TensorFlow/TFLearn聊天机器人报错：RuntimeError(尝试使用已关闭会话)

问题背景 #

按照教程制作深度学习聊天机器人，代码运行时报错RuntimeError: Attempted to use a closed Session。尝试手动添加model.tflearn文件无效，每次运行必须删除data.pickle才能执行，但仍无法解决核心报错。

原始代码 #

import nltk
from nltk.stem.lancaster import LancasterStemmer
stemmer = LancasterStemmer()

import numpy
import tflearn
import tensorflow
import random
import json
import pickle

with open("intents.json") as file:
    data = json.load(file)

try:
    with open("data.pickle", "rb") as f:
        words, labels, training, output = pickle.load(f)
except: 
    words = []
    labels = []
    docs_x = []
    docs_y = []

for intent in data["intents"]:
    for pattern in intent["patterns"]:
        wrds = nltk.word_tokenize(pattern)
        words.extend(wrds)
        docs_x.append(wrds)
        docs_y.append(intent["tag"])

    if intent["tag"] not in labels:
        labels.append(intent["tag"])

words = [stemmer.stem(w.lower()) for w in words if w != "?"]
words = sorted(list(set(words)))

labels = sorted(labels)

training = []
output = []

out_empty = [0 for _ in range(len(labels))]

for x, doc in enumerate(docs_x):
    bag = []

    wrds = [stemmer.stem(w.lower()) for w in doc]

    for w in words:
        if w in wrds:
            bag.append(1)
        else:
            bag.append(0)
    
    output_row = out_empty[:]
    output_row[labels.index(docs_y[x])] = 1

    training.append(bag)
    output.append(output_row)

training = numpy.array(training)
output = numpy.array(output)

with open("data.pickle", "wb") as f:
    pickle.dump((words, labels, training, output), f)

tensorflow.compat.v1.reset_default_graph()

net = tflearn.input_data(shape=[None, len(training[0])])
net = tflearn.fully_connected(net, 8)
net = tflearn.fully_connected(net, 8)
net = tflearn.fully_connected(net, len(output[0]), activation="softmax")
net = tflearn.regression(net)

model = tflearn.DNN(net)

try:
    model.load("model.tflearn")
except:
    model.fit(training, output, n_epoch=100, batch_size=8, show_metric=True)
    model.save("model.tflearn")
    
def bag_of_words(s, words):
    bag = [0 for _ in range(len(words))]

    s_words = nltk.word_tokenize(s)
    s_words = [stemmer.stem(word.lower()) for word in s_words]

    for se in s_words:
        for i, w in enumerate(words):
            if w == se:
                bag[i] = 1

    return numpy.array(bag)

def chat():
    print("Start talking with the bot (type quit to stop)!")
    while True:
        inp = input("You: ")
        if inp.lower() == "quit":
            break

        results = model.predict([bag_of_words(inp, words)])
        results_index = numpy.argmax(results)
        tag = labels[results_index]

        for tg in data["intents"]:
            if tg['tag'] == tag:
                responses = tg['responses']

        print(random.choice(responses))

chat()

终端报错信息 #

tflearn.is_training(True, session=self.session) File "/usr/local/lib/python3.7/site-packages/tflearn/config.py", line 95, in is_training tf.get_collection('is_training_ops')[0].eval(session=session) File "/usr/local/lib/python3.7/site-packages/tensorflow/python/framework/ops.py", line 913, in eval return _eval_using_default_session(self, feed_dict, self.graph, session) File "/usr/local/lib/python3.7/site-packages/tensorflow/python/framework/ops.py", line 5512, in _eval_using_default_session return session.run(tensors, feed_dict) File "/usr/local/lib/python3.7/site-packages/tensorflow/python/client/session.py", line 958, in run run_metadata_ptr) File "/usr/local/lib/python3.7/site-packages/tensorflow/python/client/session.py", line 1104, in _run raise RuntimeError('Attempted to use a closed Session.') RuntimeError: Attempted to use a closed Session.

问题分析 #

1. 会话关闭问题：TFlearn加载预训练模型后，TensorFlow会话被自动关闭，后续调用model.predict()时无法访问已关闭的会话，引发报错。
2. 数据逻辑错误：原始代码中，数据预处理逻辑（分词、词干提取、生成训练集）不在except块内，导致无论是否成功加载data.pickle，都会重新生成并覆盖数据，这就是必须删除data.pickle才能运行的原因。
3. 版本兼容问题：TFlearn对TensorFlow 2.x的兼容性较差，当前使用的TensorFlow 2.x环境和TFlearn的会话管理逻辑冲突。

解决步骤 #

1. 调整数据加载逻辑 #

把数据预处理的所有代码放入except块内，确保只有当data.pickle不存在时才重新生成数据：

try:
    with open("data.pickle", "rb") as f:
        words, labels, training, output = pickle.load(f)
except: 
    words = []
    labels = []
    docs_x = []
    docs_y = []

    for intent in data["intents"]:
        for pattern in intent["patterns"]:
            wrds = nltk.word_tokenize(pattern)
            words.extend(wrds)
            docs_x.append(wrds)
            docs_y.append(intent["tag"])

        if intent["tag"] not in labels:
            labels.append(intent["tag"])

    words = [stemmer.stem(w.lower()) for w in words if w != "?"]
    words = sorted(list(set(words)))

    labels = sorted(labels)

    training = []
    output = []

    out_empty = [0 for _ in range(len(labels))]

    for x, doc in enumerate(docs_x):
        bag = []

        wrds = [stemmer.stem(w.lower()) for w in doc]

        for w in words:
            if w in wrds:
                bag.append(1)
            else:
                bag.append(0)
        
        output_row = out_empty[:]
        output_row[labels.index(docs_y[x])] = 1

        training.append(bag)
        output.append(output_row)

    training = numpy.array(training)
    output = numpy.array(output)

    with open("data.pickle", "wb") as f:
        pickle.dump((words, labels, training, output), f)

2. 修复模型会话问题 #

在加载模型后，显式重新初始化TensorFlow会话，或者在创建模型时指定会话并保持活跃。以下是适配的代码：

# 替换原模型创建和加载部分
tensorflow.compat.v1.reset_default_graph()
sess = tensorflow.compat.v1.Session()

net = tflearn.input_data(shape=[None, len(training[0])])
net = tflearn.fully_connected(net, 8)
net = tflearn.fully_connected(net, 8)
net = tflearn.fully_connected(net, len(output[0]), activation="softmax")
net = tflearn.regression(net)

model = tflearn.DNN(net, session=sess)

try:
    model.load("model.tflearn")
except:
    model.fit(training, output, n_epoch=100, batch_size=8, show_metric=True)
    model.save("model.tflearn")

3. 版本兼容建议 #

如果上述调整仍有问题，建议安装TensorFlow 1.15.x版本（TFlearn的稳定兼容版本）：

pip uninstall tensorflow -y
pip install tensorflow==1.15.5

修正后的完整代码 #

import nltk
from nltk.stem.lancaster import LancasterStemmer
stemmer = LancasterStemmer()

import numpy
import tflearn
import tensorflow
import random
import json
import pickle

with open("intents.json") as file:
    data = json.load(file)

try:
    with open("data.pickle", "rb") as f:
        words, labels, training, output = pickle.load(f)
except: 
    words = []
    labels = []
    docs_x = []
    docs_y = []

    for intent in data["intents"]:
        for pattern in intent["patterns"]:
            wrds = nltk.word_tokenize(pattern)
            words.extend(wrds)
            docs_x.append(wrds)
            docs_y.append(intent["tag"])

        if intent["tag"] not in labels:
            labels.append(intent["tag"])

    words = [stemmer.stem(w.lower()) for w in words if w != "?"]
    words = sorted(list(set(words)))

    labels = sorted(labels)

    training = []
    output = []

    out_empty = [0 for _ in range(len(labels))]

    for x, doc in enumerate(docs_x):
        bag = []

        wrds = [stemmer.stem(w.lower()) for w in doc]

        for w in words:
            if w in wrds:
                bag.append(1)
            else:
                bag.append(0)
        
        output_row = out_empty[:]
        output_row[labels.index(docs_y[x])] = 1

        training.append(bag)
        output.append(output_row)

    training = numpy.array(training)
    output = numpy.array(output)

    with open("data.pickle", "wb") as f:
        pickle.dump((words, labels, training, output), f)

tensorflow.compat.v1.reset_default_graph()
sess = tensorflow.compat.v1.Session()

net = tflearn.input_data(shape=[None, len(training[0])])
net = tflearn.fully_connected(net, 8)
net = tflearn.fully_connected(net, 8)
net = tflearn.fully_connected(net, len(output[0]), activation="softmax")
net = tflearn.regression(net)

model = tflearn.DNN(net, session=sess)

try:
    model.load("model.tflearn")
except:
    model.fit(training, output, n_epoch=100, batch_size=8, show_metric=True)
    model.save("model.tflearn")
    
def bag_of_words(s, words):
    bag = [0 for _ in range(len(words))]

    s_words = nltk.word_tokenize(s)
    s_words = [stemmer.stem(word.lower()) for word in s_words]

    for se in s_words:
        for i, w in enumerate(words):
            if w == se:
                bag[i] = 1

    return numpy.array(bag)

def chat():
    print("Start talking with the bot (type quit to stop)!")
    while True:
        inp = input("You: ")
        if inp.lower() == "quit":
            break

        results = model.predict([bag_of_words(inp, words)])
        results_index = numpy.argmax(results)
        tag = labels[results_index]

        responses = ""
        for tg in data["intents"]:
            if tg['tag'] == tag:
                responses = tg['responses']

        print(random.choice(responses))

chat()

内容的提问来源于stack exchange，提问作者Hyun Sim