See also the LangChain cookbook.

Below is an example of creating a Chroma vector database using a TextSplitter with a simple text string.

from dotenv import load_dotenv
from langchain.vectorstores import Chroma
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings import OpenAIEmbeddings

load_dotenv()

# https://platform.openai.com/docs/guides/embeddings/use-cases
embeddings = OpenAIEmbeddings(model="text-embedding-ada-002")

text_splitter = RecursiveCharacterTextSplitter(
    chunk_size=10,  # default: 4000
    chunk_overlap=2  # default: 200
)

texts = text_splitter.create_documents(
    texts=['Often times your document is too long (like a book) for your LLM. You need to split it up into chunks. Text splitters help with this.'])

vectordb = Chroma.from_documents(
    documents=texts, embedding=embeddings, persist_directory="./dbtest")

You can also include PDF documents in your Chroma database. See the code below.

import os
from dotenv import load_dotenv
from langchain.vectorstores import Chroma
from langchain.document_loaders import DirectoryLoader
from langchain.document_loaders import PyPDFLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.embeddings import OpenAIEmbeddings

load_dotenv()

if not any(file.endswith('.pdf') for file in os.listdir('.')):
    exit(1)

# https://platform.openai.com/docs/guides/embeddings/use-cases
embeddings = OpenAIEmbeddings(model="text-embedding-ada-002")

text_splitter = RecursiveCharacterTextSplitter(
    chunk_size=2000,  # default: 4000
    chunk_overlap=100  # default: 200
)

loader = DirectoryLoader(".", glob='./*.pdf', loader_cls=PyPDFLoader)
documents = loader.load()

texts = text_splitter.split_documents(documents)

vectordb = Chroma.from_documents(
    documents=texts, embedding=embeddings, persist_directory="./dbtest")