tensorflow serving 模型部署

拉去tensorflow srving 镜像

docker pull tensorflow/serving:1.12.0

代码里新增tensorflow 配置代码

# 要指出输入，输出张量
#指定保存路径
# serving_save
    signature = tf.saved_model.signature_def_utils.predict_signature_def(
        inputs={'input_param': model.inputs_placeholder}, outputs={'type': model.the_outputs})

    builder = tf.saved_model.builder.SavedModelBuilder('../serving_save/001')
    builder.add_meta_graph_and_variables(
        sess=sess,
        tags=[tf.saved_model.tag_constants.SERVING],
        signature_def_map={
            tf.saved_model.signature_constants.DEFAULT_SERVING_SIGNATURE_DEF_KEY:
                signature
        })
    builder.save()

启动服务

docker run -p 8501:8501 --mount type=bind,source=F:\avaintec\aad_by_me_new\serving_save,target=/models/aad_serve -e MODEL_NAME=aad_serve -t tensorflow/serving &

访问服务

que = [1, 2, 3, 44, 5, 6, 7, 8] + [0] * 348
input = [que]
print(input)

url = 'http://localhost:8501/v1/models/aad_serve:predict'

headers = {
    "Host": "anticvarium.ru",
    "Connection": "keep-alive",
    "Content-Length": "14",
    "Origin": "http://anticvarium.ru",
    "X-CSRF-TOKEN": "u1InfvhE23slcmReIJlgvI7IGzb3xQEvXHQbP3Bc",
    "User-Agent": "Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/67.0.3396.99 Safari/537.36",
    "Content-Type": "application/json;charset=UTF-8",
    "Accept": "application/json, text/plain, */*",
    "X-Requested-With": "XMLHttpRequest",
    "Referer": "http://anticvarium.ru/auction/show/66",
    "Accept-Language": "zh",
    "Cookie": "_ga=GA1.2.819821684.1533003173; _gid=GA1.2.258681840.1533003173; _ym_uid=1533003182951117996; _ym_d=1533003182; _ym_isad=2; _ym_visorc_35976415=w; tmr_detect=0%7C1533004777867; laravel_session=eyJpdiI6InZNSTlWSlYyUW1BaitUWTNBSjJKckE9PSIsInZhbHVlIjoibWprTDFxeTZZc0lOS2VTcmxIdzZJSkwwTVFXOHphN2FwTysxbXNHSDkybXE0VUZYakpTUlZ5Ykh6U2pFTTVJTWdWUHBUaVJwVGl1Q3RlTjdRQUhKNEE9PSIsIm1hYyI6IjQ5NmIwMGYxNDllZDFiNTY2ZmFjYWY5NjEyOTQ4MTZjYzQzNmFiY2EyMzFiOTY1Mzg3ODUyZTllN2U0MTQwYTEifQ%3D%3D",
}

response = requests.post(url, headers=headers, json={"instances": input})
print(response.text)
print('ok')

预测结果

遗留问题

tensorflow serving 保存的时侯，只保存了，模型graphy相关的操作。数据预处理操作，不在serving服务中。

所以数据的预处理等，需要重新做。

可不可以再服务中，直接包含数据预处理部分？？？

参考链接

https://www.cnblogs.com/xing901022/p/10216933.html

https://blog.csdn.net/weixin_34343000/article/details/88118667