Spaces:
Runtime error
Runtime error
File size: 5,769 Bytes
8b79aed e80121c 8b79aed 2d6ef72 8b79aed 177f919 8b79aed 177f919 8b79aed 287750f 8b79aed 5039b99 8b79aed 26f32ee 8b79aed 2d6ef72 8b79aed 177f919 8b79aed 26f32ee 5039b99 26f32ee 5039b99 287750f 5039b99 207ae44 5039b99 207ae44 287750f 5039b99 8b79aed 2d6ef72 8b79aed e80121c 2d6ef72 e80121c 8b79aed 5039b99 |
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 |
import json
import time
from transformers import AutoTokenizer, AutoModel
from langchain_community.chat_models import ChatOpenAI
import pandas as pd
from config import settings
from langchain_core.utils.function_calling import convert_to_openai_function
from langchain.prompts import ChatPromptTemplate, MessagesPlaceholder
from langchain.memory import ConversationBufferWindowMemory
from langchain.schema.runnable import RunnablePassthrough
from langchain.agents.format_scratchpad import format_to_openai_functions
from langchain.agents.output_parsers import OpenAIFunctionsAgentOutputParser
from langchain.agents import AgentExecutor
from tools import MEMORY, set_chatbot_name, close_chat, recommand_podcast,app_features,joke_teller,SESSION_ID
from database_functions import get_chat_bot_name,get_chat_history, get_last_conversion, get_last_session, get_mood_data
def get_mood_summary(user_id):
data = get_mood_data(user_id)
system_prompt = """You are an descripting assistant that provides the breif description of the user data which is related to their mood tracking activity. Properly descibe the reason for their mood.Avoid times and dates in description
Here is the user data: {data}"""
llm = ChatOpenAI(model=settings.OPENAI_MODEL,
openai_api_key=settings.OPENAI_KEY, temperature=0.0)
return llm.invoke(system_prompt.format(data=data)).content
def deanonymizer(input, anonymizer):
input = anonymizer.deanonymize(input)
map = anonymizer.deanonymizer_mapping
if map:
for k in map["PERSON"]:
names = k.split(" ")
for i in names:
input = input.replace(i, map["PERSON"][k])
return input
def get_last_session_summary(last_session_id, second_last_session_id):
conversation = get_last_conversion(last_session_id,second_last_session_id)
if conversation:
system_prompt = """ context: there is one typical conversation going on between two high school gen z girls.
you are one of the high school gen z girl. your voice is edgy and raw.
must use I for AI named BMOXI and for human use my friend. in summary.
this is your conversation with your best friend. summerize whole conversation and if you found any question can be asked from chat then return summary and qustion else return summary and append None at last.
conversation: {conversation}
summary:
"""
llm = ChatOpenAI(model=settings.OPENAI_MODEL,
openai_api_key=settings.OPENAI_KEY, temperature=0.0)
response = llm.invoke(system_prompt.format(conversation=conversation)).content
return response
else:
return ""
def create_agent(user_id,is_first = False):
# print("get user Id**********************",user_id)
previous_session_id = get_last_session(user_id)
# print(previous_session_id)
tools = [set_chatbot_name,close_chat,recommand_podcast,app_features,joke_teller]
functions = [convert_to_openai_function(f) for f in tools]
model = ChatOpenAI(model_name=settings.OPENAI_MODEL,
openai_api_key=settings.OPENAI_KEY, frequency_penalty= 1, temperature=0.7).bind(functions=functions)
chat_bot_name = get_chat_bot_name(user_id)
extra_prompt = ""
previous_problem_summary = None
if is_first:
start = time.time()
mood_summary = get_mood_summary(user_id)
print(previous_session_id)
if previous_session_id['second_last_session_id']:
previous_problem_summary = get_last_session_summary(previous_session_id['last_session_id'], previous_session_id['second_last_session_id'])
print("time require for mood summary: ",time.time()-start)
if previous_problem_summary.find('None') == -1:
extra_prompt = f"""ask user her previous problem is solved or not.use previous problem summary for framming the question. must include her name which is {user_id} .nothing else."""
else:
extra_prompt = f""" Only use these templates to start conversation:-
1. Hey again {user_id}! How's it going?
2. hey again {user_id}! What's up today? Need ✨ Advice, ✨ a Mood Boost, ✨ a Chat, ✨ Resource Suggestions, ✨ App Features help? How can I help?"
use any one of the question for response based on your understanding not use anything else simply return one of these two only.
"""
prompt = ChatPromptTemplate.from_messages([("system", settings.SYSTEM_PROMPT.format(name = chat_bot_name, mood="", previous_summary=previous_problem_summary)+extra_prompt),
MessagesPlaceholder(variable_name="chat_history"), ("user", "{input}"),
MessagesPlaceholder(variable_name="agent_scratchpad")])
memory = ConversationBufferWindowMemory(memory_key="chat_history", chat_memory=get_chat_history(
previous_session_id['last_session_id']), return_messages=True, k=5)
# print("memory created")
global MEMORY,SESSION_ID
MEMORY = memory
SESSION_ID = previous_session_id['last_session_id']
chain = RunnablePassthrough.assign(agent_scratchpad=lambda x: format_to_openai_functions(x["intermediate_steps"])) | prompt | model | OpenAIFunctionsAgentOutputParser()
agent_executor = AgentExecutor(
agent=chain, tools=tools, memory=memory, verbose=True)
return agent_executor
|