File size: 3,849 Bytes
5a84661
 
 
 
 
 
 
 
 
 
 
 
 
129499e
5a84661
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
139a897
5a84661
139a897
5a84661
 
 
139a897
 
 
 
 
 
 
5a84661
680fe32
 
 
 
5a84661
 
 
 
 
 
139a897
 
 
5a84661
680fe32
1d11211
da1bd08
 
5a84661
 
 
139a897
 
 
 
680fe32
1d11211
da1bd08
 
5a84661
 
129499e
 
 
 
 
5a84661
 
139a897
5a84661
 
139a897
 
680fe32
129499e
5a84661
 
 
 
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
import logging
import os
import pickle

import chromadb
import logfire
from custom_retriever import CustomRetriever
from dotenv import load_dotenv
from llama_index.core import VectorStoreIndex
from llama_index.core.node_parser import SentenceSplitter
from llama_index.core.retrievers import VectorIndexRetriever
from llama_index.embeddings.openai import OpenAIEmbedding
from llama_index.vector_stores.chroma import ChromaVectorStore
from utils import init_mongo_db

load_dotenv()

logger = logging.getLogger(__name__)
logging.basicConfig(level=logging.INFO)
logging.getLogger("httpx").setLevel(logging.WARNING)
logfire.configure()


if not os.path.exists("data/chroma-db-transformers"):
    # Download the vector database from the Hugging Face Hub if it doesn't exist locally
    # https://huggingface.co/datasets/towardsai-buster/ai-tutor-vector-db/tree/main
    logfire.warn(
        f"Vector database does not exist at 'data/chroma-db-transformers', downloading from Hugging Face Hub"
    )
    from huggingface_hub import snapshot_download

    snapshot_download(
        repo_id="towardsai-buster/ai-tutor-vector-db",
        local_dir="data",
        repo_type="dataset",
    )
    logfire.info(f"Downloaded vector database to 'data/chroma-db-transformers'")


def setup_database(db_collection, dict_file_name):
    db = chromadb.PersistentClient(path=f"data/{db_collection}")
    chroma_collection = db.get_or_create_collection(db_collection)
    vector_store = ChromaVectorStore(chroma_collection=chroma_collection)

    index = VectorStoreIndex.from_vector_store(
        vector_store=vector_store,
        embed_model=OpenAIEmbedding(model="text-embedding-3-large", mode="similarity"),
        transformations=[SentenceSplitter(chunk_size=800, chunk_overlap=400)],
        show_progress=True,
        use_async=True,
    )
    vector_retriever = VectorIndexRetriever(
        index=index,
        similarity_top_k=10,
        use_async=True,
        embed_model=OpenAIEmbedding(model="text-embedding-3-large", mode="similarity"),
    )
    with open(f"data/{db_collection}/{dict_file_name}", "rb") as f:
        document_dict = pickle.load(f)

    return CustomRetriever(vector_retriever, document_dict)


# Setup retrievers
custom_retriever_transformers = setup_database(
    "chroma-db-transformers",
    "document_dict_transformers.pkl",
)
custom_retriever_peft = setup_database("chroma-db-peft", "document_dict_peft.pkl")
custom_retriever_trl = setup_database("chroma-db-trl", "document_dict_trl.pkl")
custom_retriever_llama_index = setup_database(
    "chroma-db-llama_index",
    "document_dict_llama_index.pkl",
)
custom_retriever_openai_cookbooks = setup_database(
    "chroma-db-openai_cookbooks",
    "document_dict_openai_cookbooks.pkl",
)
custom_retriever_langchain = setup_database(
    "chroma-db-langchain",
    "document_dict_langchain.pkl",
)

# Constants
CONCURRENCY_COUNT = int(os.getenv("CONCURRENCY_COUNT", 64))
MONGODB_URI = os.getenv("MONGODB_URI")

AVAILABLE_SOURCES_UI = [
    "Transformers Docs",
    "PEFT Docs",
    "TRL Docs",
    "LlamaIndex Docs",
    "LangChain Docs",
    "OpenAI Cookbooks",
    # "Towards AI Blog",
    # "RAG Course",
]

AVAILABLE_SOURCES = [
    "transformers",
    "peft",
    "trl",
    "llama_index",
    "langchain",
    "openai_cookbooks",
    # "towards_ai_blog",
    # "rag_course",
]

mongo_db = (
    init_mongo_db(uri=MONGODB_URI, db_name="towardsai-buster")
    if MONGODB_URI
    else logfire.warn("No mongodb uri found, you will not be able to save data.")
)

__all__ = [
    "custom_retriever_transformers",
    "custom_retriever_peft",
    "custom_retriever_trl",
    "custom_retriever_llama_index",
    "custom_retriever_openai_cookbooks",
    "custom_retriever_langchain",
    "mongo_db",
    "CONCURRENCY_COUNT",
    "AVAILABLE_SOURCES_UI",
    "AVAILABLE_SOURCES",
]