Spaces:
Build error
Build error
File size: 3,362 Bytes
21a2e46 1051bf9 2d8c319 8550dc5 91b268b 2e62dd1 91b268b 21a2e46 0e5b4a4 2e62dd1 569e45d 2e62dd1 2d8c319 0e5b4a4 21a2e46 2e62dd1 21a2e46 8550dc5 21a2e46 8550dc5 0e5b4a4 2e62dd1 21a2e46 2e62dd1 21a2e46 f8c1ecf 0e5b4a4 2e62dd1 f8c1ecf 21a2e46 f8c1ecf 21a2e46 f8c1ecf 21a2e46 8550dc5 f8c1ecf 8550dc5 0e5b4a4 f8c1ecf 21a2e46 8550dc5 21a2e46 8550dc5 2e62dd1 91b268b f8c1ecf 8550dc5 2e62dd1 8550dc5 0e5b4a4 8550dc5 |
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 |
# app.py
import gradio as gr
import os
import torch
from langchain_community.document_loaders import PyPDFLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain_community.vectorstores import FAISS
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
# Configuration
DOCS_DIR = "business_docs"
EMBEDDING_MODEL = "sentence-transformers/all-MiniLM-L6-v2"
MODEL_NAME = "microsoft/phi-2"
# Quantization config
quant_config = BitsAndBytesConfig(
load_in_4bit=True,
bnb_4bit_quant_type="nf4",
bnb_4bit_compute_dtype=torch.float16,
bnb_4bit_use_double_quant=False
)
def initialize_system():
# Document processing
if not os.path.exists(DOCS_DIR):
raise FileNotFoundError(f"Missing {DOCS_DIR} folder")
pdf_files = [os.path.join(DOCS_DIR, f)
for f in os.listdir(DOCS_DIR)
if f.endswith(".pdf")]
text_splitter = RecursiveCharacterTextSplitter(
chunk_size=800,
chunk_overlap=100
)
texts = []
for pdf in pdf_files:
loader = PyPDFLoader(pdf)
pages = loader.load_and_split(text_splitter)
texts.extend(pages)
# Create embeddings
embeddings = HuggingFaceEmbeddings(
model_name=EMBEDDING_MODEL,
model_kwargs={'device': 'cpu'}
)
# Vector store
vector_store = FAISS.from_documents(texts, embeddings)
# Model loading
tokenizer = AutoTokenizer.from_pretrained(
MODEL_NAME,
trust_remote_code=True,
padding_side="left"
)
model = AutoModelForCausalLM.from_pretrained(
MODEL_NAME,
trust_remote_code=True,
device_map="auto",
quantization_config=quant_config,
torch_dtype=torch.float16
)
return vector_store, model, tokenizer
try:
vector_store, model, tokenizer = initialize_system()
print("✅ System initialized successfully")
except Exception as e:
print(f"❌ Initialization failed: {str(e)}")
raise
def generate_response(query):
try:
docs = vector_store.similarity_search(query, k=2)
context = "\n".join([d.page_content for d in docs])
prompt = f"""<|system|>
Answer using only this context: {context}
- Max 2 sentences
- If unsure: "I'll check with the team"</s>
<|user|>{query}</s>
<|assistant|>"""
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(
**inputs,
max_new_tokens=150,
temperature=0.1,
pad_token_id=tokenizer.eos_token_id
)
return tokenizer.decode(outputs[0], skip_special_tokens=True).split("<|assistant|>")[-1].strip()
except Exception as e:
return "Please try again later."
# Gradio interface
with gr.Blocks() as demo:
gr.Markdown("# Customer Service Chatbot")
chatbot = gr.Chatbot()
msg = gr.Textbox(label="Your question")
clear = gr.ClearButton([msg, chatbot])
def respond(message, history):
response = generate_response(message)
history.append((message, response))
return "", history
msg.submit(respond, [msg, chatbot], [msg, chatbot])
demo.launch() |