Spaces:
Sleeping
Sleeping
new file
Browse files
app.py
CHANGED
|
@@ -5,36 +5,123 @@ from tools import sentiment_analysis_util
|
|
| 5 |
import numpy as np
|
| 6 |
from dotenv import load_dotenv
|
| 7 |
import os
|
| 8 |
-
st.title("💬 Chatbot")
|
| 9 |
-
st.caption("")
|
| 10 |
|
|
|
|
|
|
|
|
|
|
| 11 |
|
| 12 |
-
|
| 13 |
-
# Initialize session state for storing messages if it doesn't already exist
|
| 14 |
-
if "messages" not in st.session_state:
|
| 15 |
-
st.session_state["messages"] = [{"role": "assistant", "content": "How can I help you?"}]
|
| 16 |
|
| 17 |
-
|
| 18 |
-
for msg in st.session_state.messages:
|
| 19 |
-
st.chat_message(msg["role"]).write(msg["content"])
|
| 20 |
|
| 21 |
-
|
| 22 |
-
prompt = st.chat_input("Enter your question:")
|
| 23 |
-
if prompt:
|
| 24 |
-
if not openai_api_key:
|
| 25 |
-
st.error("No OpenAI API key found. Please set the OPENAI_API_KEY environment variable.")
|
| 26 |
-
st.stop()
|
| 27 |
|
| 28 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 29 |
st.session_state.messages.append({"role": "user", "content": prompt})
|
| 30 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 31 |
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
|
| 38 |
-
# Append and display the assistant's response
|
| 39 |
-
st.session_state.messages.append({"role": "assistant", "content": msg})
|
| 40 |
-
st.chat_message("assistant").write(msg)
|
|
|
|
| 5 |
import numpy as np
|
| 6 |
from dotenv import load_dotenv
|
| 7 |
import os
|
|
|
|
|
|
|
| 8 |
|
| 9 |
+
st.set_page_config(page_title="LangChain Agent", layout="wide")
|
| 10 |
+
load_dotenv()
|
| 11 |
+
OPENAI_API_KEY = os.environ["OPENAI_API_KEY"]
|
| 12 |
|
| 13 |
+
llm = ChatOpenAI(model="gpt-3.5-turbo")
|
|
|
|
|
|
|
|
|
|
| 14 |
|
| 15 |
+
from langchain_core.runnables import RunnableConfig
|
|
|
|
|
|
|
| 16 |
|
| 17 |
+
st.title("💬 ExpressMood")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
|
| 19 |
+
#@st.cache_resource
|
| 20 |
+
if "chat_history" not in st.session_state:
|
| 21 |
+
st.session_state["messages"] = [{"role":"system", "content":"""
|
| 22 |
+
You are a sentiment analysis expert. Answer all questions related to cryptocurrency investment reccommendations. Say I don't know if you don't know.
|
| 23 |
+
"""}]
|
| 24 |
+
|
| 25 |
+
if 'count' not in st.session_state:
|
| 26 |
+
st.session_state['count'] = 0
|
| 27 |
+
|
| 28 |
+
# If not, then initialize it
|
| 29 |
+
if 'key' not in st.session_state:
|
| 30 |
+
st.session_state['key'] = 'value'
|
| 31 |
+
|
| 32 |
+
# Session State also supports the attribute based syntax
|
| 33 |
+
if 'key' not in st.session_state:
|
| 34 |
+
st.session_state.key = 'value'
|
| 35 |
+
|
| 36 |
+
#st.image('el_pic.png')
|
| 37 |
+
|
| 38 |
+
sideb=st.sidebar
|
| 39 |
+
with st.sidebar:
|
| 40 |
+
prompt=st.text_input("Enter topic for sentiment analysis: ")
|
| 41 |
+
|
| 42 |
+
check1=sideb.button(f"analyze {prompt}")
|
| 43 |
+
|
| 44 |
+
if check1:
|
| 45 |
+
# Add user message to chat history
|
| 46 |
st.session_state.messages.append({"role": "user", "content": prompt})
|
| 47 |
+
# Display user message in chat message container
|
| 48 |
+
with st.chat_message("user"):
|
| 49 |
+
st.markdown(prompt)
|
| 50 |
+
|
| 51 |
+
# ========================== Sentiment analysis
|
| 52 |
+
#Perform sentiment analysis on the cryptocurrency news & predict dominant sentiment along with plotting the sentiment breakdown chart
|
| 53 |
+
# Downloading from reddit
|
| 54 |
+
|
| 55 |
+
# Downloading from alpaca
|
| 56 |
+
if len(prompt.split(' '))<2:
|
| 57 |
+
print('here')
|
| 58 |
+
st.write('I am analyzing Google News ...')
|
| 59 |
+
news_articles = sentiment_analysis_util.fetch_news(str(prompt))
|
| 60 |
+
st.write('Now, I am analyzing Reddit ...')
|
| 61 |
+
reddit_news_articles=sentiment_analysis_util.fetch_reddit_news(prompt)
|
| 62 |
+
analysis_results = []
|
| 63 |
+
|
| 64 |
+
#Perform sentiment analysis for each product review
|
| 65 |
+
if len(prompt.split(' '))<2:
|
| 66 |
+
print('here')
|
| 67 |
+
for article in news_articles:
|
| 68 |
+
if prompt.lower()[0:6] in article['News_Article'].lower():
|
| 69 |
+
sentiment_analysis_result = sentiment_analysis_util.analyze_sentiment(article['News_Article'])
|
| 70 |
|
| 71 |
+
# Display sentiment analysis results
|
| 72 |
+
#print(f'News Article: {sentiment_analysis_result["News_Article"]} : Sentiment: {sentiment_analysis_result["Sentiment"]}', '\n')
|
| 73 |
+
|
| 74 |
+
result = {
|
| 75 |
+
'News_Article': sentiment_analysis_result["News_Article"],
|
| 76 |
+
'Sentiment': sentiment_analysis_result["Sentiment"][0]['label'],
|
| 77 |
+
'Index': sentiment_analysis_result["Sentiment"][0]['score'],
|
| 78 |
+
'URL': article['URL']
|
| 79 |
+
}
|
| 80 |
+
|
| 81 |
+
analysis_results.append(result)
|
| 82 |
+
|
| 83 |
+
articles_url=[]
|
| 84 |
+
for article in reddit_news_articles:
|
| 85 |
+
if prompt.lower()[0:6] in article.lower():
|
| 86 |
+
sentiment_analysis_result_reddit = sentiment_analysis_util.analyze_sentiment(article)
|
| 87 |
+
|
| 88 |
+
# Display sentiment analysis results
|
| 89 |
+
#print(f'News Article: {sentiment_analysis_result_reddit["News_Article"]} : Sentiment: {sentiment_analysis_result_reddit["Sentiment"]}', '\n')
|
| 90 |
+
|
| 91 |
+
result = {
|
| 92 |
+
'News_Article': sentiment_analysis_result_reddit["News_Article"],
|
| 93 |
+
'Index':np.round(sentiment_analysis_result_reddit["Sentiment"][0]['score'],2)
|
| 94 |
+
}
|
| 95 |
+
analysis_results.append(np.append(result,np.append(article.split('URL:')[-1:], ((article.split('Date: ')[-1:])[0][0:10]))))
|
| 96 |
+
#pd.DataFrame(analysis_results).to_csv('analysis_results.csv')
|
| 97 |
+
|
| 98 |
+
#Generate summarized message rationalize dominant sentiment
|
| 99 |
+
summary = sentiment_analysis_util.generate_summary_of_sentiment(analysis_results) #, dominant_sentiment)
|
| 100 |
+
st.chat_message("assistant").write((summary))
|
| 101 |
+
st.session_state.messages.append({"role": "assistant", "content": summary})
|
| 102 |
+
#answers=np.append(res["messages"][-1].content,summary)
|
| 103 |
+
|
| 104 |
+
client = OpenAI(api_key=OPENAI_API_KEY)
|
| 105 |
+
|
| 106 |
+
if "openai_model" not in st.session_state:
|
| 107 |
+
st.session_state["openai_model"] = "gpt-3.5-turbo"
|
| 108 |
+
|
| 109 |
+
if prompt := st.chat_input("Any other questions? "):
|
| 110 |
+
# Add user message to chat history
|
| 111 |
+
st.session_state.messages.append({"role": "user", "content": prompt})
|
| 112 |
+
# Display user message in chat message container
|
| 113 |
+
with st.chat_message("user"):
|
| 114 |
+
st.markdown(prompt)
|
| 115 |
+
# Display assistant response in chat message container
|
| 116 |
+
with st.chat_message("assistant"):
|
| 117 |
+
stream = client.chat.completions.create(
|
| 118 |
+
model=st.session_state["openai_model"],
|
| 119 |
+
messages=[
|
| 120 |
+
{"role": m["role"], "content": m["content"]}
|
| 121 |
+
for m in st.session_state.messages
|
| 122 |
+
],
|
| 123 |
+
stream=True,
|
| 124 |
+
)
|
| 125 |
+
response = st.write_stream(stream)
|
| 126 |
+
st.session_state.messages.append({"role": "assistant", "content": response})
|
| 127 |
|
|
|
|
|
|
|
|
|