text')p = work(text)self.write(p)classIndexHandler(tornado.web.RequestHandler):def get(self):text = self.get_argument('text')p = work(text)self.render('index.html',text=text,result='{0}'.format(p))application=tornado.web.Application(handlers=[(r"/",MainHandler),(r'/demo',IndexHandler)],template_path=os.path.join(os.path.dirname(__file__),"templates"))if __name__=='__main__':if len(sys.argv)<2:port=8410else:port=string.atoi(sys.argv[1])application.listen(port)logger.info('Starting server:port=%d, use <Ctrl-C> to stop'% port)tornado.ioloop.IOLoop.instance().start()
templates/index.html問價內容：
<!DOCTYPE html><html><head><title>text abumse server</title></head><body><form method="get" action="/demo"><p>text content:<input type="text" name="text" value="{{text}}" style="width: 800px"><input type="submit"></p></form><p>{{result}}</p><p><img src="{{text}}" width="600px"></p></body></html>
啟動service.py檔案:
/anaconda/bin/python3 /Users/shuubiasahi/Documents/python/credit-tftextclassify-abuse/service.py2017-10-1315:58:20.625704: W tensorflow/core/platform/cpu_feature_guard.cc:45]TheTensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.2017-10-1315:58:20.625723: W tensorflow/core/platform/cpu_feature_guard.cc:45]TheTensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.2017-10-1315:58:20.625728: W tensorflow/core/platform/cpu_feature_guard.cc:45]TheTensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.2017-10-1315:58:20.625733: W tensorflow/core/platform/cpu_feature_guard.cc:45]TheTensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations.2017-10-1315:58:21,732-Starting server:port=8410, use <Ctrl-C> to stop
本身我是用的cnn做的辱罵識別模型，預測的概率還挺高，看到最後別噴啊 ，哈哈哈哈。。。。。