File size: 2,578 Bytes
e370f3a d1406e8 51ce63b d1406e8 270b445 efbfb8b bb68426 3868931 270b445 51ce63b 270b445 efbfb8b 3868931 e370f3a b3766ed 51ce63b 3868931 39f5430 270b445 3868931 e370f3a 3868931 a4b560c bb68426 3868931 e370f3a 3868931 e370f3a 0527279 3868931 0527279 e07b81a e370f3a efa5f27 51ce63b bb68426 a4b560c 51ce63b a4b560c 270b445 e370f3a 51ce63b e370f3a 3868931 e370f3a e07b81a e370f3a e07b81a e370f3a 3868931 e370f3a 3868931 e370f3a 3868931 e370f3a 3868931 e370f3a |
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 |
import gradio as gr
import requests
import os
import faiss
import numpy as np
import json
from sentence_transformers import SentenceTransformer
# ✅ Load context
with open("texts.json", "r", encoding="utf-8") as f:
texts = json.load(f)
index = faiss.read_index("faiss_index.bin")
embed_model = SentenceTransformer("all-MiniLM-L6-v2")
# ✅ API Setup
API_KEY = os.environ.get("OPENROUTER_API_KEY")
MODEL = "meta-llama/llama-3.3-8b-instruct:free"
# ✅ Context search
def get_context(query, top_k=5, min_score=0.05):
query_vec = embed_model.encode([query])
D, I = index.search(np.array(query_vec), top_k)
# If top score is too low, reject
if D[0][0] < min_score:
return None
return "\n".join([texts[i] for i in I[0]])
# ✅ Chat function
def chat_fn(message, history):
context = get_context(message)
if context is None:
return history[:-1] + [(message, "❌ Sorry! I cannot answer that.")]
headers = {
"Authorization": f"Bearer {API_KEY}",
"Content-Type": "application/json"
}
messages = [
{
"role": "system",
"content": f"You are a helpful assistant. Use ONLY this context to answer:\n{context}\nIf not found in context, reply: 'Sorry! I cannot answer that.'"
}
]
for user, assistant in history[:-1]:
messages.append({"role": "user", "content": user})
messages.append({"role": "assistant", "content": assistant})
messages.append({"role": "user", "content": message})
payload = {
"model": MODEL,
"messages": messages
}
try:
response = requests.post("https://openrouter.ai/api/v1/chat/completions", headers=headers, json=payload)
response.raise_for_status()
reply = response.json()["choices"][0]["message"]["content"]
except Exception as e:
reply = f"❌ Error: {e}"
return history[:-1] + [(message, reply)]
# ✅ UI
with gr.Blocks(css="""
.footer {display: none !important;}
#chat-window {height: 500px !important; overflow-y: auto;}
""") as demo:
chatbot = gr.Chatbot(elem_id="chat-window")
state = gr.State([])
with gr.Row():
msg = gr.Textbox(placeholder="Type your message and press enter...", scale=9)
# Typing sim
def user_send(message, history):
return "", history + [(message, "⏳ ...")]
def complete_chat(message, history):
return chat_fn(message, history)
msg.submit(user_send, [msg, state], [msg, chatbot]).then(
complete_chat, [msg, state], [chatbot]
)
demo.launch()
|