Python 预测失败：检查输入时出错：预期密集_输入具有形状（2898），但获得具有形状（1）的数组

我使用以下脚本predictor.py从GCP AI平台中托管的Keras模型中获取预测

import os
import pickle
import tensorflow as tf
import numpy as np
import logging

class MyPredictor(object):

    def __init__(self, model, bow_model):
        self._model = model
        self._bow_model = bow_model

    def predict(self, instances, **kwargs):
      
        vectors = self.embedding([instances])

        vectors = vectors.tolist()

        output = self._model.predict(vectors)

        return output

    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)

不管我得到什么

Prediction failed: Error when checking input: expected dense_input to have shape (2898,) but got array with shape (1,)

问题似乎是由于我在尝试进行实际预测时输入数据的维度造成的，行内输出=self.\u model.predict（[vectors]）。模型需要一个形状向量（2898，）

我觉得这很奇怪。。。因为当我打印向量的形状和尺寸时，我得到了以下结果

This is the shape
(1, 2898)

This is the dim number
2

This is the vector 
[[0 0 0 ... 0 0 0]]

因此，尺寸和形状都很好，它应该真的工作

此外，我做了一个测试，以获得本地存储的模型预测，它运行良好。这是测试文件：

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, **kwargs):

        print("These are the instances ", instances)

        vector = self.embedding([instances])

        output = self._model.predict(vector)

        return output

    def embedding(self, statement):
        vector = self._bow_model.transform(statement).toarray()
        #vector = vector.to_list()
        return vector



model_path = 'model.h5'
model = tf.keras.models.load_model(model_path, compile = False)

preprocessor_path = 'bow.pkl'
with open(preprocessor_path, 'rb') as f:
    bow_model = pickle.load(f)


instances = 'test'

predictor = MyPredictor(model, bow_model)

outputs = predictor.predict(instances)

print(outputs)

解决了

这就像在这行添加一组括号一样愚蠢

output=self.\u model.predict（[vectors]）

在此之后，我得到了另一个关于预测输出不可json序列化的错误。我只需将.tolist（）添加到return

return output.to\u list（）

import os
import pickle
import tensorflow as tf
import numpy as np
import logging

class MyPredictor(object):

    def __init__(self, model, bow_model):
        self._model = model
        self._bow_model = bow_model

    def predict(self, instances, **kwargs):
      
        vectors = self.embedding([instances])

        vectors = vectors.tolist()

        output = self._model.predict([vectors])

        return output.to_list()

    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)