mirrors/docling
1
0
Fork 0
mirror of https://github.com/DS4SD/docling.git synced 2025-07-27 04:24:45 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge remote-tracking branch 'origin/main' into cau/input-format-abstraction

Browse Source 
Signed-off-by: Michele Dolfi <dol@zurich.ibm.com>
This commit is contained in:
Michele Dolfi 2024-10-11 11:23:04 +02:00
commit 331ab36f04
8 changed files with 522 additions and 510 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
CHANGELOG.md
View File

@ -1,3 +1,13 @@
## [v1.19.1](https://github.com/DS4SD/docling/releases/tag/v1.19.1) - 2024-10-11
### Fix
* Remove stderr from tesseract cli and introduce fuzziness in the text validation of OCR tests ([#138](https://github.com/DS4SD/docling/issues/138)) ([`dae2a3b`](https://github.com/DS4SD/docling/commit/dae2a3b66732e1e135b00cce24226c7d9f2eb2e4))
### Documentation
* Simplify LlamaIndex example using Docling extension ([#135](https://github.com/DS4SD/docling/issues/135)) ([`5f1bd9e`](https://github.com/DS4SD/docling/commit/5f1bd9e9c8a19c667d1d587a557c3c36df494762))
## [v1.19.0](https://github.com/DS4SD/docling/releases/tag/v1.19.0) - 2024-10-08
### Feature

17
README.md
View File

@ -291,15 +291,14 @@ from docling_core.transforms.chunker import HierarchicalChunker
doc = DocumentConverter().convert_single("https://arxiv.org/pdf/2206.01062").legacy_output
chunks = list(HierarchicalChunker().chunk(doc))
# > [
# > ChunkWithMetadata(
# > path='$.main-text[0]',
# > text='DocLayNet: A Large Human-Annotated Dataset [...]',
# > page=1,
# > bbox=[107.30, 672.38, 505.19, 709.08]
# > ),
# > [...]
# > ]
print(chunks[0])
# ChunkWithMetadata(
# path='#/main-text/1',
# text='DocLayNet: A Large Human-Annotated Dataset [...]',
# page=1,
# bbox=[107.30, 672.38, 505.19, 709.08],
# [...]
# )
```

4
docling/models/tesseract_ocr_cli_model.py
View File

@ -1,7 +1,7 @@
import io
import logging
import tempfile
from subprocess import PIPE, Popen
from subprocess import DEVNULL, PIPE, Popen
from typing import Iterable, Tuple
import pandas as pd
@ -82,7 +82,7 @@ class TesseractOcrCliModel(BaseOcrModel):
cmd += [ifilename, "stdout", "tsv"]
_log.info("command: {}".format(" ".join(cmd)))
proc = Popen(cmd, stdout=PIPE)
proc = Popen(cmd, stdout=PIPE, stderr=DEVNULL)
output, _ = proc.communicate()
# _log.info(output)

627
examples/rag_llamaindex.ipynb
View File

@ -1,5 +1,12 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"<a href=\"https://colab.research.google.com/github/DS4SD/docling/blob/main/examples/rag_llamaindex.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"cell_type": "markdown",
"metadata": {},
@ -7,6 +14,38 @@
"# RAG with Docling and 🦙 LlamaIndex"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Overview"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"LlamaIndex extensions `DoclingReader` and `DoclingNodeParser` presented in this notebook seamlessly integrate Docling into LlamaIndex, enabling you to:\n",
"- use PDF documents in your LLM applications with ease and speed, and\n",
"- leverage Docling's rich format for advanced, document-native grounding."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Setup"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"- 👉 For best conversion speed, use GPU acceleration whenever available; e.g. if running on Colab, use GPU-enabled runtime.\n",
"- Notebook uses HuggingFace's Inference API; for increased LLM quota, token can be provided via env var `HF_TOKEN`.\n",
"- Requirements can be installed as shown below (`--no-warn-conflicts` meant for Colab's pre-populated Python env; feel free to remove for stricter usage):"
]
},
{
"cell_type": "code",
"execution_count": 1,
@ -21,35 +60,49 @@
}
],
"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"
"%pip install -q --progress-bar off --no-warn-conflicts llama-index-core llama-index-readers-docling llama-index-node-parser-docling llama-index-embeddings-huggingface llama-index-llms-huggingface-api llama-index-readers-file python-dotenv"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"True"
]
},
"execution_count": 2,
"metadata": {},
"output_type": "execute_result"
}
],
"outputs": [],
"source": [
"import os\n",
"from tempfile import TemporaryDirectory\n",
"from pathlib import Path\n",
"from tempfile import mkdtemp\n",
"from warnings import filterwarnings\n",
"\n",
"from dotenv import load_dotenv\n",
"from pydantic import TypeAdapter\n",
"from rich.pretty import pprint\n",
"\n",
"load_dotenv()"
"\n",
"def _get_env_from_colab_or_os(key):\n",
" try:\n",
" from google.colab import userdata\n",
"\n",
" try:\n",
" return userdata.get(key)\n",
" except userdata.SecretNotFoundError:\n",
" pass\n",
" except ImportError:\n",
" pass\n",
" return os.getenv(key)\n",
"\n",
"\n",
"load_dotenv()\n",
"\n",
"filterwarnings(action=\"ignore\", category=UserWarning, module=\"pydantic\")\n",
"filterwarnings(action=\"ignore\", category=FutureWarning, module=\"easyocr\")\n",
"# https://github.com/huggingface/transformers/issues/5486:\n",
"os.environ[\"TOKENIZERS_PARALLELISM\"] = \"false\""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can now define the main parameters:"
]
},
{
@ -58,250 +111,61 @@
"metadata": {},
"outputs": [],
"source": [
"import warnings\n",
"from llama_index.embeddings.huggingface import HuggingFaceEmbedding\n",
"from llama_index.llms.huggingface_api import HuggingFaceInferenceAPI\n",
"\n",
"warnings.filterwarnings(action=\"ignore\", category=UserWarning, module=\"pydantic|torch\")\n",
"warnings.filterwarnings(action=\"ignore\", category=FutureWarning, module=\"easyocr\")"
"EMBED_MODEL = HuggingFaceEmbedding(model_name=\"BAAI/bge-small-en-v1.5\")\n",
"MILVUS_URI = str(Path(mkdtemp()) / \"docling.db\")\n",
"GEN_MODEL = HuggingFaceInferenceAPI(\n",
" token=_get_env_from_colab_or_os(\"HF_TOKEN\"),\n",
" model_name=\"mistralai/Mixtral-8x7B-Instruct-v0.1\",\n",
")\n",
"SOURCE = \"https://arxiv.org/pdf/2408.09869\" # Docling Technical Report\n",
"QUERY = \"Which are the main AI models in Docling?\"\n",
"\n",
"embed_dim = len(EMBED_MODEL.get_text_embedding(\"hi\"))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Setup"
"## Using Markdown export"
]
},
{
"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"
"To create a simple RAG pipeline, we can:\n",
"- define a `DoclingReader`, which by default exports to Markdown, and\n",
"- use a standard node parser for these Markdown-based docs, e.g. a `MarkdownNodeParser`"
]
},
{
"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",
"name": "stdout",
"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",
"Q: Which are the main AI models in Docling?\n",
"A: 1. A layout analysis model, an accurate object-detector for page elements. 2. TableFormer, a state-of-the-art table structure recognition model.\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",
")"
"Sources:\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"
"[('3.2 AI models\\n\\nAs part of Docling, we initially release two highly capable AI models to the open-source community, which have been developed and published recently by our team. The first model is a layout analysis model, an accurate object-detector for page elements [13]. The second model is TableFormer [12, 9], a state-of-the-art table structure recognition model. We provide the pre-trained weights (hosted on huggingface) and a separate package for the inference code as docling-ibm-models . Both models are also powering the open-access deepsearch-experience, our cloud-native service for knowledge exploration tasks.',\n",
" {'dl_doc_hash': '556ad9e23b6d2245e36b3208758cf0c8a709382bb4c859eacfe8e73b14e635aa',\n",
" 'Header_2': '3.2 AI models'}),\n",
" (\"5 Applications\\n\\nThanks to the high-quality, richly structured document conversion achieved by Docling, its output qualifies for numerous downstream applications. For example, Docling can provide a base for detailed enterprise document search, passage retrieval or classification use-cases, or support knowledge extraction pipelines, allowing specific treatment of different structures in the document, such as tables, figures, section structure or references. For popular generative AI application patterns, such as retrieval-augmented generation (RAG), we provide quackling , an open-source package which capitalizes on Docling's feature-rich document output to enable document-native optimized vector embedding and chunking. It plugs in seamlessly with LLM frameworks such as LlamaIndex [8]. Since Docling is fast, stable and cheap to run, it also makes for an excellent choice to build document-derived datasets. With its powerful table structure recognition, it provides significant benefit to automated knowledge-base construction [11, 10]. Docling is also integrated within the open IBM data prep kit [6], which implements scalable data transforms to build large-scale multi-modal training datasets.\",\n",
" {'dl_doc_hash': '556ad9e23b6d2245e36b3208758cf0c8a709382bb4c859eacfe8e73b14e635aa',\n",
" 'Header_2': '5 Applications'})]"
]
},
"metadata": {},
@ -310,131 +174,83 @@
],
"source": [
"from llama_index.core import StorageContext, VectorStoreIndex\n",
"from llama_index.core.node_parser import MarkdownNodeParser\n",
"from llama_index.readers.docling import DoclingReader\n",
"from llama_index.vector_stores.milvus import MilvusVectorStore\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",
"reader = DoclingReader()\n",
"node_parser = MarkdownNodeParser()\n",
"\n",
"vector_store = MilvusVectorStore(\n",
" uri=str(Path(mkdtemp()) / \"docling.db\"), # or set as needed\n",
" dim=embed_dim,\n",
" overwrite=True,\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",
" documents=reader.load_data(SOURCE),\n",
" transformations=[node_parser],\n",
" storage_context=StorageContext.from_defaults(vector_store=vector_store),\n",
" embed_model=EMBED_MODEL,\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",
" )"
"result = index.as_query_engine(llm=GEN_MODEL).query(QUERY)\n",
"print(f\"Q: {QUERY}\\nA: {result.response.strip()}\\n\\nSources:\")\n",
"display([(n.text, n.metadata) for n in result.source_nodes])"
]
},
{
"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",
")"
"## Using Docling format"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## RAG"
"To leverage Docling's rich native format, we:\n",
"- create a `DoclingReader` with JSON export type, and\n",
"- employ a `DoclingNodeParser` in order to appropriately parse that Docling format.\n",
"\n",
"Notice how the sources now also contain document-level grounding (e.g. page number or bounding box information):"
]
},
{
"cell_type": "code",
"execution_count": 12,
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Q: Which are the main AI models in Docling?\n",
"A: The main AI models in Docling are a layout analysis model and TableFormer. The layout analysis model is an accurate object-detector for page elements, and TableFormer is a state-of-the-art table structure recognition model.\n",
"\n",
"Sources:\n"
]
},
{
"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"
"[('As part of Docling, we initially release two highly capable AI models to the open-source community, which have been developed and published recently by our team. The first model is a layout analysis model, an accurate object-detector for page elements [13]. The second model is TableFormer [12, 9], a state-of-the-art table structure recognition model. We provide the pre-trained weights (hosted on huggingface) and a separate package for the inference code as docling-ibm-models . Both models are also powering the open-access deepsearch-experience, our cloud-native service for knowledge exploration tasks.',\n",
" {'dl_doc_hash': '556ad9e23b6d2245e36b3208758cf0c8a709382bb4c859eacfe8e73b14e635aa',\n",
" 'path': '#/main-text/37',\n",
" 'heading': '3.2 AI models',\n",
" 'page': 3,\n",
" 'bbox': [107.36903381347656,\n",
" 330.07513427734375,\n",
" 506.29705810546875,\n",
" 407.3725280761719]}),\n",
" ('With Docling , we open-source a very capable and efficient document conversion tool which builds on the powerful, specialized AI models and datasets for layout analysis and table structure recognition we developed and presented in the recent past [12, 13, 9]. Docling is designed as a simple, self-contained python library with permissive license, running entirely locally on commodity hardware. Its code architecture allows for easy extensibility and addition of new features and models.',\n",
" {'dl_doc_hash': '556ad9e23b6d2245e36b3208758cf0c8a709382bb4c859eacfe8e73b14e635aa',\n",
" 'path': '#/main-text/10',\n",
" 'heading': '1 Introduction',\n",
" 'page': 1,\n",
" 'bbox': [107.33261108398438,\n",
" 83.3067626953125,\n",
" 504.0033874511719,\n",
" 136.45367431640625]})]"
]
},
"metadata": {},
@ -442,9 +258,148 @@
}
],
"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)"
"from llama_index.node_parser.docling import DoclingNodeParser\n",
"\n",
"reader = DoclingReader(export_type=DoclingReader.ExportType.JSON)\n",
"node_parser = DoclingNodeParser()\n",
"\n",
"vector_store = MilvusVectorStore(\n",
" uri=str(Path(mkdtemp()) / \"docling.db\"), # or set as needed\n",
" dim=embed_dim,\n",
" overwrite=True,\n",
")\n",
"index = VectorStoreIndex.from_documents(\n",
" documents=reader.load_data(SOURCE),\n",
" transformations=[node_parser],\n",
" storage_context=StorageContext.from_defaults(vector_store=vector_store),\n",
" embed_model=EMBED_MODEL,\n",
")\n",
"result = index.as_query_engine(llm=GEN_MODEL).query(QUERY)\n",
"print(f\"Q: {QUERY}\\nA: {result.response.strip()}\\n\\nSources:\")\n",
"display([(n.text, n.metadata) for n in result.source_nodes])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## With Simple Directory Reader"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"To demonstrate this usage pattern, we first set up a test document directory."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"from pathlib import Path\n",
"from tempfile import mkdtemp\n",
"\n",
"import requests\n",
"\n",
"tmp_dir_path = Path(mkdtemp())\n",
"r = requests.get(SOURCE)\n",
"with open(tmp_dir_path / f\"{Path(SOURCE).name}.pdf\", \"wb\") as out_file:\n",
" out_file.write(r.content)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Using the `reader` and `node_parser` definitions from any of the above variants, usage with `SimpleDirectoryReader` then looks as follows:"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Loading files: 100%|██████████| 1/1 [00:11<00:00, 11.15s/file]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Q: Which are the main AI models in Docling?\n",
"A: The main AI models in Docling are a layout analysis model and TableFormer. The layout analysis model is an accurate object-detector for page elements, and TableFormer is a state-of-the-art table structure recognition model.\n",
"\n",
"Sources:\n"
]
},
{
"data": {
"text/plain": [
"[('As part of Docling, we initially release two highly capable AI models to the open-source community, which have been developed and published recently by our team. The first model is a layout analysis model, an accurate object-detector for page elements [13]. The second model is TableFormer [12, 9], a state-of-the-art table structure recognition model. We provide the pre-trained weights (hosted on huggingface) and a separate package for the inference code as docling-ibm-models . Both models are also powering the open-access deepsearch-experience, our cloud-native service for knowledge exploration tasks.',\n",
" {'file_path': '/var/folders/76/4wwfs06x6835kcwj4186c0nc0000gn/T/tmp4vsev3_r/2408.09869.pdf',\n",
" 'file_name': '2408.09869.pdf',\n",
" 'file_type': 'application/pdf',\n",
" 'file_size': 5566574,\n",
" 'creation_date': '2024-10-09',\n",
" 'last_modified_date': '2024-10-09',\n",
" 'dl_doc_hash': '556ad9e23b6d2245e36b3208758cf0c8a709382bb4c859eacfe8e73b14e635aa',\n",
" 'path': '#/main-text/37',\n",
" 'heading': '3.2 AI models',\n",
" 'page': 3,\n",
" 'bbox': [107.36903381347656,\n",
" 330.07513427734375,\n",
" 506.29705810546875,\n",
" 407.3725280761719]}),\n",
" ('With Docling , we open-source a very capable and efficient document conversion tool which builds on the powerful, specialized AI models and datasets for layout analysis and table structure recognition we developed and presented in the recent past [12, 13, 9]. Docling is designed as a simple, self-contained python library with permissive license, running entirely locally on commodity hardware. Its code architecture allows for easy extensibility and addition of new features and models.',\n",
" {'file_path': '/var/folders/76/4wwfs06x6835kcwj4186c0nc0000gn/T/tmp4vsev3_r/2408.09869.pdf',\n",
" 'file_name': '2408.09869.pdf',\n",
" 'file_type': 'application/pdf',\n",
" 'file_size': 5566574,\n",
" 'creation_date': '2024-10-09',\n",
" 'last_modified_date': '2024-10-09',\n",
" 'dl_doc_hash': '556ad9e23b6d2245e36b3208758cf0c8a709382bb4c859eacfe8e73b14e635aa',\n",
" 'path': '#/main-text/10',\n",
" 'heading': '1 Introduction',\n",
" 'page': 1,\n",
" 'bbox': [107.33261108398438,\n",
" 83.3067626953125,\n",
" 504.0033874511719,\n",
" 136.45367431640625]})]"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"from llama_index.core import SimpleDirectoryReader\n",
"\n",
"dir_reader = SimpleDirectoryReader(\n",
" input_dir=tmp_dir_path,\n",
" file_extractor={\".pdf\": reader},\n",
")\n",
"\n",
"vector_store = MilvusVectorStore(\n",
" uri=str(Path(mkdtemp()) / \"docling.db\"), # or set as needed\n",
" dim=embed_dim,\n",
" overwrite=True,\n",
")\n",
"index = VectorStoreIndex.from_documents(\n",
" documents=dir_reader.load_data(SOURCE),\n",
" transformations=[node_parser],\n",
" storage_context=StorageContext.from_defaults(vector_store=vector_store),\n",
" embed_model=EMBED_MODEL,\n",
")\n",
"result = index.as_query_engine(llm=GEN_MODEL).query(QUERY)\n",
"print(f\"Q: {QUERY}\\nA: {result.response.strip()}\\n\\nSources:\")\n",
"display([(n.text, n.metadata) for n in result.source_nodes])"
]
},
{

272
poetry.lock generated
View File

@ -13,102 +13,102 @@ files = [
[[package]]
name = "aiohttp"
version = "3.10.9"
version = "3.10.10"
description = "Async http client/server framework (asyncio)"
optional = false
python-versions = ">=3.8"
files = [
{file = "aiohttp-3.10.9-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:8b3fb28a9ac8f2558760d8e637dbf27aef1e8b7f1d221e8669a1074d1a266bb2"},
{file = "aiohttp-3.10.9-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:91aa966858593f64c8a65cdefa3d6dc8fe3c2768b159da84c1ddbbb2c01ab4ef"},
{file = "aiohttp-3.10.9-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:63649309da83277f06a15bbdc2a54fbe75efb92caa2c25bb57ca37762789c746"},
{file = "aiohttp-3.10.9-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:e3e7fabedb3fe06933f47f1538df7b3a8d78e13d7167195f51ca47ee12690373"},
{file = "aiohttp-3.10.9-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:5c070430fda1a550a1c3a4c2d7281d3b8cfc0c6715f616e40e3332201a253067"},
{file = "aiohttp-3.10.9-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:51d0a4901b27272ae54e42067bc4b9a90e619a690b4dc43ea5950eb3070afc32"},
{file = "aiohttp-3.10.9-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:fec5fac7aea6c060f317f07494961236434928e6f4374e170ef50b3001e14581"},
{file = "aiohttp-3.10.9-cp310-cp310-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:172ad884bb61ad31ed7beed8be776eb17e7fb423f1c1be836d5cb357a096bf12"},
{file = "aiohttp-3.10.9-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:d646fdd74c25bbdd4a055414f0fe32896c400f38ffbdfc78c68e62812a9e0257"},
{file = "aiohttp-3.10.9-cp310-cp310-musllinux_1_2_i686.whl", hash = "sha256:e86260b76786c28acf0b5fe31c8dca4c2add95098c709b11e8c35b424ebd4f5b"},
{file = "aiohttp-3.10.9-cp310-cp310-musllinux_1_2_ppc64le.whl", hash = "sha256:c7d7cafc11d70fdd8801abfc2ff276744ae4cb39d8060b6b542c7e44e5f2cfc2"},
{file = "aiohttp-3.10.9-cp310-cp310-musllinux_1_2_s390x.whl", hash = "sha256:fc262c3df78c8ff6020c782d9ce02e4bcffe4900ad71c0ecdad59943cba54442"},
{file = "aiohttp-3.10.9-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:482c85cf3d429844396d939b22bc2a03849cb9ad33344689ad1c85697bcba33a"},
{file = "aiohttp-3.10.9-cp310-cp310-win32.whl", hash = "sha256:aeebd3061f6f1747c011e1d0b0b5f04f9f54ad1a2ca183e687e7277bef2e0da2"},
{file = "aiohttp-3.10.9-cp310-cp310-win_amd64.whl", hash = "sha256:fa430b871220dc62572cef9c69b41e0d70fcb9d486a4a207a5de4c1f25d82593"},
{file = "aiohttp-3.10.9-cp311-cp311-macosx_10_9_universal2.whl", hash = "sha256:16e6a51d8bc96b77f04a6764b4ad03eeef43baa32014fce71e882bd71302c7e4"},
{file = "aiohttp-3.10.9-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:8bd9125dd0cc8ebd84bff2be64b10fdba7dc6fd7be431b5eaf67723557de3a31"},
{file = "aiohttp-3.10.9-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:dcf354661f54e6a49193d0b5653a1b011ba856e0b7a76bda2c33e4c6892f34ea"},
{file = "aiohttp-3.10.9-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:42775de0ca04f90c10c5c46291535ec08e9bcc4756f1b48f02a0657febe89b10"},
{file = "aiohttp-3.10.9-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:87d1e4185c5d7187684d41ebb50c9aeaaaa06ca1875f4c57593071b0409d2444"},
{file = "aiohttp-3.10.9-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:c2695c61cf53a5d4345a43d689f37fc0f6d3a2dc520660aec27ec0f06288d1f9"},
{file = "aiohttp-3.10.9-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:8a3f063b41cc06e8d0b3fcbbfc9c05b7420f41287e0cd4f75ce0a1f3d80729e6"},
{file = "aiohttp-3.10.9-cp311-cp311-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:2d37f4718002863b82c6f391c8efd4d3a817da37030a29e2682a94d2716209de"},
{file = "aiohttp-3.10.9-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:2746d8994ebca1bdc55a1e998feff4e94222da709623bb18f6e5cfec8ec01baf"},
{file = "aiohttp-3.10.9-cp311-cp311-musllinux_1_2_i686.whl", hash = "sha256:6f3c6648aa123bcd73d6f26607d59967b607b0da8ffcc27d418a4b59f4c98c7c"},
{file = "aiohttp-3.10.9-cp311-cp311-musllinux_1_2_ppc64le.whl", hash = "sha256:558b3d223fd631ad134d89adea876e7fdb4c93c849ef195049c063ada82b7d08"},
{file = "aiohttp-3.10.9-cp311-cp311-musllinux_1_2_s390x.whl", hash = "sha256:4e6cb75f8ddd9c2132d00bc03c9716add57f4beff1263463724f6398b813e7eb"},
{file = "aiohttp-3.10.9-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:608cecd8d58d285bfd52dbca5b6251ca8d6ea567022c8a0eaae03c2589cd9af9"},
{file = "aiohttp-3.10.9-cp311-cp311-win32.whl", hash = "sha256:36d4fba838be5f083f5490ddd281813b44d69685db910907636bc5dca6322316"},
{file = "aiohttp-3.10.9-cp311-cp311-win_amd64.whl", hash = "sha256:8be1a65487bdfc285bd5e9baf3208c2132ca92a9b4020e9f27df1b16fab998a9"},
{file = "aiohttp-3.10.9-cp312-cp312-macosx_10_9_universal2.whl", hash = "sha256:4fd16b30567c5b8e167923be6e027eeae0f20cf2b8a26b98a25115f28ad48ee0"},
{file = "aiohttp-3.10.9-cp312-cp312-macosx_10_9_x86_64.whl", hash = "sha256:40ff5b7660f903dc587ed36ef08a88d46840182d9d4b5694e7607877ced698a1"},
{file = "aiohttp-3.10.9-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:4edc3fd701e2b9a0d605a7b23d3de4ad23137d23fc0dbab726aa71d92f11aaaf"},
{file = "aiohttp-3.10.9-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:e525b69ee8a92c146ae5b4da9ecd15e518df4d40003b01b454ad694a27f498b5"},
{file = "aiohttp-3.10.9-cp312-cp312-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:5002a02c17fcfd796d20bac719981d2fca9c006aac0797eb8f430a58e9d12431"},
{file = "aiohttp-3.10.9-cp312-cp312-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:fd4ceeae2fb8cabdd1b71c82bfdd39662473d3433ec95b962200e9e752fb70d0"},
{file = "aiohttp-3.10.9-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:d6e395c3d1f773cf0651cd3559e25182eb0c03a2777b53b4575d8adc1149c6e9"},
{file = "aiohttp-3.10.9-cp312-cp312-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:bbdb8def5268f3f9cd753a265756f49228a20ed14a480d151df727808b4531dd"},
{file = "aiohttp-3.10.9-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:f82ace0ec57c94aaf5b0e118d4366cff5889097412c75aa14b4fd5fc0c44ee3e"},
{file = "aiohttp-3.10.9-cp312-cp312-musllinux_1_2_i686.whl", hash = "sha256:6ebdc3b3714afe1b134b3bbeb5f745eed3ecbcff92ab25d80e4ef299e83a5465"},
{file = "aiohttp-3.10.9-cp312-cp312-musllinux_1_2_ppc64le.whl", hash = "sha256:f9ca09414003c0e96a735daa1f071f7d7ed06962ef4fa29ceb6c80d06696d900"},
{file = "aiohttp-3.10.9-cp312-cp312-musllinux_1_2_s390x.whl", hash = "sha256:1298b854fd31d0567cbb916091be9d3278168064fca88e70b8468875ef9ff7e7"},
{file = "aiohttp-3.10.9-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:60ad5b8a7452c0f5645c73d4dad7490afd6119d453d302cd5b72b678a85d6044"},
{file = "aiohttp-3.10.9-cp312-cp312-win32.whl", hash = "sha256:1a0ee6c0d590c917f1b9629371fce5f3d3f22c317aa96fbdcce3260754d7ea21"},
{file = "aiohttp-3.10.9-cp312-cp312-win_amd64.whl", hash = "sha256:c46131c6112b534b178d4e002abe450a0a29840b61413ac25243f1291613806a"},
{file = "aiohttp-3.10.9-cp313-cp313-macosx_10_13_universal2.whl", hash = "sha256:2bd9f3eac515c16c4360a6a00c38119333901b8590fe93c3257a9b536026594d"},
{file = "aiohttp-3.10.9-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:8cc0d13b4e3b1362d424ce3f4e8c79e1f7247a00d792823ffd640878abf28e56"},
{file = "aiohttp-3.10.9-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:ba1a599255ad6a41022e261e31bc2f6f9355a419575b391f9655c4d9e5df5ff5"},
{file = "aiohttp-3.10.9-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:776e9f3c9b377fcf097c4a04b241b15691e6662d850168642ff976780609303c"},
{file = "aiohttp-3.10.9-cp313-cp313-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:8debb45545ad95b58cc16c3c1cc19ad82cffcb106db12b437885dbee265f0ab5"},
{file = "aiohttp-3.10.9-cp313-cp313-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:c2555e4949c8d8782f18ef20e9d39730d2656e218a6f1a21a4c4c0b56546a02e"},
{file = "aiohttp-3.10.9-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:c54dc329cd44f7f7883a9f4baaefe686e8b9662e2c6c184ea15cceee587d8d69"},
{file = "aiohttp-3.10.9-cp313-cp313-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:e709d6ac598c5416f879bb1bae3fd751366120ac3fa235a01de763537385d036"},
{file = "aiohttp-3.10.9-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:17c272cfe7b07a5bb0c6ad3f234e0c336fb53f3bf17840f66bd77b5815ab3d16"},
{file = "aiohttp-3.10.9-cp313-cp313-musllinux_1_2_i686.whl", hash = "sha256:0c21c82df33b264216abffff9f8370f303dab65d8eee3767efbbd2734363f677"},
{file = "aiohttp-3.10.9-cp313-cp313-musllinux_1_2_ppc64le.whl", hash = "sha256:9331dd34145ff105177855017920dde140b447049cd62bb589de320fd6ddd582"},
{file = "aiohttp-3.10.9-cp313-cp313-musllinux_1_2_s390x.whl", hash = "sha256:ac3196952c673822ebed8871cf8802e17254fff2a2ed4835d9c045d9b88c5ec7"},
{file = "aiohttp-3.10.9-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:2c33fa6e10bb7ed262e3ff03cc69d52869514f16558db0626a7c5c61dde3c29f"},
{file = "aiohttp-3.10.9-cp313-cp313-win32.whl", hash = "sha256:a14e4b672c257a6b94fe934ee62666bacbc8e45b7876f9dd9502d0f0fe69db16"},
{file = "aiohttp-3.10.9-cp313-cp313-win_amd64.whl", hash = "sha256:a35ed3d03910785f7d9d6f5381f0c24002b2b888b298e6f941b2fc94c5055fcd"},
{file = "aiohttp-3.10.9-cp38-cp38-macosx_10_9_universal2.whl", hash = "sha256:5f392ef50e22c31fa49b5a46af7f983fa3f118f3eccb8522063bee8bfa6755f8"},
{file = "aiohttp-3.10.9-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:d1f5c9169e26db6a61276008582d945405b8316aae2bb198220466e68114a0f5"},
{file = "aiohttp-3.10.9-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:8d9d10d10ec27c0d46ddaecc3c5598c4db9ce4e6398ca872cdde0525765caa2f"},
{file = "aiohttp-3.10.9-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:d97273a52d7f89a75b11ec386f786d3da7723d7efae3034b4dda79f6f093edc1"},
{file = "aiohttp-3.10.9-cp38-cp38-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:d271f770b52e32236d945911b2082f9318e90ff835d45224fa9e28374303f729"},
{file = "aiohttp-3.10.9-cp38-cp38-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:7003f33f5f7da1eb02f0446b0f8d2ccf57d253ca6c2e7a5732d25889da82b517"},
{file = "aiohttp-3.10.9-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:a6e00c8a92e7663ed2be6fcc08a2997ff06ce73c8080cd0df10cc0321a3168d7"},
{file = "aiohttp-3.10.9-cp38-cp38-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:a61df62966ce6507aafab24e124e0c3a1cfbe23c59732987fc0fd0d71daa0b88"},
{file = "aiohttp-3.10.9-cp38-cp38-musllinux_1_2_aarch64.whl", hash = "sha256:60555211a006d26e1a389222e3fab8cd379f28e0fbf7472ee55b16c6c529e3a6"},
{file = "aiohttp-3.10.9-cp38-cp38-musllinux_1_2_i686.whl", hash = "sha256:d15a29424e96fad56dc2f3abed10a89c50c099f97d2416520c7a543e8fddf066"},
{file = "aiohttp-3.10.9-cp38-cp38-musllinux_1_2_ppc64le.whl", hash = "sha256:a19caae0d670771ea7854ca30df76f676eb47e0fd9b2ee4392d44708f272122d"},
{file = "aiohttp-3.10.9-cp38-cp38-musllinux_1_2_s390x.whl", hash = "sha256:99f9678bf0e2b1b695e8028fedac24ab6770937932eda695815d5a6618c37e04"},
{file = "aiohttp-3.10.9-cp38-cp38-musllinux_1_2_x86_64.whl", hash = "sha256:2914caa46054f3b5ff910468d686742ff8cff54b8a67319d75f5d5945fd0a13d"},
{file = "aiohttp-3.10.9-cp38-cp38-win32.whl", hash = "sha256:0bc059ecbce835630e635879f5f480a742e130d9821fbe3d2f76610a6698ee25"},
{file = "aiohttp-3.10.9-cp38-cp38-win_amd64.whl", hash = "sha256:e883b61b75ca6efc2541fcd52a5c8ccfe288b24d97e20ac08fdf343b8ac672ea"},
{file = "aiohttp-3.10.9-cp39-cp39-macosx_10_9_universal2.whl", hash = "sha256:fcd546782d03181b0b1d20b43d612429a90a68779659ba8045114b867971ab71"},
{file = "aiohttp-3.10.9-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:85711eec2d875cd88c7eb40e734c4ca6d9ae477d6f26bd2b5bb4f7f60e41b156"},
{file = "aiohttp-3.10.9-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:02d1d6610588bcd743fae827bd6f2e47e0d09b346f230824b4c6fb85c6065f9c"},
{file = "aiohttp-3.10.9-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:3668d0c2a4d23fb136a753eba42caa2c0abbd3d9c5c87ee150a716a16c6deec1"},
{file = "aiohttp-3.10.9-cp39-cp39-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:d7c071235a47d407b0e93aa6262b49422dbe48d7d8566e1158fecc91043dd948"},
{file = "aiohttp-3.10.9-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:ac74e794e3aee92ae8f571bfeaa103a141e409863a100ab63a253b1c53b707eb"},
{file = "aiohttp-3.10.9-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:2bbf94d4a0447705b7775417ca8bb8086cc5482023a6e17cdc8f96d0b1b5aba6"},
{file = "aiohttp-3.10.9-cp39-cp39-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:cb0b2d5d51f96b6cc19e6ab46a7b684be23240426ae951dcdac9639ab111b45e"},
{file = "aiohttp-3.10.9-cp39-cp39-musllinux_1_2_aarch64.whl", hash = "sha256:e83dfefb4f7d285c2d6a07a22268344a97d61579b3e0dce482a5be0251d672ab"},
{file = "aiohttp-3.10.9-cp39-cp39-musllinux_1_2_i686.whl", hash = "sha256:f0a44bb40b6aaa4fb9a5c1ee07880570ecda2065433a96ccff409c9c20c1624a"},
{file = "aiohttp-3.10.9-cp39-cp39-musllinux_1_2_ppc64le.whl", hash = "sha256:c2b627d3c8982691b06d89d31093cee158c30629fdfebe705a91814d49b554f8"},
{file = "aiohttp-3.10.9-cp39-cp39-musllinux_1_2_s390x.whl", hash = "sha256:03690541e4cc866eef79626cfa1ef4dd729c5c1408600c8cb9e12e1137eed6ab"},
{file = "aiohttp-3.10.9-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:ad3675c126f2a95bde637d162f8231cff6bc0bc9fbe31bd78075f9ff7921e322"},
{file = "aiohttp-3.10.9-cp39-cp39-win32.whl", hash = "sha256:1321658f12b6caffafdc35cfba6c882cb014af86bef4e78c125e7e794dfb927b"},
{file = "aiohttp-3.10.9-cp39-cp39-win_amd64.whl", hash = "sha256:9fdf5c839bf95fc67be5794c780419edb0dbef776edcfc6c2e5e2ffd5ee755fa"},
{file = "aiohttp-3.10.9.tar.gz", hash = "sha256:143b0026a9dab07a05ad2dd9e46aa859bffdd6348ddc5967b42161168c24f857"},
{file = "aiohttp-3.10.10-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:be7443669ae9c016b71f402e43208e13ddf00912f47f623ee5994e12fc7d4b3f"},
{file = "aiohttp-3.10.10-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:7b06b7843929e41a94ea09eb1ce3927865387e3e23ebe108e0d0d09b08d25be9"},
{file = "aiohttp-3.10.10-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:333cf6cf8e65f6a1e06e9eb3e643a0c515bb850d470902274239fea02033e9a8"},
{file = "aiohttp-3.10.10-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:274cfa632350225ce3fdeb318c23b4a10ec25c0e2c880eff951a3842cf358ac1"},
{file = "aiohttp-3.10.10-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:d9e5e4a85bdb56d224f412d9c98ae4cbd032cc4f3161818f692cd81766eee65a"},
{file = "aiohttp-3.10.10-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:2b606353da03edcc71130b52388d25f9a30a126e04caef1fd637e31683033abd"},
{file = "aiohttp-3.10.10-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:ab5a5a0c7a7991d90446a198689c0535be89bbd6b410a1f9a66688f0880ec026"},
{file = "aiohttp-3.10.10-cp310-cp310-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:578a4b875af3e0daaf1ac6fa983d93e0bbfec3ead753b6d6f33d467100cdc67b"},
{file = "aiohttp-3.10.10-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:8105fd8a890df77b76dd3054cddf01a879fc13e8af576805d667e0fa0224c35d"},
{file = "aiohttp-3.10.10-cp310-cp310-musllinux_1_2_i686.whl", hash = "sha256:3bcd391d083f636c06a68715e69467963d1f9600f85ef556ea82e9ef25f043f7"},
{file = "aiohttp-3.10.10-cp310-cp310-musllinux_1_2_ppc64le.whl", hash = "sha256:fbc6264158392bad9df19537e872d476f7c57adf718944cc1e4495cbabf38e2a"},
{file = "aiohttp-3.10.10-cp310-cp310-musllinux_1_2_s390x.whl", hash = "sha256:e48d5021a84d341bcaf95c8460b152cfbad770d28e5fe14a768988c461b821bc"},
{file = "aiohttp-3.10.10-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:2609e9ab08474702cc67b7702dbb8a80e392c54613ebe80db7e8dbdb79837c68"},
{file = "aiohttp-3.10.10-cp310-cp310-win32.whl", hash = "sha256:84afcdea18eda514c25bc68b9af2a2b1adea7c08899175a51fe7c4fb6d551257"},
{file = "aiohttp-3.10.10-cp310-cp310-win_amd64.whl", hash = "sha256:9c72109213eb9d3874f7ac8c0c5fa90e072d678e117d9061c06e30c85b4cf0e6"},
{file = "aiohttp-3.10.10-cp311-cp311-macosx_10_9_universal2.whl", hash = "sha256:c30a0eafc89d28e7f959281b58198a9fa5e99405f716c0289b7892ca345fe45f"},
{file = "aiohttp-3.10.10-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:258c5dd01afc10015866114e210fb7365f0d02d9d059c3c3415382ab633fcbcb"},
{file = "aiohttp-3.10.10-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:15ecd889a709b0080f02721255b3f80bb261c2293d3c748151274dfea93ac871"},
{file = "aiohttp-3.10.10-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:f3935f82f6f4a3820270842e90456ebad3af15810cf65932bd24da4463bc0a4c"},
{file = "aiohttp-3.10.10-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:413251f6fcf552a33c981c4709a6bba37b12710982fec8e558ae944bfb2abd38"},
{file = "aiohttp-3.10.10-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:d1720b4f14c78a3089562b8875b53e36b51c97c51adc53325a69b79b4b48ebcb"},
{file = "aiohttp-3.10.10-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:679abe5d3858b33c2cf74faec299fda60ea9de62916e8b67e625d65bf069a3b7"},
{file = "aiohttp-3.10.10-cp311-cp311-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:79019094f87c9fb44f8d769e41dbb664d6e8fcfd62f665ccce36762deaa0e911"},
{file = "aiohttp-3.10.10-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:fe2fb38c2ed905a2582948e2de560675e9dfbee94c6d5ccdb1301c6d0a5bf092"},
{file = "aiohttp-3.10.10-cp311-cp311-musllinux_1_2_i686.whl", hash = "sha256:a3f00003de6eba42d6e94fabb4125600d6e484846dbf90ea8e48a800430cc142"},
{file = "aiohttp-3.10.10-cp311-cp311-musllinux_1_2_ppc64le.whl", hash = "sha256:1bbb122c557a16fafc10354b9d99ebf2f2808a660d78202f10ba9d50786384b9"},
{file = "aiohttp-3.10.10-cp311-cp311-musllinux_1_2_s390x.whl", hash = "sha256:30ca7c3b94708a9d7ae76ff281b2f47d8eaf2579cd05971b5dc681db8caac6e1"},
{file = "aiohttp-3.10.10-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:df9270660711670e68803107d55c2b5949c2e0f2e4896da176e1ecfc068b974a"},
{file = "aiohttp-3.10.10-cp311-cp311-win32.whl", hash = "sha256:aafc8ee9b742ce75044ae9a4d3e60e3d918d15a4c2e08a6c3c3e38fa59b92d94"},
{file = "aiohttp-3.10.10-cp311-cp311-win_amd64.whl", hash = "sha256:362f641f9071e5f3ee6f8e7d37d5ed0d95aae656adf4ef578313ee585b585959"},
{file = "aiohttp-3.10.10-cp312-cp312-macosx_10_9_universal2.whl", hash = "sha256:9294bbb581f92770e6ed5c19559e1e99255e4ca604a22c5c6397b2f9dd3ee42c"},
{file = "aiohttp-3.10.10-cp312-cp312-macosx_10_9_x86_64.whl", hash = "sha256:a8fa23fe62c436ccf23ff930149c047f060c7126eae3ccea005f0483f27b2e28"},
{file = "aiohttp-3.10.10-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:5c6a5b8c7926ba5d8545c7dd22961a107526562da31a7a32fa2456baf040939f"},
{file = "aiohttp-3.10.10-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:007ec22fbc573e5eb2fb7dec4198ef8f6bf2fe4ce20020798b2eb5d0abda6138"},
{file = "aiohttp-3.10.10-cp312-cp312-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:9627cc1a10c8c409b5822a92d57a77f383b554463d1884008e051c32ab1b3742"},
{file = "aiohttp-3.10.10-cp312-cp312-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:50edbcad60d8f0e3eccc68da67f37268b5144ecc34d59f27a02f9611c1d4eec7"},
{file = "aiohttp-3.10.10-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:a45d85cf20b5e0d0aa5a8dca27cce8eddef3292bc29d72dcad1641f4ed50aa16"},
{file = "aiohttp-3.10.10-cp312-cp312-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:0b00807e2605f16e1e198f33a53ce3c4523114059b0c09c337209ae55e3823a8"},
{file = "aiohttp-3.10.10-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:f2d4324a98062be0525d16f768a03e0bbb3b9fe301ceee99611dc9a7953124e6"},
{file = "aiohttp-3.10.10-cp312-cp312-musllinux_1_2_i686.whl", hash = "sha256:438cd072f75bb6612f2aca29f8bd7cdf6e35e8f160bc312e49fbecab77c99e3a"},
{file = "aiohttp-3.10.10-cp312-cp312-musllinux_1_2_ppc64le.whl", hash = "sha256:baa42524a82f75303f714108fea528ccacf0386af429b69fff141ffef1c534f9"},
{file = "aiohttp-3.10.10-cp312-cp312-musllinux_1_2_s390x.whl", hash = "sha256:a7d8d14fe962153fc681f6366bdec33d4356f98a3e3567782aac1b6e0e40109a"},
{file = "aiohttp-3.10.10-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:c1277cd707c465cd09572a774559a3cc7c7a28802eb3a2a9472588f062097205"},
{file = "aiohttp-3.10.10-cp312-cp312-win32.whl", hash = "sha256:59bb3c54aa420521dc4ce3cc2c3fe2ad82adf7b09403fa1f48ae45c0cbde6628"},
{file = "aiohttp-3.10.10-cp312-cp312-win_amd64.whl", hash = "sha256:0e1b370d8007c4ae31ee6db7f9a2fe801a42b146cec80a86766e7ad5c4a259cf"},
{file = "aiohttp-3.10.10-cp313-cp313-macosx_10_13_universal2.whl", hash = "sha256:ad7593bb24b2ab09e65e8a1d385606f0f47c65b5a2ae6c551db67d6653e78c28"},
{file = "aiohttp-3.10.10-cp313-cp313-macosx_10_13_x86_64.whl", hash = "sha256:1eb89d3d29adaf533588f209768a9c02e44e4baf832b08118749c5fad191781d"},
{file = "aiohttp-3.10.10-cp313-cp313-macosx_11_0_arm64.whl", hash = "sha256:3fe407bf93533a6fa82dece0e74dbcaaf5d684e5a51862887f9eaebe6372cd79"},
{file = "aiohttp-3.10.10-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:50aed5155f819873d23520919e16703fc8925e509abbb1a1491b0087d1cd969e"},
{file = "aiohttp-3.10.10-cp313-cp313-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:4f05e9727ce409358baa615dbeb9b969db94324a79b5a5cea45d39bdb01d82e6"},
{file = "aiohttp-3.10.10-cp313-cp313-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:3dffb610a30d643983aeb185ce134f97f290f8935f0abccdd32c77bed9388b42"},
{file = "aiohttp-3.10.10-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:aa6658732517ddabe22c9036479eabce6036655ba87a0224c612e1ae6af2087e"},
{file = "aiohttp-3.10.10-cp313-cp313-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:741a46d58677d8c733175d7e5aa618d277cd9d880301a380fd296975a9cdd7bc"},
{file = "aiohttp-3.10.10-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:e00e3505cd80440f6c98c6d69269dcc2a119f86ad0a9fd70bccc59504bebd68a"},
{file = "aiohttp-3.10.10-cp313-cp313-musllinux_1_2_i686.whl", hash = "sha256:ffe595f10566f8276b76dc3a11ae4bb7eba1aac8ddd75811736a15b0d5311414"},
{file = "aiohttp-3.10.10-cp313-cp313-musllinux_1_2_ppc64le.whl", hash = "sha256:bdfcf6443637c148c4e1a20c48c566aa694fa5e288d34b20fcdc58507882fed3"},
{file = "aiohttp-3.10.10-cp313-cp313-musllinux_1_2_s390x.whl", hash = "sha256:d183cf9c797a5291e8301790ed6d053480ed94070637bfaad914dd38b0981f67"},
{file = "aiohttp-3.10.10-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:77abf6665ae54000b98b3c742bc6ea1d1fb31c394bcabf8b5d2c1ac3ebfe7f3b"},
{file = "aiohttp-3.10.10-cp313-cp313-win32.whl", hash = "sha256:4470c73c12cd9109db8277287d11f9dd98f77fc54155fc71a7738a83ffcc8ea8"},
{file = "aiohttp-3.10.10-cp313-cp313-win_amd64.whl", hash = "sha256:486f7aabfa292719a2753c016cc3a8f8172965cabb3ea2e7f7436c7f5a22a151"},
{file = "aiohttp-3.10.10-cp38-cp38-macosx_10_9_universal2.whl", hash = "sha256:1b66ccafef7336a1e1f0e389901f60c1d920102315a56df85e49552308fc0486"},
{file = "aiohttp-3.10.10-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:acd48d5b80ee80f9432a165c0ac8cbf9253eaddb6113269a5e18699b33958dbb"},
{file = "aiohttp-3.10.10-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:3455522392fb15ff549d92fbf4b73b559d5e43dc522588f7eb3e54c3f38beee7"},
{file = "aiohttp-3.10.10-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:45c3b868724137f713a38376fef8120c166d1eadd50da1855c112fe97954aed8"},
{file = "aiohttp-3.10.10-cp38-cp38-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:da1dee8948d2137bb51fbb8a53cce6b1bcc86003c6b42565f008438b806cccd8"},
{file = "aiohttp-3.10.10-cp38-cp38-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:c5ce2ce7c997e1971b7184ee37deb6ea9922ef5163c6ee5aa3c274b05f9e12fa"},
{file = "aiohttp-3.10.10-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:28529e08fde6f12eba8677f5a8608500ed33c086f974de68cc65ab218713a59d"},
{file = "aiohttp-3.10.10-cp38-cp38-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:f7db54c7914cc99d901d93a34704833568d86c20925b2762f9fa779f9cd2e70f"},
{file = "aiohttp-3.10.10-cp38-cp38-musllinux_1_2_aarch64.whl", hash = "sha256:03a42ac7895406220124c88911ebee31ba8b2d24c98507f4a8bf826b2937c7f2"},
{file = "aiohttp-3.10.10-cp38-cp38-musllinux_1_2_i686.whl", hash = "sha256:7e338c0523d024fad378b376a79faff37fafb3c001872a618cde1d322400a572"},
{file = "aiohttp-3.10.10-cp38-cp38-musllinux_1_2_ppc64le.whl", hash = "sha256:038f514fe39e235e9fef6717fbf944057bfa24f9b3db9ee551a7ecf584b5b480"},
{file = "aiohttp-3.10.10-cp38-cp38-musllinux_1_2_s390x.whl", hash = "sha256:64f6c17757251e2b8d885d728b6433d9d970573586a78b78ba8929b0f41d045a"},
{file = "aiohttp-3.10.10-cp38-cp38-musllinux_1_2_x86_64.whl", hash = "sha256:93429602396f3383a797a2a70e5f1de5df8e35535d7806c9f91df06f297e109b"},
{file = "aiohttp-3.10.10-cp38-cp38-win32.whl", hash = "sha256:c823bc3971c44ab93e611ab1a46b1eafeae474c0c844aff4b7474287b75fe49c"},
{file = "aiohttp-3.10.10-cp38-cp38-win_amd64.whl", hash = "sha256:54ca74df1be3c7ca1cf7f4c971c79c2daf48d9aa65dea1a662ae18926f5bc8ce"},
{file = "aiohttp-3.10.10-cp39-cp39-macosx_10_9_universal2.whl", hash = "sha256:01948b1d570f83ee7bbf5a60ea2375a89dfb09fd419170e7f5af029510033d24"},
{file = "aiohttp-3.10.10-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:9fc1500fd2a952c5c8e3b29aaf7e3cc6e27e9cfc0a8819b3bce48cc1b849e4cc"},
{file = "aiohttp-3.10.10-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:f614ab0c76397661b90b6851a030004dac502e48260ea10f2441abd2207fbcc7"},
{file = "aiohttp-3.10.10-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:00819de9e45d42584bed046314c40ea7e9aea95411b38971082cad449392b08c"},
{file = "aiohttp-3.10.10-cp39-cp39-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:05646ebe6b94cc93407b3bf34b9eb26c20722384d068eb7339de802154d61bc5"},
{file = "aiohttp-3.10.10-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:998f3bd3cfc95e9424a6acd7840cbdd39e45bc09ef87533c006f94ac47296090"},
{file = "aiohttp-3.10.10-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:d9010c31cd6fa59438da4e58a7f19e4753f7f264300cd152e7f90d4602449762"},
{file = "aiohttp-3.10.10-cp39-cp39-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:7ea7ffc6d6d6f8a11e6f40091a1040995cdff02cfc9ba4c2f30a516cb2633554"},
{file = "aiohttp-3.10.10-cp39-cp39-musllinux_1_2_aarch64.whl", hash = "sha256:ef9c33cc5cbca35808f6c74be11eb7f5f6b14d2311be84a15b594bd3e58b5527"},
{file = "aiohttp-3.10.10-cp39-cp39-musllinux_1_2_i686.whl", hash = "sha256:ce0cdc074d540265bfeb31336e678b4e37316849d13b308607efa527e981f5c2"},
{file = "aiohttp-3.10.10-cp39-cp39-musllinux_1_2_ppc64le.whl", hash = "sha256:597a079284b7ee65ee102bc3a6ea226a37d2b96d0418cc9047490f231dc09fe8"},
{file = "aiohttp-3.10.10-cp39-cp39-musllinux_1_2_s390x.whl", hash = "sha256:7789050d9e5d0c309c706953e5e8876e38662d57d45f936902e176d19f1c58ab"},
{file = "aiohttp-3.10.10-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:e7f8b04d83483577fd9200461b057c9f14ced334dcb053090cea1da9c8321a91"},
{file = "aiohttp-3.10.10-cp39-cp39-win32.whl", hash = "sha256:c02a30b904282777d872266b87b20ed8cc0d1501855e27f831320f471d54d983"},
{file = "aiohttp-3.10.10-cp39-cp39-win_amd64.whl", hash = "sha256:edfe3341033a6b53a5c522c802deb2079eee5cbfbb0af032a55064bd65c73a23"},
{file = "aiohttp-3.10.10.tar.gz", hash = "sha256:0631dd7c9f0822cc61c88586ca76d5b5ada26538097d0f1df510b082bad3411a"},
]
[package.dependencies]
@ -831,33 +831,37 @@ vision = ["Pillow (>=9.4.0)"]
[[package]]
name = "debugpy"
version = "1.8.6"
version = "1.8.7"
description = "An implementation of the Debug Adapter Protocol for Python"
optional = false
python-versions = ">=3.8"
files = [
{file = "debugpy-1.8.6-cp310-cp310-macosx_14_0_x86_64.whl", hash = "sha256:30f467c5345d9dfdcc0afdb10e018e47f092e383447500f125b4e013236bf14b"},
{file = "debugpy-1.8.6-cp310-cp310-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:5d73d8c52614432f4215d0fe79a7e595d0dd162b5c15233762565be2f014803b"},
{file = "debugpy-1.8.6-cp310-cp310-win32.whl", hash = "sha256:e3e182cd98eac20ee23a00653503315085b29ab44ed66269482349d307b08df9"},
{file = "debugpy-1.8.6-cp310-cp310-win_amd64.whl", hash = "sha256:e3a82da039cfe717b6fb1886cbbe5c4a3f15d7df4765af857f4307585121c2dd"},
{file = "debugpy-1.8.6-cp311-cp311-macosx_14_0_universal2.whl", hash = "sha256:67479a94cf5fd2c2d88f9615e087fcb4fec169ec780464a3f2ba4a9a2bb79955"},
{file = "debugpy-1.8.6-cp311-cp311-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:9fb8653f6cbf1dd0a305ac1aa66ec246002145074ea57933978346ea5afdf70b"},
{file = "debugpy-1.8.6-cp311-cp311-win32.whl", hash = "sha256:cdaf0b9691879da2d13fa39b61c01887c34558d1ff6e5c30e2eb698f5384cd43"},
{file = "debugpy-1.8.6-cp311-cp311-win_amd64.whl", hash = "sha256:43996632bee7435583952155c06881074b9a742a86cee74e701d87ca532fe833"},
{file = "debugpy-1.8.6-cp312-cp312-macosx_14_0_universal2.whl", hash = "sha256:db891b141fc6ee4b5fc6d1cc8035ec329cabc64bdd2ae672b4550c87d4ecb128"},
{file = "debugpy-1.8.6-cp312-cp312-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:567419081ff67da766c898ccf21e79f1adad0e321381b0dfc7a9c8f7a9347972"},
{file = "debugpy-1.8.6-cp312-cp312-win32.whl", hash = "sha256:c9834dfd701a1f6bf0f7f0b8b1573970ae99ebbeee68314116e0ccc5c78eea3c"},
{file = "debugpy-1.8.6-cp312-cp312-win_amd64.whl", hash = "sha256:e4ce0570aa4aca87137890d23b86faeadf184924ad892d20c54237bcaab75d8f"},
{file = "debugpy-1.8.6-cp38-cp38-macosx_14_0_x86_64.whl", hash = "sha256:df5dc9eb4ca050273b8e374a4cd967c43be1327eeb42bfe2f58b3cdfe7c68dcb"},
{file = "debugpy-1.8.6-cp38-cp38-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:0a85707c6a84b0c5b3db92a2df685b5230dd8fb8c108298ba4f11dba157a615a"},
{file = "debugpy-1.8.6-cp38-cp38-win32.whl", hash = "sha256:538c6cdcdcdad310bbefd96d7850be1cd46e703079cc9e67d42a9ca776cdc8a8"},
{file = "debugpy-1.8.6-cp38-cp38-win_amd64.whl", hash = "sha256:22140bc02c66cda6053b6eb56dfe01bbe22a4447846581ba1dd6df2c9f97982d"},
{file = "debugpy-1.8.6-cp39-cp39-macosx_14_0_x86_64.whl", hash = "sha256:c1cef65cffbc96e7b392d9178dbfd524ab0750da6c0023c027ddcac968fd1caa"},
{file = "debugpy-1.8.6-cp39-cp39-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f1e60bd06bb3cc5c0e957df748d1fab501e01416c43a7bdc756d2a992ea1b881"},
{file = "debugpy-1.8.6-cp39-cp39-win32.whl", hash = "sha256:f7158252803d0752ed5398d291dee4c553bb12d14547c0e1843ab74ee9c31123"},
{file = "debugpy-1.8.6-cp39-cp39-win_amd64.whl", hash = "sha256:3358aa619a073b620cd0d51d8a6176590af24abcc3fe2e479929a154bf591b51"},
{file = "debugpy-1.8.6-py2.py3-none-any.whl", hash = "sha256:b48892df4d810eff21d3ef37274f4c60d32cdcafc462ad5647239036b0f0649f"},
{file = "debugpy-1.8.6.zip", hash = "sha256:c931a9371a86784cee25dec8d65bc2dc7a21f3f1552e3833d9ef8f919d22280a"},
{file = "debugpy-1.8.7-cp310-cp310-macosx_14_0_x86_64.whl", hash = "sha256:95fe04a573b8b22896c404365e03f4eda0ce0ba135b7667a1e57bd079793b96b"},
{file = "debugpy-1.8.7-cp310-cp310-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:628a11f4b295ffb4141d8242a9bb52b77ad4a63a2ad19217a93be0f77f2c28c9"},
{file = "debugpy-1.8.7-cp310-cp310-win32.whl", hash = "sha256:85ce9c1d0eebf622f86cc68618ad64bf66c4fc3197d88f74bb695a416837dd55"},
{file = "debugpy-1.8.7-cp310-cp310-win_amd64.whl", hash = "sha256:29e1571c276d643757ea126d014abda081eb5ea4c851628b33de0c2b6245b037"},
{file = "debugpy-1.8.7-cp311-cp311-macosx_14_0_universal2.whl", hash = "sha256:caf528ff9e7308b74a1749c183d6808ffbedbb9fb6af78b033c28974d9b8831f"},
{file = "debugpy-1.8.7-cp311-cp311-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:cba1d078cf2e1e0b8402e6bda528bf8fda7ccd158c3dba6c012b7897747c41a0"},
{file = "debugpy-1.8.7-cp311-cp311-win32.whl", hash = "sha256:171899588bcd412151e593bd40d9907133a7622cd6ecdbdb75f89d1551df13c2"},
{file = "debugpy-1.8.7-cp311-cp311-win_amd64.whl", hash = "sha256:6e1c4ffb0c79f66e89dfd97944f335880f0d50ad29525dc792785384923e2211"},
{file = "debugpy-1.8.7-cp312-cp312-macosx_14_0_universal2.whl", hash = "sha256:4d27d842311353ede0ad572600c62e4bcd74f458ee01ab0dd3a1a4457e7e3706"},
{file = "debugpy-1.8.7-cp312-cp312-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:703c1fd62ae0356e194f3e7b7a92acd931f71fe81c4b3be2c17a7b8a4b546ec2"},
{file = "debugpy-1.8.7-cp312-cp312-win32.whl", hash = "sha256:2f729228430ef191c1e4df72a75ac94e9bf77413ce5f3f900018712c9da0aaca"},
{file = "debugpy-1.8.7-cp312-cp312-win_amd64.whl", hash = "sha256:45c30aaefb3e1975e8a0258f5bbd26cd40cde9bfe71e9e5a7ac82e79bad64e39"},
{file = "debugpy-1.8.7-cp313-cp313-macosx_14_0_universal2.whl", hash = "sha256:d050a1ec7e925f514f0f6594a1e522580317da31fbda1af71d1530d6ea1f2b40"},
{file = "debugpy-1.8.7-cp313-cp313-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f2f4349a28e3228a42958f8ddaa6333d6f8282d5edaea456070e48609c5983b7"},
{file = "debugpy-1.8.7-cp313-cp313-win32.whl", hash = "sha256:11ad72eb9ddb436afb8337891a986302e14944f0f755fd94e90d0d71e9100bba"},
{file = "debugpy-1.8.7-cp313-cp313-win_amd64.whl", hash = "sha256:2efb84d6789352d7950b03d7f866e6d180284bc02c7e12cb37b489b7083d81aa"},
{file = "debugpy-1.8.7-cp38-cp38-macosx_14_0_x86_64.whl", hash = "sha256:4b908291a1d051ef3331484de8e959ef3e66f12b5e610c203b5b75d2725613a7"},
{file = "debugpy-1.8.7-cp38-cp38-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:da8df5b89a41f1fd31503b179d0a84a5fdb752dddd5b5388dbd1ae23cda31ce9"},
{file = "debugpy-1.8.7-cp38-cp38-win32.whl", hash = "sha256:b12515e04720e9e5c2216cc7086d0edadf25d7ab7e3564ec8b4521cf111b4f8c"},
{file = "debugpy-1.8.7-cp38-cp38-win_amd64.whl", hash = "sha256:93176e7672551cb5281577cdb62c63aadc87ec036f0c6a486f0ded337c504596"},
{file = "debugpy-1.8.7-cp39-cp39-macosx_14_0_x86_64.whl", hash = "sha256:90d93e4f2db442f8222dec5ec55ccfc8005821028982f1968ebf551d32b28907"},
{file = "debugpy-1.8.7-cp39-cp39-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:b6db2a370e2700557a976eaadb16243ec9c91bd46f1b3bb15376d7aaa7632c81"},
{file = "debugpy-1.8.7-cp39-cp39-win32.whl", hash = "sha256:a6cf2510740e0c0b4a40330640e4b454f928c7b99b0c9dbf48b11efba08a8cda"},
{file = "debugpy-1.8.7-cp39-cp39-win_amd64.whl", hash = "sha256:6a9d9d6d31846d8e34f52987ee0f1a904c7baa4912bf4843ab39dadf9b8f3e0d"},
{file = "debugpy-1.8.7-py2.py3-none-any.whl", hash = "sha256:57b00de1c8d2c84a61b90880f7e5b6deaf4c312ecbde3a0e8912f2a56c4ac9ae"},
{file = "debugpy-1.8.7.zip", hash = "sha256:18b8f731ed3e2e1df8e9cdaa23fb1fc9c24e570cd0081625308ec51c82efe42e"},
]
[[package]]
@ -2325,13 +2329,13 @@ langchain-core = ">=0.2.38,<0.3.0"
[[package]]
name = "langsmith"
version = "0.1.133"
version = "0.1.134"
description = "Client library to connect to the LangSmith LLM Tracing and Evaluation Platform."
optional = false
python-versions = "<4.0,>=3.8.1"
files = [
{file = "langsmith-0.1.133-py3-none-any.whl", hash = "sha256:82e837a6039c483beadbe19c2ba7ebafbd402d3e8105234f5ef334425cff7b45"},
{file = "langsmith-0.1.133.tar.gz", hash = "sha256:7bfd8bef166b9a64ee540a11bee4aa7bf43b1d9229f95b0fc19086454955185d"},
{file = "langsmith-0.1.134-py3-none-any.whl", hash = "sha256:ada98ad80ef38807725f32441a472da3dd28394010877751f48f458d3289da04"},
{file = "langsmith-0.1.134.tar.gz", hash = "sha256:23abee3b508875a0e63c602afafffc02442a19cfd88f9daae05b3e9054fd6b61"},
]
[package.dependencies]
@ -3037,20 +3041,21 @@ files = [
[[package]]
name = "networkx"
version = "3.3"
version = "3.4"
description = "Python package for creating and manipulating graphs and networks"
optional = false
python-versions = ">=3.10"
files = [
{file = "networkx-3.3-py3-none-any.whl", hash = "sha256:28575580c6ebdaf4505b22c6256a2b9de86b316dc63ba9e93abde3d78dfdbcf2"},
{file = "networkx-3.3.tar.gz", hash = "sha256:0c127d8b2f4865f59ae9cb8aafcd60b5c70f3241ebd66f7defad7c4ab90126c9"},
{file = "networkx-3.4-py3-none-any.whl", hash = "sha256:46dad0ec74a825a968e2b36c37ef5b91faa3868f017b2283d9cbff33112222ce"},
{file = "networkx-3.4.tar.gz", hash = "sha256:1269b90f8f0d3a4095f016f49650f35ac169729f49b69d0572b2bb142748162b"},
]
[package.extras]
default = ["matplotlib (>=3.6)", "numpy (>=1.23)", "pandas (>=1.4)", "scipy (>=1.9,!=1.11.0,!=1.11.1)"]
default = ["matplotlib (>=3.7)", "numpy (>=1.24)", "pandas (>=2.0)", "scipy (>=1.10,!=1.11.0,!=1.11.1)"]
developer = ["changelist (==0.5)", "mypy (>=1.1)", "pre-commit (>=3.2)", "rtoml"]
doc = ["myst-nb (>=1.0)", "numpydoc (>=1.7)", "pillow (>=9.4)", "pydata-sphinx-theme (>=0.14)", "sphinx (>=7)", "sphinx-gallery (>=0.14)", "texext (>=0.6.7)"]
extra = ["lxml (>=4.6)", "pydot (>=2.0)", "pygraphviz (>=1.12)", "sympy (>=1.10)"]
doc = ["intersphinx-registry", "myst-nb (>=1.1)", "numpydoc (>=1.8.0)", "pillow (>=9.4)", "pydata-sphinx-theme (>=0.15)", "sphinx (>=7.3)", "sphinx-gallery (>=0.16)", "texext (>=0.6.7)"]
example = ["cairocffi (>=1.7)", "contextily (>=1.6)", "igraph (>=0.11)", "momepy (>=0.7.2)", "osmnx (>=1.9)", "scikit-learn (>=1.5)", "seaborn (>=0.13)"]
extra = ["lxml (>=4.6)", "pydot (>=3.0.1)", "pygraphviz (>=1.14)", "sympy (>=1.10)"]
test = ["pytest (>=7.2)", "pytest-cov (>=4.0)"]
[[package]]
@ -3757,13 +3762,13 @@ xmp = ["defusedxml"]
[[package]]
name = "pkginfo"
version = "1.11.1"
version = "1.11.2"
description = "Query metadata from sdists / bdists / installed packages."
optional = false
python-versions = ">=3.8"
files = [
{file = "pkginfo-1.11.1-py3-none-any.whl", hash = "sha256:bfa76a714fdfc18a045fcd684dbfc3816b603d9d075febef17cb6582bea29573"},
{file = "pkginfo-1.11.1.tar.gz", hash = "sha256:2e0dca1cf4c8e39644eed32408ea9966ee15e0d324c62ba899a393b3c6b467aa"},
{file = "pkginfo-1.11.2-py3-none-any.whl", hash = "sha256:9ec518eefccd159de7ed45386a6bb4c6ca5fa2cb3bd9b71154fae44f6f1b36a3"},
{file = "pkginfo-1.11.2.tar.gz", hash = "sha256:c6bc916b8298d159e31f2c216e35ee5b86da7da18874f879798d0a1983537c86"},
]
[package.extras]
@ -5561,26 +5566,29 @@ files = [
[[package]]
name = "sentence-transformers"
version = "3.1.1"
version = "3.2.0"
description = "State-of-the-Art Text Embeddings"
optional = false
python-versions = ">=3.8"
files = [
{file = "sentence_transformers-3.1.1-py3-none-any.whl", hash = "sha256:c73bf6f17e3676bb9372a6133a254ebfb5907586b470f2bac5a840c64c3cf97e"},
{file = "sentence_transformers-3.1.1.tar.gz", hash = "sha256:8f00020ef4ad6b918475c38af545c22f61403b67eb22d994860bab06902db160"},
{file = "sentence_transformers-3.2.0-py3-none-any.whl", hash = "sha256:02dbe96d669d30084ea11af94e7c17895c31748e86f20af4dbcc4ea6522b3506"},
{file = "sentence_transformers-3.2.0.tar.gz", hash = "sha256:4da78ba340fffc48d60a25415145cbe665216d374c948ec54c3163bd352bcc27"},
]
[package.dependencies]
huggingface-hub = ">=0.19.3"
huggingface-hub = ">=0.20.0"
Pillow = "*"
scikit-learn = "*"
scipy = "*"
torch = ">=1.11.0"
tqdm = "*"
transformers = ">=4.38.0,<5.0.0"
transformers = ">=4.41.0,<5.0.0"
[package.extras]
dev = ["accelerate (>=0.20.3)", "datasets", "pre-commit", "pytest", "pytest-cov"]
onnx = ["optimum[onnxruntime] (>=1.23.0)"]
onnx-gpu = ["optimum[onnxruntime-gpu] (>=1.23.0)"]
openvino = ["optimum-intel[openvino] (>=1.20.0)"]
train = ["accelerate (>=0.20.3)", "datasets"]
[[package]]

5
pyproject.toml
View File

@ -1,6 +1,6 @@
[tool.poetry]
name = "docling"
version = "2.0.0" # DO NOT EDIT, updated automatically
version = "2.0.0-dev1" # DO NOT EDIT, updated automatically
description = "Docling PDF conversion package"
authors = ["Christoph Auer <cau@zurich.ibm.com>", "Michele Dolfi <dol@zurich.ibm.com>", "Maxim Lysak <mly@zurich.ibm.com>", "Nikos Livathinos <nli@zurich.ibm.com>", "Ahmed Nassar <ahn@zurich.ibm.com>", "Peter Staar <taa@zurich.ibm.com>"]
license = "MIT"
@ -80,6 +80,9 @@ nbqa = "^1.9.0"
[tool.poetry.group.examples.dependencies]
datasets = "^2.21.0"
python-dotenv = "^1.0.1"
# llama-index-readers-docling = { version = "^0.1.0", markers = 'python_version < "3.13"' }
# llama-index-node-parser-docling = { version = "^0.1.0", markers = 'python_version < "3.13"' }
# llama-index-readers-file = { version = "^0.2.2", markers = 'python_version < "3.13"' }
# llama-index-embeddings-huggingface = { version = "^0.3.1", markers = 'python_version < "3.13"' }
# llama-index-llms-huggingface-api = { version = "^0.2.0", markers = 'python_version < "3.13"' }
# llama-index-vector-stores-milvus ={ version = "^0.2.1", markers = 'python_version < "3.13"' }

4
tests/test_e2e_ocr_conversion.py
View File

@ -16,7 +16,7 @@ from docling.document_converter import DocumentConverter, PdfFormatOption
from .verify_utils import verify_conversion_result_v1, verify_conversion_result_v2
GENERATE = True
GENERATE = False
# Debug
@ -107,5 +107,5 @@ def test_e2e_conversions():
input_path=pdf_path,
doc_result=doc_result,
generate=GENERATE,
skip_cells=True,
fuzzy=True,
)

83
tests/verify_utils.py
View File

@ -12,6 +12,42 @@ from docling.datamodel.base_models import ConversionStatus, Page
from docling.datamodel.document import ConversionResult
def levenshtein(str1: str, str2: str) -> int:
# Ensure str1 is the shorter string to optimize memory usage
if len(str1) > len(str2):
str1, str2 = str2, str1
# Previous and current row buffers
previous_row = list(range(len(str2) + 1))
current_row = [0] * (len(str2) + 1)
# Compute the Levenshtein distance row by row
for i, c1 in enumerate(str1, start=1):
current_row[0] = i
for j, c2 in enumerate(str2, start=1):
insertions = previous_row[j] + 1
deletions = current_row[j - 1] + 1
substitutions = previous_row[j - 1] + (c1 != c2)
current_row[j] = min(insertions, deletions, substitutions)
# Swap rows for the next iteration
previous_row, current_row = current_row, previous_row
# The result is in the last element of the previous row
return previous_row[-1]
def verify_text(gt: str, pred: str, fuzzy: bool, fuzzy_threshold: float = 0.4):
if len(gt) == 0 or not fuzzy:
assert gt == pred, f"{gt}!={pred}"
else:
dist = levenshtein(gt, pred)
diff = dist / len(gt)
assert diff < fuzzy_threshold, f"{gt}!~{pred}"
return True
def verify_cells(doc_pred_pages: List[Page], doc_true_pages: List[Page]):
assert len(doc_pred_pages) == len(
@ -33,7 +69,6 @@ def verify_cells(doc_pred_pages: List[Page], doc_true_pages: List[Page]):
true_text = cell_true_item.text
pred_text = cell_pred_item.text
assert true_text == pred_text, f"{true_text}!={pred_text}"
true_bbox = cell_true_item.bbox.as_tuple()
@ -70,7 +105,7 @@ def verify_cells(doc_pred_pages: List[Page], doc_true_pages: List[Page]):
# return True
def verify_tables_v1(doc_pred: DsDocument, doc_true: DsDocument):
def verify_tables_v1(doc_pred: DsDocument, doc_true: DsDocument, fuzzy: bool):
if doc_true.tables is None:
# No tables to check
assert doc_pred.tables is None, "not expecting any table on this document"
@ -103,9 +138,9 @@ def verify_tables_v1(doc_pred: DsDocument, doc_true: DsDocument):
# print("pred: ", pred_item.data[i][j].text)
# print("")
assert (
true_item.data[i][j].text == pred_item.data[i][j].text
), "table-cell does not have the same text"
verify_text(
true_item.data[i][j].text, pred_item.data[i][j].text, fuzzy=fuzzy
)
assert (
true_item.data[i][j].obj_type == pred_item.data[i][j].obj_type
@ -114,7 +149,7 @@ def verify_tables_v1(doc_pred: DsDocument, doc_true: DsDocument):
return True
def verify_tables_v2(doc_pred: DoclingDocument, doc_true: DoclingDocument):
def verify_tables_v2(doc_pred: DoclingDocument, doc_true: DoclingDocument, fuzzy: bool):
if not len(doc_true.tables) > 0:
# No tables to check
assert len(doc_pred.tables) == 0, "not expecting any table on this document"
@ -147,9 +182,11 @@ def verify_tables_v2(doc_pred: DoclingDocument, doc_true: DoclingDocument):
# print("pred: ", pred_item.data[i][j].text)
# print("")
assert (
true_item.data.grid[i][j].text == pred_item.data.grid[i][j].text
), "table-cell does not have the same text"
verify_text(
true_item.data.grid[i][j].text,
pred_item.data.grid[i][j].text,
fuzzy=fuzzy,
)
assert (
true_item.data.grid[i][j].column_header
@ -175,12 +212,12 @@ def verify_tables_v2(doc_pred: DoclingDocument, doc_true: DoclingDocument):
# return True
def verify_md(doc_pred_md, doc_true_md):
return doc_pred_md == doc_true_md
def verify_md(doc_pred_md: str, doc_true_md: str, fuzzy: bool):
return verify_text(doc_true_md, doc_pred_md, fuzzy)
def verify_dt(doc_pred_dt, doc_true_dt):
return doc_pred_dt == doc_true_dt
def verify_dt(doc_pred_dt: str, doc_true_dt: str, fuzzy: bool):
return verify_text(doc_true_dt, doc_pred_dt, fuzzy)
def verify_conversion_result_v1(
@ -188,7 +225,7 @@ def verify_conversion_result_v1(
doc_result: ConversionResult,
generate: bool = False,
ocr_engine: str = None,
skip_cells: bool = False,
fuzzy: bool = False,
):
PageList = TypeAdapter(List[Page])
@ -233,7 +270,7 @@ def verify_conversion_result_v1(
with open(dt_path, "r") as fr:
doc_true_dt = fr.read()
if not skip_cells:
if not fuzzy:
assert verify_cells(
doc_pred_pages, doc_true_pages
), f"Mismatch in PDF cell prediction for {input_path}"
@ -243,15 +280,15 @@ def verify_conversion_result_v1(
# ), f"Mismatch in JSON prediction for {input_path}"
assert verify_tables_v1(
doc_pred, doc_true
doc_pred, doc_true, fuzzy=fuzzy
), f"verify_tables(doc_pred, doc_true) mismatch for {input_path}"
assert verify_md(
doc_pred_md, doc_true_md
doc_pred_md, doc_true_md, fuzzy=fuzzy
), f"Mismatch in Markdown prediction for {input_path}"
assert verify_dt(
doc_pred_dt, doc_true_dt
doc_pred_dt, doc_true_dt, fuzzy=fuzzy
), f"Mismatch in DocTags prediction for {input_path}"
@ -260,7 +297,7 @@ def verify_conversion_result_v2(
doc_result: ConversionResult,
generate: bool = False,
ocr_engine: str = None,
skip_cells: bool = False,
fuzzy: bool = False,
):
PageList = TypeAdapter(List[Page])
@ -305,7 +342,7 @@ def verify_conversion_result_v2(
with open(dt_path, "r") as fr:
doc_true_dt = fr.read()
if not skip_cells:
if not fuzzy:
assert verify_cells(
doc_pred_pages, doc_true_pages
), f"Mismatch in PDF cell prediction for {input_path}"
@ -315,13 +352,13 @@ def verify_conversion_result_v2(
# ), f"Mismatch in JSON prediction for {input_path}"
assert verify_tables_v2(
doc_pred, doc_true
doc_pred, doc_true, fuzzy=fuzzy
), f"verify_tables(doc_pred, doc_true) mismatch for {input_path}"
assert verify_md(
doc_pred_md, doc_true_md
doc_pred_md, doc_true_md, fuzzy=fuzzy
), f"Mismatch in Markdown prediction for {input_path}"
assert verify_dt(
doc_pred_dt, doc_true_dt
doc_pred_dt, doc_true_dt, fuzzy=fuzzy
), f"Mismatch in DocTags prediction for {input_path}"