【发布时间】：2020-02-13 17:14:11
【问题描述】：
我使用 chatterbot 和 tkinter 库创建了一个 Chatbot。 但是每当我打开文件时，它就会启动 training model 并花费大量时间，因此我搜索并找到了 pickle 模块。 但现在我也尝试了 pickle 它不是工作并显示错误。
有什么方法可以保存不会每次都开始训练的模型。这是我的代码
import chatterbot
import pickle
from chatterbot import ChatBot
from chatterbot.trainers import ListTrainer
import os
import tkinter as tk
try:
import ttk as ttk
import ScrolledText
except ImportError:
import tkinter.ttk as ttk
import tkinter.scrolledtext as ScrolledText
import time
class TkinterGUIExample(tk.Tk):
def __init__(self, *args, **kwargs):
"""
Create & set window variables.
"""
tk.Tk.__init__(self, *args, **kwargs)
self.chatbot = ChatBot(
"GUI Bot",
storage_adapter="chatterbot.storage.SQLStorageAdapter",
logic_adapters=[{
'import_path': 'chatterbot.logic.BestMatch',
'default_response': 'I am sorry, but I do not understand.',
'maximum_similarity_threshold': 0.80
} ]
)
for files in os.listdir('C:/Users/HP/Desktop/FYP BOT/training_data/'):
con=open('C:/Users/HP/Desktop/FYP BOT/training_data/'+files,'r').readlines()
trainer = ListTrainer(self.chatbot)
trainer.train(con)
self.title("Chatterbot")
self.initialize()
def initialize(self):
"""
Set window layout.
"""
self.grid()
ttk.Style().configure("TButton", padding=6, relief="flat",background="#ccc")
style = ttk.Style()
style.map("C.TButton",
foreground=[('pressed', 'red'), ('active', 'blue')],
background=[('pressed', '!disabled', 'black'), ('active', 'white')]
)
self.respond = ttk.Button(self, text='Get Response',cursor='hand2' ,command=self.get_response)
self.respond.grid(column=1, row=2, sticky='nesw', padx=3, pady=10)
self.usr_input = tk.Entry(self, state='normal',text='Enter your query here!')
self.usr_input.grid(column=0, row=2, sticky='nesw', padx=1, pady=5)
#Binding entry
self.usr_input.bind('<Return>',self.get_response)
self.conversation_lbl = tk.Label(self,
text='English',
anchor='center',
font=('Arial Bold ',18),
bg="#3a8fc5",
fg="white")
self.conversation_lbl.grid(column=0, row=0,columnspan=2, padx=3, pady=3,sticky='news')
self.conversation = ScrolledText.ScrolledText(self,
state='disabled',borderwidth=5,
highlightthickness=1,
bg='#15202b',fg='#16202A',
font=('Arial Bold',8))
self.conversation.grid(column=0, row=1, columnspan=2, sticky='nesw', padx=3, pady=3)
def get_response(self,*args):
"""
Get a response from the chatbot and display it.
"""
user_input = self.usr_input.get()
self.usr_input.delete(0, tk.END)
response = self.chatbot.get_response(user_input)
self.conversation['state'] = 'normal'
'''----------------------------------------------
self.conversation.tag_configure('tag-left', justify='left')
self.conversation.insert('end',"Human: " + user_input + "\n", 'tag-left')
self.conversation.tag_configure('tag-left', justify='right')
self.conversation.insert('end',"ChatBot: " + str(response.text) + "\n\n\n", 'tag-right')'''
label1 = tk.Label(self.conversation,
text="Human: \n"+user_input,
background='#3B5566',
fg='white',
font=("Helvetica", 12),
justify='left',
wraplength=300,
anchor='w',
padx=10, pady=5)
label2 = tk.Label(self.conversation,
text="ChatBot: \n"+str(response.text),
wraplength=300,
anchor='w',
background='#1D9DFC',
fg='white',
font=("Helvetica", 12),
justify='left',
padx=10, pady=5)
self.conversation.tag_configure('tag-left', justify='left')
self.conversation.tag_configure('tag-right', justify='right')
self.conversation.insert('end', '\n\n\n')
self.conversation.window_create('end', window=label1)
self.conversation.insert('end', '\n\n\n ', 'tag-right') # space to move Label to the right
self.conversation.window_create('end', window=label2)
'''self.conversation.insert(
tk.END, "Human: " + user_input + "\n" + "ChatBot: " + str(response.text) + "\n\n\n"
)'''
self.conversation['state'] = 'disabled'
time.sleep(0.2)
gui_example = TkinterGUIExample()
gui_example.attributes('-topmost', True)
gui_example.update()
gui_example.attributes('-topmost', False)
gui_example.geometry('810x550+460+100')
gui_example.resizable(0, 0)
gui_example.configure(background='#3a8fc5')
gui_example.mainloop()
我还创建了 .exi 文件，但它也开始了训练，所以有什么方法可以在没有错误的情况下保存它，当我在主窗口中调用这个脚本时，这个脚本开始工作而不是训练等。
【问题讨论】：
标签： python machine-learning pickle