Spaces:
Sleeping
Sleeping
File size: 3,993 Bytes
6c7823c |
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 |
import streamlit as st
import requests
from deepgram import (
DeepgramClient,
DeepgramClientOptions,
PrerecordedOptions,
FileSource,
)
from io import BytesIO
import os
# Page configuration
st.set_page_config(page_title="Market Brief Chat", page_icon="💬", layout="wide")
DG_API = os.getenv("DG_API")
# Initialize session state for chat history
if "messages" not in st.session_state:
st.session_state.messages = [
{"role": "assistant", "content": "How can I help you today?"}
]
if "um" not in st.session_state:
st.session_state.um = None
# Initialize other session state variables
if "is_recording" not in st.session_state:
st.session_state.is_recording = False
def STT(buffer):
config: DeepgramClientOptions = DeepgramClientOptions(api_key=DG_API)
deepgram: DeepgramClient = DeepgramClient("", config)
payload: FileSource = {
"buffer": buffer,
}
# STEP 2: Configure Deepgram options for audio analysis
options = PrerecordedOptions(
model="nova-3",
smart_format=True,
)
# STEP 3: Call the transcribe_file method with the text payload and options
response = deepgram.listen.rest.v("1").transcribe_file(payload, options)
data = response.to_json()
transcript = data["results"]["channels"][0]["alternatives"][0]["transcript"]
return transcript
def TTS(text):
DEEPGRAM_URL = "https://api.deepgram.com/v1/speak?model=aura-2-thalia-en"
DEEPGRAM_API_KEY = DG_API
payload = {"text": text}
headers = {
"Authorization": f"Token {DEEPGRAM_API_KEY}",
"Content-Type": "application/json",
}
# Create a BytesIO buffer to store audio
audio_buffer = BytesIO()
response = requests.post(DEEPGRAM_URL, headers=headers, json=payload)
audio_buffer.write(response.content)
# Move cursor to the beginning of the buffer
audio_buffer.seek(0)
return audio_buffer
# App title
st.markdown("<h1>💬 Market Brief Chat</h1>", unsafe_allow_html=True)
st.markdown("---")
# Display chat history
chat_col, audio_col = st.columns([0.85, 0.15])
with chat_col:
c = st.container(height=400, border=True)
with c:
for message in st.session_state.messages:
with st.chat_message(message["role"]):
st.write(message["content"])
with audio_col:
# Voice input buttonif
data = st.audio_input(label="🎤 Record")
if data:
st.session_state.um = STT(data)
if st.button("🔊 listen"):
text = st.session_state.messages[-1]["content"]
buffer = TTS(text)
st.audio(data=buffer)
st.success("Playing")
# Voice input button (beside text input conceptually)
# Handle text input
user_input = st.chat_input("Ask me about market trends...")
st.session_state.um = user_input
if st.session_state.um:
with c:
with st.chat_message("user"):
st.markdown(user_input)
st.session_state.messages.append(
{"role": "user", "content": st.session_state.um}
)
data = {"query": user_input, "history": []}
or_response = requests.post(
url="http://127.0.0.1:8000/orchestrator/orchestrator_decision", json=data
).json()
print(or_response)
agent_response = ""
data = {"query": user_input, "context": or_response["response"]}
full_response = requests.post(
url="http://127.0.0.1:8000/orchestrator/final_response",
json=data,
stream=True,
)
with st.chat_message("assistant"):
placeholder = st.empty()
for chunk in full_response.iter_content(
decode_unicode=True, chunk_size=None
):
agent_response += chunk
placeholder.markdown(agent_response + "▌")
st.session_state.messages.append(
{"role": "assistant", "content": agent_response}
)
st.session_state.um = None
st.rerun()
|