Spaces:
Running
Running
File size: 2,189 Bytes
6517515 3cfcbf6 6517515 |
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 |
import streamlit as st
from huggingface_hub import InferenceClient
from dotenv import load_dotenv
import os
# Load .env file
load_dotenv()
# Get API key from environment variable
api_key = os.getenv("HUGGINGFACEHUB_API_TOKEN")
st.set_page_config(page_title="Intellicounsel AI Chat", page_icon="🤖")
# Add system prompt once at the start
system_prompt = {
"role": "system",
"content": (
"You are Intellicounsel, an intelligent and friendly AI advisor that helps students "
"with college applications, SOP reviews, resume tips, and academic advice. "
"Respond clearly and helpfully, always tailored to the student's needs."
)
}
# Initialize chat history
if "messages" not in st.session_state:
st.session_state.messages = []
# Title
st.title("🧠 Intellicounsel — AI College Advisor")
# User input
user_input = st.chat_input("Ask something like SOP tips or university suggestions...")
# Show past messages
for msg in st.session_state.messages:
with st.chat_message(msg["role"]):
st.markdown(msg["content"])
# If new input
if user_input:
st.session_state.messages.append({"role": "user", "content": user_input})
with st.chat_message("user"):
st.markdown(user_input)
with st.chat_message("assistant"):
with st.spinner("Thinking..."):
try:
client = InferenceClient(
model="nvidia/Llama-3_1-Nemotron-Ultra-253B-v1",
provider="nebius",
api_key=api_key,
)
# Add system prompt only once at the start of the context
full_context = [system_prompt] + st.session_state.messages
completion = client.chat.completions.create(
model="nvidia/Llama-3_1-Nemotron-Ultra-253B-v1",
messages=full_context,
max_tokens=2048,
)
response = completion.choices[0].message.content
except Exception as e:
response = f"❌ Error: {e}"
st.markdown(response)
st.session_state.messages.append({"role": "assistant", "content": response})
|