使用 pickle 保存 keras 模型时面临“无法腌制 _thread.rlock 对象”错误答案

【问题标题】：Facing 'can't pickle _thread.rlock objects' error while saving keras model using pickle使用 pickle 保存 keras 模型时面临“无法腌制 _thread.rlock 对象”错误
【发布时间】：2020-09-25 12:38:54
【问题描述】：

我正在尝试使用 keras 构建分类器，下面是我的代码：

import pandas as pd
import numpy as np

import matplotlib.pyplot as plt
import seaborn as sns

from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score, confusion_matrix
from sklearn.preprocessing import MinMaxScaler

from collections import Counter
from imblearn.over_sampling import SMOTE

from keras.models import Sequential
from keras.layers import Dense

import pickle
import joblib

import warnings
warnings.filterwarnings('ignore')

df = pd.read_csv('bankloan.csv')
df = df.dropna()
df.isna().any()
df = df.drop('Loan_ID', axis=1)
df['LoanAmount'] = (df['LoanAmount']*1000).astype(int)

pre_y = df['Loan_Status']
pre_X = df.drop('Loan_Status', axis=1)
dm_X = pd.get_dummies(pre_X)
dm_y = pre_y.map(dict(Y=1, N=0))

smote = SMOTE()
X1, y = smote.fit_sample(dm_X, dm_y)
sc = MinMaxScaler()
X = sc.fit_transform(X1)

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42, shuffle=True)

classifier = Sequential()
classifier.add(Dense(200, activation='relu', input_dim=X_test.shape[1]))
classifier.add(Dense(400, activation='relu'))
classifier.add(Dense(4, activation='relu'))
classifier.add(Dense(1, activation='sigmoid'))
classifier.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
classifier.fit(X_train, y_train, batch_size=20, epochs=100, verbose=0)

filename = 'loan_model.pkl'
joblib.dump(classifier, filename)

这是我正在尝试做的，模型非常好，但是在我尝试保存模型的最后一步时，它显示了这个错误：

TypeError: can't pickle _thread.RLock objects

【问题讨论】：

看看at this Github issue，这似乎是一个库问题，没有什么可以帮助解决的
没有帮助。
确实，因为问题在 keras 方面没有得到解决（而且我怀疑它永远不会解决，考虑到 keras 已完全集成到 TF 中，我不认为 tf.keras 是永远腌制）

标签： python keras pickle

【解决方案1】：

在酸洗模型之前运行以下代码：

import types
import tempfile
import keras.models

def make_keras_picklable():
    def __getstate__(self):
        model_str = ""
        with tempfile.NamedTemporaryFile(suffix='.hdf5', delete=True) as fd:
            keras.models.save_model(self, fd.name, overwrite=True)
            model_str = fd.read()
        d = { 'model_str': model_str }
        return d

    def __setstate__(self, state):
        with tempfile.NamedTemporaryFile(suffix='.hdf5', delete=True) as fd:
            fd.write(state['model_str'])
            fd.flush()
            model = keras.models.load_model(fd.name)
        self._dict_ = model._dict_


    cls = keras.models.Model
    cls.__getstate__ = __getstate__
    cls.__setstate__ = __setstate__

make_keras_picklable()

【讨论】：