docs: showcase RAG with LlamaIndex and LangChain (#71)
Signed-off-by: Panos Vagenas <35837085+vagenas@users.noreply.github.com>
This commit is contained in:
Panos Vagenas
GitHub
parent
79932b7d69
commit
53569a1023
@ -2,39 +2,43 @@ fail_fast: true
|
||||
repos:
|
||||
- repo: local
|
||||
hooks:
|
||||
- id: system
|
||||
- id: black
|
||||
name: Black
|
||||
entry: poetry run black docling examples tests
|
||||
pass_filenames: false
|
||||
language: system
|
||||
files: '\.py$'
|
||||
- repo: local
|
||||
hooks:
|
||||
- id: system
|
||||
- id: isort
|
||||
name: isort
|
||||
entry: poetry run isort docling examples tests
|
||||
pass_filenames: false
|
||||
language: system
|
||||
files: '\.py$'
|
||||
# - repo: local
|
||||
# hooks:
|
||||
# - id: system
|
||||
# - id: flake8
|
||||
# name: flake8
|
||||
# entry: poetry run flake8 docling
|
||||
# pass_filenames: false
|
||||
# language: system
|
||||
# files: '\.py$'
|
||||
# - repo: local
|
||||
# hooks:
|
||||
# - id: system
|
||||
# - id: mypy
|
||||
# name: MyPy
|
||||
# entry: poetry run mypy docling
|
||||
# pass_filenames: false
|
||||
# language: system
|
||||
# files: '\.py$'
|
||||
- repo: local
|
||||
hooks:
|
||||
- id: system
|
||||
- id: nbqa_black
|
||||
name: nbQA Black
|
||||
entry: poetry run nbqa black examples
|
||||
pass_filenames: false
|
||||
language: system
|
||||
files: '\.ipynb$'
|
||||
- id: nbqa_isort
|
||||
name: nbQA isort
|
||||
entry: poetry run nbqa isort examples
|
||||
pass_filenames: false
|
||||
language: system
|
||||
files: '\.ipynb$'
|
||||
- id: poetry
|
||||
name: Poetry check
|
||||
entry: poetry check --lock
|
||||
pass_filenames: false
|
||||
|
||||
@ -23,8 +23,7 @@ Docling bundles PDF document conversion to JSON and Markdown in an easy, self-co
|
||||
* 📑 Understands detailed page layout, reading order and recovers table structures
|
||||
* 📝 Extracts metadata from the document, such as title, authors, references and language
|
||||
* 🔍 Optionally applies OCR (use with scanned PDFs)
|
||||
|
||||
For RAG, check out [Quackling](https://github.com/DS4SD/quackling) to get the most out of your docs, be it using LlamaIndex, LangChain or your pipeline.
|
||||
* 🤖 Integrates easily with LLM app / RAG frameworks like 🦙 LlamaIndex and 🦜🔗 LangChain
|
||||
|
||||
## Installation
|
||||
|
||||
@ -143,6 +142,10 @@ results = doc_converter.convert(conv_input)
|
||||
|
||||
You can limit the CPU threads used by Docling by setting the environment variable `OMP_NUM_THREADS` accordingly. The default setting is using 4 CPU threads.
|
||||
|
||||
### RAG
|
||||
Check out the following examples showcasing RAG using Docling with standard LLM application frameworks:
|
||||
- [Basic RAG pipeline with 🦙 LlamaIndex](https://github.com/DS4SD/docling/tree/main/examples/rag_llamaindex.ipynb)
|
||||
- [Basic RAG pipeline with 🦜🔗 LangChain](https://github.com/DS4SD/docling/tree/main/examples/rag_langchain.ipynb)
|
||||
|
||||
## Technical report
|
||||
|
||||
|
||||
362
examples/rag_langchain.ipynb
Normal file
362
examples/rag_langchain.ipynb
Normal file
@ -0,0 +1,362 @@
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 1,
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stdout",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
"Note: you may need to restart the kernel to use updated packages.\n"
|
||||
]
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"# requirements for this example:\n",
|
||||
"%pip install -qq docling docling-core python-dotenv langchain-text-splitters langchain-huggingface langchain-milvus"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 2,
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"True"
|
||||
]
|
||||
},
|
||||
"execution_count": 2,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"import os\n",
|
||||
"\n",
|
||||
"from dotenv import load_dotenv\n",
|
||||
"\n",
|
||||
"load_dotenv()"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 3,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"import warnings\n",
|
||||
"\n",
|
||||
"warnings.filterwarnings(action=\"ignore\", category=UserWarning, module=\"pydantic|torch\")\n",
|
||||
"warnings.filterwarnings(action=\"ignore\", category=FutureWarning, module=\"easyocr\")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"## Setup"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"### Loader and splitter"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"Below we set up:\n",
|
||||
"- a `Loader` which will be used to create LangChain documents, and\n",
|
||||
"- a splitter, which will be used to split these documents"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 4,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from enum import Enum\n",
|
||||
"from typing import Iterator\n",
|
||||
"\n",
|
||||
"from langchain_core.document_loaders import BaseLoader\n",
|
||||
"from langchain_core.documents import Document as LCDocument\n",
|
||||
"from pydantic import BaseModel\n",
|
||||
"\n",
|
||||
"from docling.document_converter import DocumentConverter\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"class DocumentMetadata(BaseModel):\n",
|
||||
" dl_doc_hash: str\n",
|
||||
" # source: str\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"class DoclingPDFLoader(BaseLoader):\n",
|
||||
" class ParseType(str, Enum):\n",
|
||||
" MARKDOWN = \"markdown\"\n",
|
||||
" # JSON = \"json\"\n",
|
||||
"\n",
|
||||
" def __init__(self, file_path: str | list[str], parse_type: ParseType) -> None:\n",
|
||||
" self._file_paths = file_path if isinstance(file_path, list) else [file_path]\n",
|
||||
" self._parse_type = parse_type\n",
|
||||
" self._converter = DocumentConverter()\n",
|
||||
"\n",
|
||||
" def lazy_load(self) -> Iterator[LCDocument]:\n",
|
||||
" for source in self._file_paths:\n",
|
||||
" dl_doc = self._converter.convert_single(source).output\n",
|
||||
" match self._parse_type:\n",
|
||||
" case self.ParseType.MARKDOWN:\n",
|
||||
" text = dl_doc.export_to_markdown()\n",
|
||||
" # case self.ParseType.JSON:\n",
|
||||
" # text = dl_doc.model_dump_json()\n",
|
||||
" case _:\n",
|
||||
" raise RuntimeError(\n",
|
||||
" f\"Unexpected parse type encountered: {self._parse_type}\"\n",
|
||||
" )\n",
|
||||
" lc_doc = LCDocument(\n",
|
||||
" page_content=text,\n",
|
||||
" metadata=DocumentMetadata(\n",
|
||||
" dl_doc_hash=dl_doc.file_info.document_hash,\n",
|
||||
" ).model_dump(),\n",
|
||||
" )\n",
|
||||
" yield lc_doc"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 5,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"FILE_PATH = \"https://arxiv.org/pdf/2206.01062\" # DocLayNet paper"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 6,
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"application/vnd.jupyter.widget-view+json": {
|
||||
"model_id": "1b38d07d5fed4618a44ecf261e1e5c44",
|
||||
"version_major": 2,
|
||||
"version_minor": 0
|
||||
},
|
||||
"text/plain": [
|
||||
"Fetching 7 files: 0%| | 0/7 [00:00<?, ?it/s]"
|
||||
]
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"from langchain_text_splitters import RecursiveCharacterTextSplitter\n",
|
||||
"\n",
|
||||
"loader = DoclingPDFLoader(\n",
|
||||
" file_path=FILE_PATH,\n",
|
||||
" parse_type=DoclingPDFLoader.ParseType.MARKDOWN,\n",
|
||||
")\n",
|
||||
"text_splitter = RecursiveCharacterTextSplitter(\n",
|
||||
" chunk_size=1000,\n",
|
||||
" chunk_overlap=200,\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"We now used the above-defined objects to get the document splits:"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 7,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"docs = loader.load()\n",
|
||||
"splits = text_splitter.split_documents(docs)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"### Embeddings"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 8,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from langchain_huggingface.embeddings import HuggingFaceEmbeddings\n",
|
||||
"\n",
|
||||
"HF_EMBED_MODEL_ID = \"BAAI/bge-small-en-v1.5\"\n",
|
||||
"embeddings = HuggingFaceEmbeddings(model_name=HF_EMBED_MODEL_ID)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"### Vector store"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 9,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from tempfile import TemporaryDirectory\n",
|
||||
"\n",
|
||||
"from langchain_milvus import Milvus\n",
|
||||
"\n",
|
||||
"MILVUS_URI = os.environ.get(\n",
|
||||
" \"MILVUS_URL\", f\"{(tmp_dir := TemporaryDirectory()).name}/milvus_demo.db\"\n",
|
||||
")\n",
|
||||
"\n",
|
||||
"vectorstore = Milvus.from_documents(\n",
|
||||
" splits,\n",
|
||||
" embeddings,\n",
|
||||
" connection_args={\"uri\": MILVUS_URI},\n",
|
||||
" drop_old=True,\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"### LLM"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 10,
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stdout",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
"The token has not been saved to the git credentials helper. Pass `add_to_git_credential=True` in this function directly or `--add-to-git-credential` if using via `huggingface-cli` if you want to set the git credential as well.\n",
|
||||
"Token is valid (permission: write).\n",
|
||||
"Your token has been saved to /Users/pva/.cache/huggingface/token\n",
|
||||
"Login successful\n"
|
||||
]
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"from langchain_huggingface import HuggingFaceEndpoint\n",
|
||||
"\n",
|
||||
"HF_API_KEY = os.environ.get(\"HF_API_KEY\")\n",
|
||||
"HF_LLM_MODEL_ID = \"mistralai/Mistral-7B-Instruct-v0.3\"\n",
|
||||
"\n",
|
||||
"llm = HuggingFaceEndpoint(\n",
|
||||
" repo_id=HF_LLM_MODEL_ID,\n",
|
||||
" huggingfacehub_api_token=HF_API_KEY,\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"## RAG"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 11,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from typing import Iterable\n",
|
||||
"\n",
|
||||
"from langchain_core.documents import Document as LCDocument\n",
|
||||
"from langchain_core.output_parsers import StrOutputParser\n",
|
||||
"from langchain_core.prompts import PromptTemplate\n",
|
||||
"from langchain_core.runnables import RunnablePassthrough\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"def format_docs(docs: Iterable[LCDocument]):\n",
|
||||
" return \"\\n\\n\".join(doc.page_content for doc in docs)\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"retriever = vectorstore.as_retriever()\n",
|
||||
"\n",
|
||||
"prompt = PromptTemplate.from_template(\n",
|
||||
" \"Context information is below.\\n---------------------\\n{context}\\n---------------------\\nGiven the context information and not prior knowledge, answer the query.\\nQuery: {question}\\nAnswer:\\n\"\n",
|
||||
")\n",
|
||||
"\n",
|
||||
"rag_chain = (\n",
|
||||
" {\"context\": retriever | format_docs, \"question\": RunnablePassthrough()}\n",
|
||||
" | prompt\n",
|
||||
" | llm\n",
|
||||
" | StrOutputParser()\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 12,
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"'The human annotation of DocLayNet was performed on 80863 pages.\\n\\nExplanation:\\nThe information is found in the paragraph \"DocLayNet contains 80863 PDF pages\" in the context.'"
|
||||
]
|
||||
},
|
||||
"execution_count": 12,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"rag_chain.invoke(\"How many pages were human annotated for DocLayNet?\")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
"kernelspec": {
|
||||
"display_name": ".venv",
|
||||
"language": "python",
|
||||
"name": "python3"
|
||||
},
|
||||
"language_info": {
|
||||
"codemirror_mode": {
|
||||
"name": "ipython",
|
||||
"version": 3
|
||||
},
|
||||
"file_extension": ".py",
|
||||
"mimetype": "text/x-python",
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.12.4"
|
||||
}
|
||||
},
|
||||
"nbformat": 4,
|
||||
"nbformat_minor": 2
|
||||
}
|
||||
479
examples/rag_llamaindex.ipynb
Normal file
479
examples/rag_llamaindex.ipynb
Normal file
@ -0,0 +1,479 @@
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"# Quackling — Basic Pipeline"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 1,
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stdout",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
"Note: you may need to restart the kernel to use updated packages.\n"
|
||||
]
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"# requirements for this example:\n",
|
||||
"%pip install -qq docling docling-core python-dotenv llama-index-embeddings-huggingface llama-index-llms-huggingface-api llama-index-vector-stores-milvus"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 2,
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"True"
|
||||
]
|
||||
},
|
||||
"execution_count": 2,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"import os\n",
|
||||
"from tempfile import TemporaryDirectory\n",
|
||||
"\n",
|
||||
"from dotenv import load_dotenv\n",
|
||||
"from pydantic import TypeAdapter\n",
|
||||
"from rich.pretty import pprint\n",
|
||||
"\n",
|
||||
"load_dotenv()"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 3,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"import warnings\n",
|
||||
"\n",
|
||||
"warnings.filterwarnings(action=\"ignore\", category=UserWarning, module=\"pydantic|torch\")\n",
|
||||
"warnings.filterwarnings(action=\"ignore\", category=FutureWarning, module=\"easyocr\")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"## Setup"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"### Reader and node parser"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"Below we set up:\n",
|
||||
"- a `Reader` which will be used to create LlamaIndex documents, and\n",
|
||||
"- a `NodeParser`, which will be used to create LlamaIndex nodes out of the documents"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 4,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from enum import Enum\n",
|
||||
"from typing import Iterable\n",
|
||||
"\n",
|
||||
"from llama_index.core.readers.base import BasePydanticReader\n",
|
||||
"from llama_index.core.schema import Document as LIDocument\n",
|
||||
"from pydantic import BaseModel\n",
|
||||
"\n",
|
||||
"from docling.document_converter import DocumentConverter\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"class DocumentMetadata(BaseModel):\n",
|
||||
" dl_doc_hash: str\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"class DoclingPDFReader(BasePydanticReader):\n",
|
||||
" class ParseType(str, Enum):\n",
|
||||
" MARKDOWN = \"markdown\"\n",
|
||||
" # JSON = \"json\"\n",
|
||||
"\n",
|
||||
" parse_type: ParseType = ParseType.MARKDOWN\n",
|
||||
"\n",
|
||||
" def lazy_load_data(self, file_path: str | list[str]) -> Iterable[LIDocument]:\n",
|
||||
" file_paths = file_path if isinstance(file_path, list) else [file_path]\n",
|
||||
" converter = DocumentConverter()\n",
|
||||
" for source in file_paths:\n",
|
||||
" dl_doc = converter.convert_single(source).output\n",
|
||||
" match self.parse_type:\n",
|
||||
" case self.ParseType.MARKDOWN:\n",
|
||||
" text = dl_doc.export_to_markdown()\n",
|
||||
" # case self.ParseType.JSON:\n",
|
||||
" # text = dl_doc.model_dump_json()\n",
|
||||
" case _:\n",
|
||||
" raise RuntimeError(\n",
|
||||
" f\"Unexpected parse type encountered: {self.parse_type}\"\n",
|
||||
" )\n",
|
||||
" excl_metadata_keys = [\"dl_doc_hash\"]\n",
|
||||
" li_doc = LIDocument(\n",
|
||||
" doc_id=dl_doc.file_info.document_hash,\n",
|
||||
" text=text,\n",
|
||||
" excluded_embed_metadata_keys=excl_metadata_keys,\n",
|
||||
" excluded_llm_metadata_keys=excl_metadata_keys,\n",
|
||||
" )\n",
|
||||
" li_doc.metadata = DocumentMetadata(\n",
|
||||
" dl_doc_hash=dl_doc.file_info.document_hash,\n",
|
||||
" ).model_dump()\n",
|
||||
" yield li_doc"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 5,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from llama_index.core.node_parser import MarkdownNodeParser\n",
|
||||
"\n",
|
||||
"reader = DoclingPDFReader(parse_type=DoclingPDFReader.ParseType.MARKDOWN)\n",
|
||||
"node_parser = MarkdownNodeParser()\n",
|
||||
"transformations = [node_parser]"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"One can include add more transformations, e.g. further chunking based on text size / overlap, as shown below:"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 6,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"# from llama_index.core.node_parser import TokenTextSplitter\n",
|
||||
"\n",
|
||||
"# splitter = TokenTextSplitter(\n",
|
||||
"# chunk_size=1024,\n",
|
||||
"# chunk_overlap=20,\n",
|
||||
"# )\n",
|
||||
"# transformations.append(splitter)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"### Embed model"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 7,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from llama_index.embeddings.huggingface import HuggingFaceEmbedding\n",
|
||||
"\n",
|
||||
"embed_model = HuggingFaceEmbedding(model_name=\"BAAI/bge-small-en-v1.5\")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"### Vector store"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 8,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"INGEST = True # whether to ingest from scratch or reuse an existing vector store"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 9,
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"name": "stderr",
|
||||
"output_type": "stream",
|
||||
"text": [
|
||||
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
|
||||
"To disable this warning, you can either:\n",
|
||||
"\t- Avoid using `tokenizers` before the fork if possible\n",
|
||||
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n"
|
||||
]
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"from llama_index.vector_stores.milvus import MilvusVectorStore\n",
|
||||
"\n",
|
||||
"MILVUS_URL = os.environ.get(\n",
|
||||
" \"MILVUS_URL\", f\"{(tmp_dir := TemporaryDirectory()).name}/milvus_demo.db\"\n",
|
||||
")\n",
|
||||
"MILVUS_COLL_NAME = os.environ.get(\"MILVUS_COLL_NAME\", \"basic_llamaindex_pipeline\")\n",
|
||||
"MILVUS_KWARGS = TypeAdapter(dict).validate_json(os.environ.get(\"MILVUS_KWARGS\", \"{}\"))\n",
|
||||
"vector_store = MilvusVectorStore(\n",
|
||||
" uri=MILVUS_URL,\n",
|
||||
" collection_name=MILVUS_COLL_NAME,\n",
|
||||
" dim=len(embed_model.get_text_embedding(\"hi\")),\n",
|
||||
" overwrite=INGEST,\n",
|
||||
" **MILVUS_KWARGS,\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 10,
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"application/vnd.jupyter.widget-view+json": {
|
||||
"model_id": "536daee038de4d52a793445c6d853c72",
|
||||
"version_major": 2,
|
||||
"version_minor": 0
|
||||
},
|
||||
"text/plain": [
|
||||
"Fetching 7 files: 0%| | 0/7 [00:00<?, ?it/s]"
|
||||
]
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
},
|
||||
{
|
||||
"data": {
|
||||
"text/html": [
|
||||
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">[</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ </span><span style=\"color: #800080; text-decoration-color: #800080; font-weight: bold\">Document</span><span style=\"font-weight: bold\">(</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">id_</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'5dfbd8c115a15fd3396b68409124cfee29fc8efac7b5c84663'</span>+<span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">14</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">embedding</span>=<span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">metadata</span>=<span style=\"font-weight: bold\">{</span><span style=\"color: #008000; text-decoration-color: #008000\">'dl_doc_hash'</span>: <span style=\"color: #008000; text-decoration-color: #008000\">'5dfbd8c115a15fd3396b68409124cfee29fc8efac7b5c84663'</span>+<span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">14</span><span style=\"font-weight: bold\">}</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">excluded_embed_metadata_keys</span>=<span style=\"font-weight: bold\">[</span><span style=\"color: #008000; text-decoration-color: #008000\">'dl_doc_hash'</span><span style=\"font-weight: bold\">]</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">excluded_llm_metadata_keys</span>=<span style=\"font-weight: bold\">[</span><span style=\"color: #008000; text-decoration-color: #008000\">'dl_doc_hash'</span><span style=\"font-weight: bold\">]</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">relationships</span>=<span style=\"font-weight: bold\">{}</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">text</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'## DocLayNet: A Large Human-Annotated Dataset for '</span>+<span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">50593</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">mimetype</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'text/plain'</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">start_char_idx</span>=<span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">end_char_idx</span>=<span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">text_template</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'{metadata_str}\\n\\n{content}'</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">metadata_template</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'{key}: {value}'</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">metadata_seperator</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'\\n'</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ </span><span style=\"font-weight: bold\">)</span>\n",
|
||||
"<span style=\"font-weight: bold\">]</span>\n",
|
||||
"</pre>\n"
|
||||
],
|
||||
"text/plain": [
|
||||
"\u001b[1m[\u001b[0m\n",
|
||||
"\u001b[2;32m│ \u001b[0m\u001b[1;35mDocument\u001b[0m\u001b[1m(\u001b[0m\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mid_\u001b[0m=\u001b[32m'5dfbd8c115a15fd3396b68409124cfee29fc8efac7b5c84663'\u001b[0m+\u001b[1;36m14\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33membedding\u001b[0m=\u001b[3;35mNone\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mmetadata\u001b[0m=\u001b[1m{\u001b[0m\u001b[32m'dl_doc_hash'\u001b[0m: \u001b[32m'5dfbd8c115a15fd3396b68409124cfee29fc8efac7b5c84663'\u001b[0m+\u001b[1;36m14\u001b[0m\u001b[1m}\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mexcluded_embed_metadata_keys\u001b[0m=\u001b[1m[\u001b[0m\u001b[32m'dl_doc_hash'\u001b[0m\u001b[1m]\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mexcluded_llm_metadata_keys\u001b[0m=\u001b[1m[\u001b[0m\u001b[32m'dl_doc_hash'\u001b[0m\u001b[1m]\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mrelationships\u001b[0m=\u001b[1m{\u001b[0m\u001b[1m}\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mtext\u001b[0m=\u001b[32m'## DocLayNet: A Large Human-Annotated Dataset for '\u001b[0m+\u001b[1;36m50593\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mmimetype\u001b[0m=\u001b[32m'text/plain'\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mstart_char_idx\u001b[0m=\u001b[3;35mNone\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mend_char_idx\u001b[0m=\u001b[3;35mNone\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mtext_template\u001b[0m=\u001b[32m'\u001b[0m\u001b[32m{\u001b[0m\u001b[32mmetadata_str\u001b[0m\u001b[32m}\u001b[0m\u001b[32m\\n\\n\u001b[0m\u001b[32m{\u001b[0m\u001b[32mcontent\u001b[0m\u001b[32m}\u001b[0m\u001b[32m'\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mmetadata_template\u001b[0m=\u001b[32m'\u001b[0m\u001b[32m{\u001b[0m\u001b[32mkey\u001b[0m\u001b[32m}\u001b[0m\u001b[32m: \u001b[0m\u001b[32m{\u001b[0m\u001b[32mvalue\u001b[0m\u001b[32m}\u001b[0m\u001b[32m'\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33mmetadata_seperator\u001b[0m=\u001b[32m'\\n'\u001b[0m\n",
|
||||
"\u001b[2;32m│ \u001b[0m\u001b[1m)\u001b[0m\n",
|
||||
"\u001b[1m]\u001b[0m\n"
|
||||
]
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"from llama_index.core import StorageContext, VectorStoreIndex\n",
|
||||
"\n",
|
||||
"if INGEST:\n",
|
||||
" # in this case we ingest the data into the vector store\n",
|
||||
" docs = reader.load_data(\n",
|
||||
" file_path=\"https://arxiv.org/pdf/2206.01062\", # DocLayNet paper\n",
|
||||
" )\n",
|
||||
" pprint(docs, max_length=1, max_string=50, max_depth=4)\n",
|
||||
" storage_context = StorageContext.from_defaults(vector_store=vector_store)\n",
|
||||
" index = VectorStoreIndex.from_documents(\n",
|
||||
" documents=docs,\n",
|
||||
" embed_model=embed_model,\n",
|
||||
" storage_context=storage_context,\n",
|
||||
" transformations=transformations,\n",
|
||||
" )\n",
|
||||
"else:\n",
|
||||
" # in this case we just load the vector store index\n",
|
||||
" index = VectorStoreIndex.from_vector_store(\n",
|
||||
" vector_store=vector_store,\n",
|
||||
" embed_model=embed_model,\n",
|
||||
" )"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"### LLM"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 11,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from llama_index.llms.huggingface_api import HuggingFaceInferenceAPI\n",
|
||||
"\n",
|
||||
"HF_API_KEY = os.environ.get(\"HF_API_KEY\")\n",
|
||||
"\n",
|
||||
"llm = HuggingFaceInferenceAPI(\n",
|
||||
" token=HF_API_KEY,\n",
|
||||
" model_name=\"mistralai/Mistral-7B-Instruct-v0.3\",\n",
|
||||
")"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"## RAG"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 12,
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/html": [
|
||||
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"color: #800080; text-decoration-color: #800080; font-weight: bold\">Response</span><span style=\"font-weight: bold\">(</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ </span><span style=\"color: #808000; text-decoration-color: #808000\">response</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'80863 pages were human annotated.'</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ </span><span style=\"color: #808000; text-decoration-color: #808000\">source_nodes</span>=<span style=\"font-weight: bold\">[</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #800080; text-decoration-color: #800080; font-weight: bold\">NodeWithScore</span><span style=\"font-weight: bold\">(</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">node</span>=<span style=\"color: #800080; text-decoration-color: #800080; font-weight: bold\">TextNode</span><span style=\"font-weight: bold\">(</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">id_</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'8874a117-d181-4f4f-a30b-0b5604370d77'</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">embedding</span>=<span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">metadata</span>=<span style=\"font-weight: bold\">{</span><span style=\"color: #808000; text-decoration-color: #808000\">...</span><span style=\"font-weight: bold\">}</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">excluded_embed_metadata_keys</span>=<span style=\"font-weight: bold\">[</span><span style=\"color: #808000; text-decoration-color: #808000\">...</span><span style=\"font-weight: bold\">]</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">excluded_llm_metadata_keys</span>=<span style=\"font-weight: bold\">[</span><span style=\"color: #808000; text-decoration-color: #808000\">...</span><span style=\"font-weight: bold\">]</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">relationships</span>=<span style=\"font-weight: bold\">{</span><span style=\"color: #808000; text-decoration-color: #808000\">...</span><span style=\"font-weight: bold\">}</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">text</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'3 THE DOCLAYNET DATASET\\n\\nDocLayNet contains 80863 PDF pages. Among these, 7059 carry two instances of human annotations, and 1591 carry three. This amounts to 91104 total annotation instances. The annotations provide layout information in the shape o'</span>+<span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">5775</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">mimetype</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'text/plain'</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">start_char_idx</span>=<span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">9089</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">end_char_idx</span>=<span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">15114</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">text_template</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'{metadata_str}\\n\\n{content}'</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">metadata_template</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'{key}: {value}'</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">metadata_seperator</span>=<span style=\"color: #008000; text-decoration-color: #008000\">'\\n'</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ </span><span style=\"font-weight: bold\">)</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">score</span>=<span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">0.7367570400238037</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"font-weight: bold\">)</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">...</span> +<span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">1</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ </span><span style=\"font-weight: bold\">]</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ </span><span style=\"color: #808000; text-decoration-color: #808000\">metadata</span>=<span style=\"font-weight: bold\">{</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #008000; text-decoration-color: #008000\">'8874a117-d181-4f4f-a30b-0b5604370d77'</span>: <span style=\"font-weight: bold\">{</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ </span><span style=\"color: #008000; text-decoration-color: #008000\">'dl_doc_hash'</span>: <span style=\"color: #008000; text-decoration-color: #008000\">'5dfbd8c115a15fd3396b68409124cfee29fc8efac7b5c846634ff924e635e0dc'</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">...</span> +<span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">1</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"font-weight: bold\">}</span>,\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ │ </span><span style=\"color: #808000; text-decoration-color: #808000\">...</span> +<span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">1</span>\n",
|
||||
"<span style=\"color: #7fbf7f; text-decoration-color: #7fbf7f\">│ </span><span style=\"font-weight: bold\">}</span>\n",
|
||||
"<span style=\"font-weight: bold\">)</span>\n",
|
||||
"</pre>\n"
|
||||
],
|
||||
"text/plain": [
|
||||
"\u001b[1;35mResponse\u001b[0m\u001b[1m(\u001b[0m\n",
|
||||
"\u001b[2;32m│ \u001b[0m\u001b[33mresponse\u001b[0m=\u001b[32m'80863 pages were human annotated.'\u001b[0m,\n",
|
||||
"\u001b[2;32m│ \u001b[0m\u001b[33msource_nodes\u001b[0m=\u001b[1m[\u001b[0m\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[1;35mNodeWithScore\u001b[0m\u001b[1m(\u001b[0m\n",
|
||||
"\u001b[2;32m│ │ │ \u001b[0m\u001b[33mnode\u001b[0m=\u001b[1;35mTextNode\u001b[0m\u001b[1m(\u001b[0m\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mid_\u001b[0m=\u001b[32m'8874a117-d181-4f4f-a30b-0b5604370d77'\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33membedding\u001b[0m=\u001b[3;35mNone\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mmetadata\u001b[0m=\u001b[1m{\u001b[0m\u001b[33m...\u001b[0m\u001b[1m}\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mexcluded_embed_metadata_keys\u001b[0m=\u001b[1m[\u001b[0m\u001b[33m...\u001b[0m\u001b[1m]\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mexcluded_llm_metadata_keys\u001b[0m=\u001b[1m[\u001b[0m\u001b[33m...\u001b[0m\u001b[1m]\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mrelationships\u001b[0m=\u001b[1m{\u001b[0m\u001b[33m...\u001b[0m\u001b[1m}\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mtext\u001b[0m=\u001b[32m'3 THE DOCLAYNET DATASET\\n\\nDocLayNet contains 80863 PDF pages. Among these, 7059 carry two instances of human annotations, and 1591 carry three. This amounts to 91104 total annotation instances. The annotations provide layout information in the shape o'\u001b[0m+\u001b[1;36m5775\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mmimetype\u001b[0m=\u001b[32m'text/plain'\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mstart_char_idx\u001b[0m=\u001b[1;36m9089\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mend_char_idx\u001b[0m=\u001b[1;36m15114\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mtext_template\u001b[0m=\u001b[32m'\u001b[0m\u001b[32m{\u001b[0m\u001b[32mmetadata_str\u001b[0m\u001b[32m}\u001b[0m\u001b[32m\\n\\n\u001b[0m\u001b[32m{\u001b[0m\u001b[32mcontent\u001b[0m\u001b[32m}\u001b[0m\u001b[32m'\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mmetadata_template\u001b[0m=\u001b[32m'\u001b[0m\u001b[32m{\u001b[0m\u001b[32mkey\u001b[0m\u001b[32m}\u001b[0m\u001b[32m: \u001b[0m\u001b[32m{\u001b[0m\u001b[32mvalue\u001b[0m\u001b[32m}\u001b[0m\u001b[32m'\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ │ \u001b[0m\u001b[33mmetadata_seperator\u001b[0m=\u001b[32m'\\n'\u001b[0m\n",
|
||||
"\u001b[2;32m│ │ │ \u001b[0m\u001b[1m)\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ \u001b[0m\u001b[33mscore\u001b[0m=\u001b[1;36m0\u001b[0m\u001b[1;36m.7367570400238037\u001b[0m\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[1m)\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33m...\u001b[0m +\u001b[1;36m1\u001b[0m\n",
|
||||
"\u001b[2;32m│ \u001b[0m\u001b[1m]\u001b[0m,\n",
|
||||
"\u001b[2;32m│ \u001b[0m\u001b[33mmetadata\u001b[0m=\u001b[1m{\u001b[0m\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[32m'8874a117-d181-4f4f-a30b-0b5604370d77'\u001b[0m: \u001b[1m{\u001b[0m\n",
|
||||
"\u001b[2;32m│ │ │ \u001b[0m\u001b[32m'dl_doc_hash'\u001b[0m: \u001b[32m'5dfbd8c115a15fd3396b68409124cfee29fc8efac7b5c846634ff924e635e0dc'\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ │ \u001b[0m\u001b[33m...\u001b[0m +\u001b[1;36m1\u001b[0m\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[1m}\u001b[0m,\n",
|
||||
"\u001b[2;32m│ │ \u001b[0m\u001b[33m...\u001b[0m +\u001b[1;36m1\u001b[0m\n",
|
||||
"\u001b[2;32m│ \u001b[0m\u001b[1m}\u001b[0m\n",
|
||||
"\u001b[1m)\u001b[0m\n"
|
||||
]
|
||||
},
|
||||
"metadata": {},
|
||||
"output_type": "display_data"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"query_engine = index.as_query_engine(llm=llm)\n",
|
||||
"query_res = query_engine.query(\"How many pages were human annotated?\")\n",
|
||||
"pprint(query_res, max_length=1, max_string=250, max_depth=4)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": null,
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": []
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
"kernelspec": {
|
||||
"display_name": ".venv",
|
||||
"language": "python",
|
||||
"name": "python3"
|
||||
},
|
||||
"language_info": {
|
||||
"codemirror_mode": {
|
||||
"name": "ipython",
|
||||
"version": 3
|
||||
},
|
||||
"file_extension": ".py",
|
||||
"mimetype": "text/x-python",
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.12.4"
|
||||
}
|
||||
},
|
||||
"nbformat": 4,
|
||||
"nbformat_minor": 2
|
||||
}
|
||||
1823
poetry.lock
generated
1823
poetry.lock
generated
File diff suppressed because it is too large
Load Diff
@ -23,7 +23,7 @@ packages = [{include = "docling"}]
|
||||
[tool.poetry.dependencies]
|
||||
python = "^3.10"
|
||||
pydantic = "^2.0.0"
|
||||
docling-core = "^1.2.0"
|
||||
docling-core = "^1.3.0"
|
||||
docling-ibm-models = "^1.1.7"
|
||||
deepsearch-glm = "^0.21.0"
|
||||
filetype = "^1.2.0"
|
||||
@ -38,6 +38,29 @@ rtree = "^1.3.0"
|
||||
scipy = "^1.14.1"
|
||||
pyarrow = "^16.1.0"
|
||||
|
||||
#########
|
||||
# extras:
|
||||
#########
|
||||
python-dotenv = { version = "^1.0.1", optional = true }
|
||||
llama-index-embeddings-huggingface = { version = "^0.3.1", optional = true }
|
||||
llama-index-llms-huggingface-api = { version = "^0.2.0", optional = true }
|
||||
llama-index-vector-stores-milvus = { version = "^0.2.1", optional = true }
|
||||
langchain-huggingface = { version = "^0.0.3", optional = true}
|
||||
langchain-milvus = { version = "^0.1.4", optional = true }
|
||||
langchain-text-splitters = { version = "^0.2.4", optional = true }
|
||||
|
||||
##############
|
||||
# constraints:
|
||||
##############
|
||||
torch = [
|
||||
{version = "^2.2.2", optional = true, markers = "sys_platform != 'darwin' or platform_machine != 'x86_64'"},
|
||||
{version = "~2.2.2", optional = true, markers = "sys_platform == 'darwin' and platform_machine == 'x86_64'"}
|
||||
]
|
||||
torchvision = [
|
||||
{version = "^0", optional = true, markers = "sys_platform != 'darwin' or platform_machine != 'x86_64'"},
|
||||
{version = "~0.17.2", optional = true, markers = "sys_platform == 'darwin' and platform_machine == 'x86_64'"}
|
||||
]
|
||||
|
||||
[tool.poetry.group.dev.dependencies]
|
||||
black = {extras = ["jupyter"], version = "^24.4.2"}
|
||||
pytest = "^7.2.2"
|
||||
@ -51,11 +74,26 @@ pytest-xdist = "^3.3.1"
|
||||
types-requests = "^2.31.0.2"
|
||||
flake8-pyproject = "^1.2.3"
|
||||
pylint = "^2.17.5"
|
||||
|
||||
ipykernel = "^6.29.5"
|
||||
ipywidgets = "^8.1.5"
|
||||
nbqa = "^1.9.0"
|
||||
|
||||
[tool.poetry.group.examples.dependencies]
|
||||
datasets = "^2.21.0"
|
||||
|
||||
[tool.poetry.extras]
|
||||
examples = [
|
||||
"python-dotenv",
|
||||
# LlamaIndex examples:
|
||||
"llama-index-embeddings-huggingface",
|
||||
"llama-index-llms-huggingface-api",
|
||||
"llama-index-vector-stores-milvus",
|
||||
# LangChain examples:
|
||||
"langchain-huggingface",
|
||||
"langchain-milvus",
|
||||
"langchain-text-splitters",
|
||||
]
|
||||
|
||||
[build-system]
|
||||
requires = ["poetry-core"]
|
||||
build-backend = "poetry.core.masonry.api"
|
||||
|
||||
Loading…
Reference in New Issue
Block a user