app.py

import os
import sys
import copy
import time
import numpy as np
import streamlit as st
from typing import Optional
from stqdm import stqdm 
from numpy import ndarray
from typing import Iterable
from qdrant_client import QdrantClient, models
from fastembed.sparse.splade_pp import supported_splade_models
from fastembed import SparseTextEmbedding, SparseEmbedding
from langchain_community.llms.exllamav2 import ExLlamaV2
from langchain_core.callbacks import CallbackManager, StreamingStdOutCallbackHandler
from fastembed_ext import FastEmbedEmbeddingsLc
from langchain_community.document_loaders.wikipedia import WikipediaLoader
from langchain_community.document_loaders.unstructured import UnstructuredFileLoader
from langchain_experimental.text_splitter import SemanticChunker
from langchain_core.documents import Document
from qdrant_client.models import (
    NamedSparseVector,
    NamedVector,
    SparseVector,
    PointStruct,
    SearchRequest,
    ScoredPoint,
)
from langchain_core.prompts import PromptTemplate
from langchain.chains.summarize import load_summarize_chain
from huggingface_hub import snapshot_download
from exllamav2.generator import ExLlamaV2Sampler

MAP_PROMPT = """
You will be given a single passage of a book. This section will be enclosed in triple backticks (```)
Your goal is to give a summary of this section so that a reader will have a full understanding of what happened.
Your response should be at least three paragraphs and fully encompass what said in the passage.

```{text}```
FULL SUMMARY:
"""

COMBINE_PROMPT = """
You will be given a series of summaries from a book. The summaries will be enclosed in triple backticks (```)
Your goal is to give a verbose summary of what happened in the story.
The reader should be able to grasp what happened in the book.

```{text}```
VERBOSE SUMMARY:
"""

supported_splade_models[0] = {
        "model": "prithivida/Splade_PP_en_v2",
        "vocab_size": 30522,
        "description": "Implementation of SPLADE++ Model for English v2",
        "size_in_GB": 0.532,
        "sources": {
            "hf": "devve1/Splade_PP_en_v2_onnx"
        },
        "model_file": "model.onnx"
    }

def make_points(chunks: list[str], dense: list[ndarray], sparse)-> Iterable[PointStruct]:
    for idx, ((indices, values), chunk, dense_vector) in stqdm(enumerate(zip(sparse, chunks, dense)), desc='Save embeddings on disk...', backend=True):
        sparse_vector = SparseVector(indices=indices.tolist(), values=values.tolist())
        point = PointStruct(
            id=idx,
            vector={
                "text-sparse": sparse_vector,
                "text-dense": dense_vector,
            },
            payload={
                "text": chunk
            }
        )
        yield point

def search(client: QdrantClient, collection_name: str, dense: ndarray, sparse: list[SparseEmbedding]):
    search_results = client.search_batch(
        collection_name,
        [
            SearchRequest(
                vector=NamedVector(
                    name="text-dense",
                    vector=dense,
                ),
                limit=10
            ),
            SearchRequest(
                vector=NamedSparseVector(
                    name="text-sparse",
                    vector=SparseVector(
                        indices=sparse[0].indices.tolist(),
                        values=sparse[0].values.tolist(),
                    ),
                ),
                limit=10
            ),
        ],
    )

    return search_results

def rank_list(search_result: list[ScoredPoint]):
    return [(point.id, rank + 1) for rank, point in enumerate(search_result)]

def rrf(rank_lists, alpha=60, default_rank=1000):
    """
    Optimized Reciprocal Rank Fusion (RRF) using NumPy for large rank lists.

    :param rank_lists: A list of rank lists. Each rank list should be a list of (item, rank) tuples.
    :param alpha: The parameter alpha used in the RRF formula. Default is 60.
    :param default_rank: The default rank assigned to items not present in a rank list. Default is 1000.
    :return: Sorted list of items based on their RRF scores.
    """
    all_items = set(item for rank_list in rank_lists for item, _ in rank_list)
    item_to_index = {item: idx for idx, item in enumerate(all_items)}
    rank_matrix = np.full((len(all_items), len(rank_lists)), default_rank)

    for list_idx, rank_list in enumerate(rank_lists):
        for item, rank in rank_list:
            rank_matrix[item_to_index[item], list_idx] = rank

    rrf_scores = np.sum(1.0 / (alpha + rank_matrix), axis=1)
    sorted_indices = np.argsort(-rrf_scores)
    sorted_items = [(list(item_to_index.keys())[idx], rrf_scores[idx]) for idx in sorted_indices]

    return sorted_items


def main(query: str, client: QdrantClient, collection_name: str, llm, dense_model, sparse_model):
     # name = 'Kia_EV6'
    # filepath = os.path.join(os.getcwd(), name + '.pdf')

    # docs = UnstructuredFileLoader(
    #     file_path=filepath,
    #     mode='elements',
    #     strategy='hi_res',
    #     skip_infer_table_types=['png', 'pdf', 'jpg', 'xls', 'xlsx', 'heic'],
    #     hi_res_model_name='yolox',
    #     include_page_breaks=True
    # )

    # docs = docs.load()
    
    dense_query = list(dense_model.embed_query(query, 32))
    sparse_query = list(sparse_model.embed(query, 32))

    search_results = search(
        client,
        collection_name,
        dense_query,
        sparse_query
    )

    dense_rank_list, sparse_rank_list = rank_list(search_results[0]), rank_list(search_results[1])
    rrf_rank_list = rrf([dense_rank_list, sparse_rank_list])

    records_list = client.retrieve(
        collection_name,
        [item[0] for item in stqdm(rrf_rank_list, desc='Retrieve relevant chunk of texts...', backend=True)]
    )
    
    docs = [Document(record.payload['text']) for record in records_list[:3]]

    print(docs)

    map_prompt = PromptTemplate(
        template=MAP_PROMPT,
        input_variables=['text']
    )
    combine_prompt = PromptTemplate(
        template=COMBINE_PROMPT,
        input_variables=['text']
    )

    map_chain = load_summarize_chain(llm,
                                     "stuff",
                                     prompt=map_prompt
                                     )
    
    summary_list = []

    for doc in docs:
        chunk_summary = map_chain.invoke([doc])
        summary_list.append(chunk_summary['output_text'])

    summaries = Document(page_content="\n".join(summary_list))
    
    reduce_chain = load_summarize_chain(llm,
                                        "stuff",
                                        prompt=combine_prompt
                                        )
    
    output = reduce_chain.invoke([summaries])
    return output['output_text']

@st.cache_resource
def load_models_and_components(show_spinner="Loading models..."):
    settings = ExLlamaV2Sampler.Settings()
    settings.temperature = 0.75
    settings.top_k = 50
    settings.top_p = 0.8
    settings.token_repetition_penalty = 1.05
    
    model_path = snapshot_download(repo_id='Zoyd/NousResearch_Hermes-2-Theta-Llama-3-8B-6_5bpw_exl2')
    callbacks = [StreamingStdOutCallbackHandler()]

    llm = ExLlamaV2(
        model_path=model_path,
        callbacks=callbacks,
        settings=settings,
        streaming=True,
        max_new_tokens=3000
    )
        
    provider = ['CPUExecutionProvider']

    sparse_model = SparseTextEmbedding(
        'prithivida/Splade_PP_en_v2',
        cache_dir=os.getenv('HF_HOME'),
        providers=provider
    )

    dense_model = FastEmbedEmbeddingsLc(
        model_name='mixedbread-ai/mxbai-embed-large-v1',
        providers=provider,
        cache_dir=os.getenv('HF_HOME'),
        batch_size=32
    )

    return llm, dense_model, sparse_model

def chunk_documents(docs, dense_model, sparse_model, show_spinner="Parsing and chunking texts..."):
    text_splitter = SemanticChunker(
        dense_model,
        breakpoint_threshold_type='standard_deviation'
    )

    documents = [doc.page_content for doc in text_splitter.transform_documents(list(docs))]

    dense_embeddings = _dense_model.embed_documents(stqdm(documents,desc='Generate dense embeddings...', backend=True), 32)
    sparse_embeddings = list(_sparse_model.embed(stqdm(documents, desc='Generate sparse embeddings...', backend=True), 32))
    
    return documents, dense_embeddings, sparse_embeddings
    
if __name__ == '__main__':
    st.set_page_config(page_title="Video Game Assistant",
                       layout="wide"
                      )
    
    llm, dense_model, sparse_model = load_models_and_components()

    client = QdrantClient(url="http://localhost:6333")
    collection_name = 'collection_demo'
    
    if not client.collection_exists(collection_name):
        client.create_collection(
            collection_name,
            {
                "text-dense": models.VectorParams(
                    size=1024,
                    distance=models.Distance.COSINE,
                    on_disk=True,
                    quantization_config=models.BinaryQuantization(
                        binary=models.BinaryQuantizationConfig(
                            always_ram=False
                        )
                    )
                )
            },
            {
                "text-sparse": models.SparseVectorParams(
                    index=models.SparseIndexParams(
                        on_disk=True
                    )
                )
            },
            2,
            optimizers_config=models.OptimizersConfigDiff(
                memmap_threshold=10000
            ),
            hnsw_config=models.HnswConfigDiff(
                on_disk=True, 
                m=16, 
                ef_construct=100
            )
        )

        docs = WikipediaLoader(query='Action-RPG').load() 
        chunks, dense, sparse = chunk_documents(docs, dense_model, sparse_model)

        client.upload_points(
            collection_name,
            make_points(
                chunks, 
                dense, 
                sparse
            )
        )

    st.title("Video Game Assistant")
    
    if "messages" not in st.session_state:
        st.session_state.messages = []

    for message in st.session_state.messages:
         with st.chat_message(message["role"]):
            st.markdown(message["content"])

    if prompt := st.chat_input("Message Video Game Assistant"):
        st.chat_message("user").markdown(prompt)
        st.session_state.messages.append({"role": "user", "content": prompt})

        ai_response = main(prompt, client, collection_name, llm, dense_model, sparse_model)
        response = f"Echo: {ai_response}"
        with st.chat_message("assistant"):
            message_placeholder = st.empty()
            full_response = ""
            for chunk in re.split(r'(\s+)', response):
                full_response += chunk + " "
                time.sleep(0.01)
                message_placeholder.markdown(full_response + "▌")
        st.session_state.messages.append({"role": "assistant", "content": full_response})