langchain-huggingface==0.1.2
torch==2.6.0
loguru==0.7.3
chromadb==0.6.3
sentence-transformers==3.4.1
langchain-chroma==0.2.2
pydantic-settings==2.8.1
langchain-text-splitters==0.3.7
langchain-deepseek==0.1.3
langchain-openai==0.3.11

SHOP_DATA = [
    {
        "text": 'Portátil Lenovo IdeaPad 5: 16 Gb RAM, SSD 512 Gb, LED 15.6", precio 550 $.',
        "metadata": {
            "id": "1",
            "type": "product",
            "category": "laptops",
            "price": 55000,
            "stock": 3,
        },
    },
    {
        "text": 'Smartphone Xiaomi Redmi Note 12: 128 Gb RAM, SSD 512 Gb, LED 15.6", precio 1550 $.',
        "metadata": {
            "id": "2",
            "type": "product",
            "category": "phones",
            "price": 18000,
            "stock": 10,
        },
    },
]

from langchain_huggingface import HuggingFaceEmbeddings
import torch
from langchain_chroma import Chroma

CHROMA_PATH = "./shop_chroma_db"
COLLECTION_NAME = "shop_data"

def generate_chroma_db():
    embeddings = HuggingFaceEmbeddings(
        model_name="sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2",
        model_kwargs={"device": "cuda" if torch.cuda.is_available() else "cpu"},
        encode_kwargs={"normalize_embeddings": True},
    )

    chroma_db = Chroma.from_texts(
        texts=[item["text"] for item in SHOP_DATA],
        embedding=embeddings,
        ids=[str(item["metadata"]["id"]) for item in SHOP_DATA],
        metadatas=[item["metadata"] for item in SHOP_DATA],
        persist_directory=CHROMA_PATH,
        collection_name=COLLECTION_NAME,
    )
    return chroma_db

if __name__ == "__main__":
    db = generate_chroma_db()

def connect_to_chroma():
    embeddings = HuggingFaceEmbeddings(
        model_name="sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2",
        model_kwargs={"device": "cuda" if torch.cuda.is_available() else "cpu"},
        encode_kwargs={"normalize_embeddings": True},
    )
    chroma_db = Chroma(
        persist_directory=CHROMA_PATH,
        embedding_function=embeddings,
        collection_name=COLLECTION_NAME,
    )
    return chroma_db


def search_products(query: str, k: int = 4, metadata_filter: dict = None):
    chroma_db = connect_to_chroma()
    results = chroma_db.similarity_search_with_score(
        query, k=k, filter=metadata_filter
    )
    formatted_results = [
        {"text": doc.page_content, "metadata": doc.metadata, "similarity_score": score}
        for doc, score in results
    ]
    return formatted_results

filter_condition = {
    "category": "laptops",
    "price": {"$lt": 600}
}
results = search_products("notebook", k=5, metadata_filter=filter_condition)

from langchain_deepseek import ChatDeepSeek
from langchain_openai import ChatOpenAI
from langchain_chroma import Chroma
from langchain_huggingface import HuggingFaceEmbeddings

class ChatWithAI:
    def __init__(self, provider="deepseek"):
        self.provider = provider
        self.embeddings = HuggingFaceEmbeddings(
            model_name="sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2",
            model_kwargs={"device": "cuda" if torch.cuda.is_available() else "cpu"},
            encode_kwargs={"normalize_embeddings": True},
        )
        if provider == "deepseek":
            self.llm = ChatDeepSeek(api_key="YOUR_DEEPSEEK_API_KEY", model="deepseek-chat")
        elif provider == "openai":
            self.llm = ChatOpenAI(api_key="YOUR_OPENAI_API_KEY", model="gpt-3.5-turbo")
        else:
            raise ValueError("Unsupported provider")
        self.chroma_db = Chroma(persist_directory=CHROMA_PATH, embedding_function=self.embeddings, collection_name=COLLECTION_NAME)

    def get_relevant_context(self, query, k=3):
        results = self.chroma_db.similarity_search(query, k=k)
        return [{"text": doc.page_content, "metadata": doc.metadata} for doc in results]

    def format_context(self, context):
        formatted = []
        for item in context:
            metadata_str = "\n".join(f"{k}: {v}" for k, v in item["metadata"].items())
            formatted.append(f"Text: {item['text']}\nMetadata:\n{metadata_str}\n")
        return "\n---\n".join(formatted)

    def generate_response(self, query):
        context = self.get_relevant_context(query)
        if not context:
            return "Xin lỗi, không tìm thấy ngữ cảnh phù hợp."
        formatted_context = self.format_context(context)
        messages = [
            {"role": "system", "content": "Bạn là trợ lý AI giúp trả lời chính xác câu hỏi về cơ sở dữ liệu."},
            {"role": "user", "content": f"Câu hỏi: {query}\nNgữ cảnh: {formatted_context}"}
        ]
        response = self.llm.invoke(messages)
        return response.content if hasattr(response, "content") else str(response)

if __name__ == "__main__":
    chat = ChatWithAI(provider="deepseek")
    print("=== Bắt đầu hội thoại với AI ===")
    while True:
        user_input = input("Bạn: ")
        if user_input.lower() == "exit":
            print("Kết thúc phiên trò chuyện.")
            break
        print("AI đang suy nghĩ...", end="\r")
        answer = chat.generate_response(user_input)
        print(f"AI: {answer}\n")