File size: 5,416 Bytes
65e3d9a 1f9735b 9b4432e 1f9735b bc45c0c ad1351e 1f9735b 987341e 6cbf495 2824ce7 6cbf495 2824ce7 9b4432e 2824ce7 9b4432e 2824ce7 6cbf495 1f9735b e0c7e62 1f9735b 57060b1 1f9735b |
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 |
import streamlit as st
from bs4 import BeautifulSoup
from openai import OpenAI
from dotenv import load_dotenv
import os
from streamlit.components.v1 import html
load_dotenv()
os.makedirs(".streamlit", exist_ok=True)
hf_api_key = os.getenv("SECRET")
if not hf_api_key:
st.error("❌ API key not found. Please set HUGGINGFACE_API_KEY in your .env file.")
st.stop()
# ==== Initialize Client ====
client = OpenAI(
base_url="https://router.huggingface.co/v1",
api_key=hf_api_key,
)
# ==== Streamlit Page Setup ====
st.set_page_config(page_title="OSS ChatGPT", layout="wide")
st.title("🤖 ChatGPT")
# ==== Sidebar ====
st.sidebar.title("🛠️ Settings")
model_choice = st.sidebar.selectbox("Choose a model", [
"openai/gpt-oss-20b:hyperbolic",
"openai/gpt-oss-120b:hyperbolic"
])
# if st.sidebar.button("🧹 Clear Chat"):
# st.session_state.messages = []
# st.rerun()
def clear_chat():
# COMPLETELY reset session state (not just messages)
st.session_state.clear() # 🔥 This wipes ALL session variables
# Then re-initialize just what you need
st.session_state.messages = []
# Force a hard refresh (JavaScript method)
js = """
<script>
window.parent.location.reload(true);
</script>
"""
html(js)
st.sidebar.button(
"🧹 Clear Chat",
on_click=clear_chat,
key="clear_chat_button_unique_123" # Unique key prevents duplicate ID errors
)
# ==== Session Initialization ====
if "messages" not in st.session_state:
st.session_state.messages = []
# ==== LaTeX Rendering Helper ====
def render_markdown_with_latex(text: str):
mathjax_script = """
<script type="text/javascript">
MathJax = {
tex: {
inlineMath: [['$', '$']],
displayMath: [['$$', '$$']]
},
svg: {
fontCache: 'global'
}
};
</script>
<script async src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-svg.js"></script>
"""
st.markdown(mathjax_script, unsafe_allow_html=True)
st.markdown(text, unsafe_allow_html=True)
for i, msg in enumerate(st.session_state.messages):
with st.chat_message(msg["role"]):
if msg["role"] == "assistant":
# Check if content contains thinking pattern
if msg["content"].startswith("<think>") and "</think>" in msg["content"]:
# Extract thinking and response parts
thinking_part = msg["content"].split("<think>")[1].split("</think>")[0].strip()
response_part = msg["content"].split("</think>")[1].strip()
# Display the main response
render_markdown_with_latex(response_part)
# Create expander for thinking
with st.expander("💭 Show Thinking"):
st.markdown(f"<span style='color: #666; font-style: italic'>{thinking_part}</span>",
unsafe_allow_html=True)
else:
render_markdown_with_latex(msg["content"])
else:
# Display user messages
render_markdown_with_latex(msg["content"])
# ==== Chat Input
if prompt := st.chat_input("Type your message...",key="unique_chat_input_key"):
print(f"=> {prompt}")
with st.chat_message("user"):
st.markdown(prompt)
# Add user message to history
st.session_state.messages.append({"role": "user", "content": prompt})
try:
with st.chat_message("assistant"):
response_placeholder = st.empty()
with st.spinner("Thinking..."):
completion = client.chat.completions.create(
model=model_choice,
messages=[
{"role": m["role"], "content": m["content"]}
for m in st.session_state.messages
]
)
raw = completion.choices[0].message.content
# Parse thinking and response
if "<think>" in raw and "</think>" in raw:
thinking_part = raw.split("<think>")[1].split("</think>")[0].strip()
response_part = raw.split("</think>")[1].strip()
# Display main response
clean_response = BeautifulSoup(response_part, "html.parser").get_text()
with st.expander("💭 Show Thinking"):
st.markdown(f"<span style='color: #666; font-style: italic'>{thinking_part}</span>",
unsafe_allow_html=True)
response_placeholder.markdown(clean_response)
# Store both parts in history
full_content = f"<think>{thinking_part}</think>{clean_response}"
else:
clean_response = BeautifulSoup(raw, "html.parser").get_text()
response_placeholder.markdown(clean_response)
full_content = clean_response
# Add assistant response to history
st.session_state.messages.append({"role": "assistant", "content": full_content})
except Exception as e:
error_msg = f"❌ Error: {str(e)}"
with st.chat_message("assistant"):
st.error(error_msg)
st.session_state.messages.append({"role": "assistant", "content": error_msg}) |