问题描述
我正在尝试从 AI 平台上提供的 Tensor Flow 自定义例程中获取预测。
我设法使用以下设置为其提供服务:--runtime-version 2.3 --python-version 3.7 --machine-type mls1-c4-m2
但是当我尝试做出任何预测时,我总是收到此错误。
ERROR:root:Prediction Failed: predict() got an unexpected keyword argument 'stats'
ERROR:root:Prediction Failed: unkNown error.
该例程有两个步骤:
这是我的 setup.py
from setuptools import setup
required_PACKAGES = ['Keras==2.3.1','sklearn==0.0','h5py<3.0.0','numpy==1.16.0','scipy==1.4.1','pyyaml==5.2']
setup(
name='my_custom_code',version='0.1',scripts=['predictor.py'],install_requires=required_PACKAGES,packages=find_packages(),include_package_data=False,description=''
)
这是我的predictor.py
import os
import pickle
import tensorflow as tf
import numpy as np
class MyPredictor(object):
def __init__(self,model,bow_model):
self._model = model
self._bow_model = bow_model
def predict(self,instances):
outputs = []
for x in instances:
vector = self.embedding(x)
output = self._model.predict(vector)
outputs.append(output)
return outputs
def embedding(self,statement):
vector = self._bow_model.transform(statement).toarray()
vector = vector.to_list()
return vector
@classmethod
def from_path(cls,model_dir):
model_path = os.path.join(model_dir,'model.h5')
model = tf.keras.models.load_model(model_path,compile = False)
preprocessor_path = os.path.join(model_dir,'bow.pkl')
with open(preprocessor_path,'rb') as f:
bow_model = pickle.load(f)
return cls(model,bow_model)
我用于测试的脚本是:
import googleapiclient.discovery
instances = ['test','test']]
service = googleapiclient.discovery.build('ml','v1')
name = 'projects/{}/models/{}/versions/{}'.format(PROJECT_ID,MODEL_NAME,VERSION_NAME)
response = service.projects().predict(
name=name,body={'instances': instances}
).execute()
if 'error' in response:
raise RuntimeError(response['error'])
else:
print(response['predictions'])
解决方法
根据自定义预测例程 documentation,一旦创建预测器 class,predict()
方法应提供 self,instances,**kwargs
参数以正确处理预测请求。
instances:预测输入实例列表。
**kwargs:在预测请求正文中作为附加字段提供的关键字参数字典。