-
Notifications
You must be signed in to change notification settings - Fork 6
/
Interivew.py
226 lines (198 loc) · 9.82 KB
/
Interivew.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
import openai
import os
from textblob import TextBlob
import re
from flask import Flask, request, jsonify, render_template
from flask import session
from reportMailscript import send_mail
from flask_cors import CORS
# from profanity_check import predict,predict_prob
# from scipy.integrate import solve_ivp
try:
openai.api_key = os.environ["OPENAI_API_KEY"]
except:
print("OpenAI API key not found. Please set the OPENAI_API_KEY environment variable.")
def safe_division(numerator, denominator):
try:
return numerator / denominator
except ZeroDivisionError:
return numerator
def report_maker(questions, format_style, answers, model = "gpt-3.5-turbo", max_tokens = 250):
messages = [
{"role": "system", "content": f"In a recent interivew with following questions {questions}\n and candidate has responded with following answers to each of the questions {answers}\n\n Make a report of the interview and areas where the candidate can improve that can be sent to candidate for future use!\n No need to write each question and answer, just write the report and areas where the candidate can improve!"},
{"role": "user", "content": format_style}
]
response = openai.ChatCompletion.create(
model=model,
messages=messages,
temperature=0,
max_tokens=max_tokens,
)
return response.choices[0].message['content']
def score_maker(question, answer, model = "gpt-3.5-turbo", max_tokens = 1):
# parse format_style to comma separated string
messages = [
{"role": "system", "content": f"Rate the message given out of 10 (Be more lenient), based on its correctness, tone and overall behaviour comparing with the question asked! Only write the score where you can (no need to explain the score) otherwise score 0'."},
{"role": "user", "content": f"Question: {question}, Anwer: {answer}, Score: "}
]
response = openai.ChatCompletion.create(
model=model,
messages=messages,
temperature=0,
max_tokens=max_tokens,
)
return response.choices[0].message['content']
class Interview:
def __init__(self, name):
self.name = name
self.messages = [
{
"role": "system",
"content": "You are a interviewer named DevHire and you are interviewing a candidate for a job. \
You are asking the candidate about his/her experience. \
You can ask about the candidate's previous work experience, education, or anything else you think is relevant to the job. \
Assess the candidate's ability to analyze problems and provide effective solutions. Can they think critically and creatively to solve complex problems? \
You can also ask the candidate about various programming questions such as code completion and bug fixes. \
Assess whether the candidate's values, personality, and work style align with the company culture and team dynamics \
User will answer your questions one by one and you have to verify whether it is correct or not, its overall behaviour and tone throughout the interview should also be noted. "
},
]
self.score_message = [
{
"role": "assistant",
"content": "Rate the message given out of 10, based on its correctness, tone and overall behaviour comparing with the question asked! Only write the score where you can (no need to explain the score) otherwise score 0'.",
}
]
self.score_pattern = re.compile(r'\d+')
self.scoreByAnswer = 0
self.scoreByTone = 0
self.scoreByBot = 0
self.x_value = 0
self.questionsAsked = 0
self.currentTone = 0.0;
self.currentUnderstanding = 0.0;
self.currentBot = 0.0
self.profanity_flag_1 = False
self.profanity_flag_2 = False
self.scoreCount_0 = 0
self.scoreCount_1 = 0
self.scoreCount_2 = 0
self.stop_patterns = [
r".*(stop|end|pause|break|reschedule|leave|close|not interested|not ready).*interview.*",
]
self.questions = []
self.answers = []
self.patterns = [
r'(if|do) you have any (questions|concerns)',
r'Let me know if you change your mind\b',
r'(?:Is there anything else you would like to add|Do you have any questions for us?)\??',
r'Good luck with your job search(?!.*Good luck with your job search)',
r"we'll keep you updated on any developments",
r'thank you for your time',
r"thank\s+you(\s+very\s+much)?\s+for\s+your\s+time",
r"(?i)\b(thank\s*you(?:\s*very\s*much)?|(?:thanks|thankyou)(?:\s+(?:very\s+much))?)(?:\s*(?:for)\s*your\s*time)\b",
]
def process_message(self, message):
if self.questionsAsked == 0 and not (message.startswith(("My name is", r"Hi(,?) I am", r"Hi(,?) My name is", r"Hello(,?) My name is", r"Hello(,?) I am"))):
message = "My name is {} ".format(self.name) + message
# if predict_prob([message]) > 0.9:
# # profanity detected
# reply = "You are using profanity, please refrain from using it!"
# self.currentBot = 0.0
# self.currentTone = 0.0
# self.currentUnderstanding = 0.0
# if self.profanity_flag_1:
# self.scoreByTone = 0.0
# self.scoreByAnswer = 0.0
# self.scoreByBot = 0.0
# self.questionsAsked += 1
# reply = "You are using profanity again, your scores are reset to 0.0!"
# else:
# self.profanity_flag_1 = True
# return reply
# Bot Prompt
self.messages.append({"role": "user", "content": message})
try:
self.chatbot = openai.ChatCompletion.create(model="gpt-3.5-turbo", messages=self.messages)
except Exception as e:
return -2
reply = self.chatbot.choices[0].message.content
self.messages.append({"role": "system", "content": reply})
# Stop_words checkings for the bot
if any(re.match(pattern, message, re.IGNORECASE) for pattern in self.stop_patterns):
return 0
if self.questionsAsked > 20:
return 1
# Stop_words checkings for the bot
for pattern in self.patterns:
match = re.search(pattern, reply, flags=re.IGNORECASE)
if match or self.questionsAsked > 4:
return 1
self.questions.append(reply)
if self.questionsAsked > 0:
self.answers.append(message)
# Score Calculations
blobReply = TextBlob(reply)
blobMessage = TextBlob(message)
tone_score = blobMessage.sentiment.polarity
understanding_score = blobReply.sentiment.polarity
self.currentTone = 0.65 * self.currentTone
self.currentUnderstanding = 0.65 * self.currentUnderstanding
self.currentBot = 0.65 * self.currentBot
if self.questionsAsked >= 1:
score1 = ((tone_score + 1) * 5)
score2 = ((understanding_score + 1) * 5)
score0 = score_maker(self.questions[-2], self.answers[-1], model="gpt-3.5-turbo", max_tokens=1)
# print(score0)
if score0.isdigit() and score0 is not None and score0 != "0" and score0 != "0.0":
self.scoreCount_0 += 1
self.currentBot = float(score0)
self.scoreByBot += self.currentBot
if score1 > 0.5:
self.scoreCount_1 += 1
self.currentTone = score1
if score2 > 0.5:
self.scoreCount_2 += 1
self.currentUnderstanding = score2
self.scoreByTone += score1
self.scoreByAnswer += score2
self.questionsAsked += 1
# Checking Tone
if self.questionsAsked > 5 and self.scoreByTone < 2:
return -1
print(reply)
return reply
def get_report_data(self):
format_style = 'Report:: Candidate Background: , Strengths: ,Areas To Improve: ,Recomendations: ,'
report = report_maker(self.questions, format_style, self.answers, model="gpt-3.5-turbo", max_tokens=250)
return report
def run(self,message):
if message:
try:
scores = []
response = self.process_message(message)
if response == 0:
BotAnswer = "Interview stopped. Thank you for your time."
BotStatus = 0
elif response == 1:
BotAnswer = "Interview completed. It was great interviewing you\n"
BotStatus = 0
elif response == -1:
response = "Interview stopped. \nUnfortunately, we can't continue with the interview at this time due to the tone of our interaction. Thank you for your time."
elif response == -2:
return "Something went wrong. Please try again."
else:
BotAnswer = response
BotStatus = 1
if BotStatus == 0:
scores.append(round(safe_division(self.scoreByTone, self.scoreCount_1), 1))
scores.append(round(safe_division(self.scoreByAnswer, self.scoreCount_2), 1))
scores.append(round(safe_division(self.scoreByBot, self.scoreCount_0), 1))
elif BotStatus == 1:
scores.append(round(self.currentTone, 1))
scores.append(round(self.currentUnderstanding, 1))
scores.append(round(self.currentBot, 1))
return BotAnswer,BotStatus, scores
except Exception as e:
print(e)
print("Something went wrong. Please try again. dfjndfsdjf")