From b4149c8d87d9625a7e136b63002387b08bcc0db8 Mon Sep 17 00:00:00 2001 From: zoazhyga Date: Sat, 10 Aug 2024 19:28:22 +0200 Subject: [PATCH] [dev] add support for .xml files --- Makefile | 2 +- experiments/llama_index.ipynb | 37 +++ poetry.lock | 217 +++++++++++++++++- private_gpt/__init__.py | 2 +- private_gpt/__main__.py | 5 +- .../components/ingest/ingest_helper.py | 65 ++++-- pyproject.toml | 3 + settings-huglama.yaml | 22 +- settings-mock.yaml | 1 - settings.yaml | 12 +- 10 files changed, 329 insertions(+), 37 deletions(-) create mode 100644 experiments/llama_index.ipynb diff --git a/Makefile b/Makefile index 7b9a8469..e1055e9a 100644 --- a/Makefile +++ b/Makefile @@ -39,7 +39,7 @@ dev-windows: (set PGPT_PROFILES=local & poetry run python -m uvicorn private_gpt.main:app --reload --port 8001) dev: - PYTHONUNBUFFERED=1 PGPT_PROFILES=local poetry run python -m uvicorn private_gpt.main:app --reload --port 8001 + PYTHONUNBUFFERED=1 PGPT_PROFILES=huglama poetry run python -m uvicorn private_gpt.main:app --reload --port 8001 ######################################################################################################################## # Misc diff --git a/experiments/llama_index.ipynb b/experiments/llama_index.ipynb new file mode 100644 index 00000000..54f657bb --- /dev/null +++ b/experiments/llama_index.ipynb @@ -0,0 +1,37 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": null, + "id": "initial_id", + "metadata": { + "collapsed": true + }, + "outputs": [], + "source": [ + "" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 2 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython2", + "version": "2.7.6" + } + }, + "nbformat": 4, + "nbformat_minor": 5 +} diff --git a/poetry.lock b/poetry.lock index 3ce94fd7..0cdec5cf 100644 --- a/poetry.lock +++ b/poetry.lock @@ -2642,6 +2642,21 @@ striprtf = ">=0.0.26,<0.0.27" [package.extras] pymupdf = ["pymupdf (>=1.23.21,<2.0.0)"] +[[package]] +name = "llama-index-readers-smart-pdf-loader" +version = "0.1.4" +description = "llama-index readers smart_pdf_loader integration" +optional = false +python-versions = "<4.0,>=3.8.1" +files = [ + {file = "llama_index_readers_smart_pdf_loader-0.1.4-py3-none-any.whl", hash = "sha256:ba21a0ba8b46cd5a5f1205b4601147e469dccccac57de2902cab44794047f57a"}, + {file = "llama_index_readers_smart_pdf_loader-0.1.4.tar.gz", hash = "sha256:40b135750612b175eea16297749946a82238776b398683fc6b1a605e7208300d"}, +] + +[package.dependencies] +llama-index-core = ">=0.10.1,<0.11.0" +llmsherpa = ">=0.1.4,<0.2.0" + [[package]] name = "llama-index-storage-docstore-postgres" version = "0.1.3" @@ -2765,6 +2780,178 @@ grpcio = ">=1.60.0,<2.0.0" llama-index-core = ">=0.10.1,<0.11.0" qdrant-client = ">=1.7.1" +[[package]] +name = "llmsherpa" +version = "0.1.4" +description = "Strategic APIs to Accelerate LLM Use Cases" +optional = false +python-versions = "*" +files = [ + {file = "llmsherpa-0.1.4-py3-none-any.whl", hash = "sha256:88a651eacf221a997a9a9795903fbb70ddb211cd2e82831507de01284c917667"}, + {file = "llmsherpa-0.1.4.tar.gz", hash = "sha256:18789840ded3d7e37db6a5c668dac2b0a58f20e426beaed3e60554dee0a5ecb0"}, +] + +[package.dependencies] +urllib3 = "*" + +[[package]] +name = "lxml" +version = "5.2.2" +description = "Powerful and Pythonic XML processing library combining libxml2/libxslt with the ElementTree API." +optional = false +python-versions = ">=3.6" +files = [ + {file = "lxml-5.2.2-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:364d03207f3e603922d0d3932ef363d55bbf48e3647395765f9bfcbdf6d23632"}, + {file = "lxml-5.2.2-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:50127c186f191b8917ea2fb8b206fbebe87fd414a6084d15568c27d0a21d60db"}, + {file = "lxml-5.2.2-cp310-cp310-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:74e4f025ef3db1c6da4460dd27c118d8cd136d0391da4e387a15e48e5c975147"}, + {file = "lxml-5.2.2-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:981a06a3076997adf7c743dcd0d7a0415582661e2517c7d961493572e909aa1d"}, + {file = "lxml-5.2.2-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:aef5474d913d3b05e613906ba4090433c515e13ea49c837aca18bde190853dff"}, + {file = "lxml-5.2.2-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:1e275ea572389e41e8b039ac076a46cb87ee6b8542df3fff26f5baab43713bca"}, + {file = "lxml-5.2.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f5b65529bb2f21ac7861a0e94fdbf5dc0daab41497d18223b46ee8515e5ad297"}, + {file = "lxml-5.2.2-cp310-cp310-manylinux_2_28_aarch64.whl", hash = "sha256:bcc98f911f10278d1daf14b87d65325851a1d29153caaf146877ec37031d5f36"}, + {file = "lxml-5.2.2-cp310-cp310-manylinux_2_28_ppc64le.whl", hash = "sha256:b47633251727c8fe279f34025844b3b3a3e40cd1b198356d003aa146258d13a2"}, + {file = "lxml-5.2.2-cp310-cp310-manylinux_2_28_s390x.whl", hash = "sha256:fbc9d316552f9ef7bba39f4edfad4a734d3d6f93341232a9dddadec4f15d425f"}, + {file = "lxml-5.2.2-cp310-cp310-manylinux_2_28_x86_64.whl", hash = "sha256:13e69be35391ce72712184f69000cda04fc89689429179bc4c0ae5f0b7a8c21b"}, + {file = "lxml-5.2.2-cp310-cp310-musllinux_1_1_aarch64.whl", hash = "sha256:3b6a30a9ab040b3f545b697cb3adbf3696c05a3a68aad172e3fd7ca73ab3c835"}, + {file = "lxml-5.2.2-cp310-cp310-musllinux_1_1_ppc64le.whl", hash = "sha256:a233bb68625a85126ac9f1fc66d24337d6e8a0f9207b688eec2e7c880f012ec0"}, + {file = "lxml-5.2.2-cp310-cp310-musllinux_1_1_s390x.whl", hash = "sha256:dfa7c241073d8f2b8e8dbc7803c434f57dbb83ae2a3d7892dd068d99e96efe2c"}, + {file = "lxml-5.2.2-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:1a7aca7964ac4bb07680d5c9d63b9d7028cace3e2d43175cb50bba8c5ad33316"}, + {file = "lxml-5.2.2-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:ae4073a60ab98529ab8a72ebf429f2a8cc612619a8c04e08bed27450d52103c0"}, + {file = "lxml-5.2.2-cp310-cp310-musllinux_1_2_ppc64le.whl", hash = "sha256:ffb2be176fed4457e445fe540617f0252a72a8bc56208fd65a690fdb1f57660b"}, + {file = "lxml-5.2.2-cp310-cp310-musllinux_1_2_s390x.whl", hash = "sha256:e290d79a4107d7d794634ce3e985b9ae4f920380a813717adf61804904dc4393"}, + {file = "lxml-5.2.2-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:96e85aa09274955bb6bd483eaf5b12abadade01010478154b0ec70284c1b1526"}, + {file = "lxml-5.2.2-cp310-cp310-win32.whl", hash = "sha256:f956196ef61369f1685d14dad80611488d8dc1ef00be57c0c5a03064005b0f30"}, + {file = "lxml-5.2.2-cp310-cp310-win_amd64.whl", hash = "sha256:875a3f90d7eb5c5d77e529080d95140eacb3c6d13ad5b616ee8095447b1d22e7"}, + {file = "lxml-5.2.2-cp311-cp311-macosx_10_9_universal2.whl", hash = "sha256:45f9494613160d0405682f9eee781c7e6d1bf45f819654eb249f8f46a2c22545"}, + {file = "lxml-5.2.2-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:b0b3f2df149efb242cee2ffdeb6674b7f30d23c9a7af26595099afaf46ef4e88"}, + {file = "lxml-5.2.2-cp311-cp311-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:d28cb356f119a437cc58a13f8135ab8a4c8ece18159eb9194b0d269ec4e28083"}, + {file = "lxml-5.2.2-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:657a972f46bbefdbba2d4f14413c0d079f9ae243bd68193cb5061b9732fa54c1"}, + {file = "lxml-5.2.2-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:b74b9ea10063efb77a965a8d5f4182806fbf59ed068b3c3fd6f30d2ac7bee734"}, + {file = "lxml-5.2.2-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:07542787f86112d46d07d4f3c4e7c760282011b354d012dc4141cc12a68cef5f"}, + {file = "lxml-5.2.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:303f540ad2dddd35b92415b74b900c749ec2010e703ab3bfd6660979d01fd4ed"}, + {file = "lxml-5.2.2-cp311-cp311-manylinux_2_28_aarch64.whl", hash = "sha256:2eb2227ce1ff998faf0cd7fe85bbf086aa41dfc5af3b1d80867ecfe75fb68df3"}, + {file = "lxml-5.2.2-cp311-cp311-manylinux_2_28_ppc64le.whl", hash = "sha256:1d8a701774dfc42a2f0b8ccdfe7dbc140500d1049e0632a611985d943fcf12df"}, + {file = "lxml-5.2.2-cp311-cp311-manylinux_2_28_s390x.whl", hash = "sha256:56793b7a1a091a7c286b5f4aa1fe4ae5d1446fe742d00cdf2ffb1077865db10d"}, + {file = "lxml-5.2.2-cp311-cp311-manylinux_2_28_x86_64.whl", hash = "sha256:eb00b549b13bd6d884c863554566095bf6fa9c3cecb2e7b399c4bc7904cb33b5"}, + {file = "lxml-5.2.2-cp311-cp311-musllinux_1_1_aarch64.whl", hash = "sha256:1a2569a1f15ae6c8c64108a2cd2b4a858fc1e13d25846be0666fc144715e32ab"}, + {file = "lxml-5.2.2-cp311-cp311-musllinux_1_1_ppc64le.whl", hash = "sha256:8cf85a6e40ff1f37fe0f25719aadf443686b1ac7652593dc53c7ef9b8492b115"}, + {file = "lxml-5.2.2-cp311-cp311-musllinux_1_1_s390x.whl", hash = "sha256:d237ba6664b8e60fd90b8549a149a74fcc675272e0e95539a00522e4ca688b04"}, + {file = "lxml-5.2.2-cp311-cp311-musllinux_1_1_x86_64.whl", hash = "sha256:0b3f5016e00ae7630a4b83d0868fca1e3d494c78a75b1c7252606a3a1c5fc2ad"}, + {file = "lxml-5.2.2-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:23441e2b5339bc54dc949e9e675fa35efe858108404ef9aa92f0456929ef6fe8"}, + {file = "lxml-5.2.2-cp311-cp311-musllinux_1_2_ppc64le.whl", hash = "sha256:2fb0ba3e8566548d6c8e7dd82a8229ff47bd8fb8c2da237607ac8e5a1b8312e5"}, + {file = "lxml-5.2.2-cp311-cp311-musllinux_1_2_s390x.whl", hash = "sha256:79d1fb9252e7e2cfe4de6e9a6610c7cbb99b9708e2c3e29057f487de5a9eaefa"}, + {file = "lxml-5.2.2-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:6dcc3d17eac1df7859ae01202e9bb11ffa8c98949dcbeb1069c8b9a75917e01b"}, + {file = "lxml-5.2.2-cp311-cp311-win32.whl", hash = "sha256:4c30a2f83677876465f44c018830f608fa3c6a8a466eb223535035fbc16f3438"}, + {file = "lxml-5.2.2-cp311-cp311-win_amd64.whl", hash = "sha256:49095a38eb333aaf44c06052fd2ec3b8f23e19747ca7ec6f6c954ffea6dbf7be"}, + {file = "lxml-5.2.2-cp312-cp312-macosx_10_9_universal2.whl", hash = "sha256:7429e7faa1a60cad26ae4227f4dd0459efde239e494c7312624ce228e04f6391"}, + {file = "lxml-5.2.2-cp312-cp312-macosx_10_9_x86_64.whl", hash = "sha256:50ccb5d355961c0f12f6cf24b7187dbabd5433f29e15147a67995474f27d1776"}, + {file = "lxml-5.2.2-cp312-cp312-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:dc911208b18842a3a57266d8e51fc3cfaccee90a5351b92079beed912a7914c2"}, + {file = "lxml-5.2.2-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:33ce9e786753743159799fdf8e92a5da351158c4bfb6f2db0bf31e7892a1feb5"}, + {file = "lxml-5.2.2-cp312-cp312-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:ec87c44f619380878bd49ca109669c9f221d9ae6883a5bcb3616785fa8f94c97"}, + {file = "lxml-5.2.2-cp312-cp312-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:08ea0f606808354eb8f2dfaac095963cb25d9d28e27edcc375d7b30ab01abbf6"}, + {file = "lxml-5.2.2-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:75a9632f1d4f698b2e6e2e1ada40e71f369b15d69baddb8968dcc8e683839b18"}, + {file = "lxml-5.2.2-cp312-cp312-manylinux_2_28_aarch64.whl", hash = "sha256:74da9f97daec6928567b48c90ea2c82a106b2d500f397eeb8941e47d30b1ca85"}, + {file = "lxml-5.2.2-cp312-cp312-manylinux_2_28_ppc64le.whl", hash = "sha256:0969e92af09c5687d769731e3f39ed62427cc72176cebb54b7a9d52cc4fa3b73"}, + {file = "lxml-5.2.2-cp312-cp312-manylinux_2_28_s390x.whl", hash = "sha256:9164361769b6ca7769079f4d426a41df6164879f7f3568be9086e15baca61466"}, + {file = "lxml-5.2.2-cp312-cp312-manylinux_2_28_x86_64.whl", hash = "sha256:d26a618ae1766279f2660aca0081b2220aca6bd1aa06b2cf73f07383faf48927"}, + {file = "lxml-5.2.2-cp312-cp312-musllinux_1_1_aarch64.whl", hash = "sha256:ab67ed772c584b7ef2379797bf14b82df9aa5f7438c5b9a09624dd834c1c1aaf"}, + {file = "lxml-5.2.2-cp312-cp312-musllinux_1_1_ppc64le.whl", hash = "sha256:3d1e35572a56941b32c239774d7e9ad724074d37f90c7a7d499ab98761bd80cf"}, + {file = "lxml-5.2.2-cp312-cp312-musllinux_1_1_s390x.whl", hash = "sha256:8268cbcd48c5375f46e000adb1390572c98879eb4f77910c6053d25cc3ac2c67"}, + {file = "lxml-5.2.2-cp312-cp312-musllinux_1_1_x86_64.whl", hash = "sha256:e282aedd63c639c07c3857097fc0e236f984ceb4089a8b284da1c526491e3f3d"}, + {file = "lxml-5.2.2-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:6dfdc2bfe69e9adf0df4915949c22a25b39d175d599bf98e7ddf620a13678585"}, + {file = "lxml-5.2.2-cp312-cp312-musllinux_1_2_ppc64le.whl", hash = "sha256:4aefd911793b5d2d7a921233a54c90329bf3d4a6817dc465f12ffdfe4fc7b8fe"}, + {file = "lxml-5.2.2-cp312-cp312-musllinux_1_2_s390x.whl", hash = "sha256:8b8df03a9e995b6211dafa63b32f9d405881518ff1ddd775db4e7b98fb545e1c"}, + {file = "lxml-5.2.2-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:f11ae142f3a322d44513de1018b50f474f8f736bc3cd91d969f464b5bfef8836"}, + {file = "lxml-5.2.2-cp312-cp312-win32.whl", hash = "sha256:16a8326e51fcdffc886294c1e70b11ddccec836516a343f9ed0f82aac043c24a"}, + {file = "lxml-5.2.2-cp312-cp312-win_amd64.whl", hash = "sha256:bbc4b80af581e18568ff07f6395c02114d05f4865c2812a1f02f2eaecf0bfd48"}, + {file = "lxml-5.2.2-cp36-cp36m-macosx_10_9_x86_64.whl", hash = "sha256:e3d9d13603410b72787579769469af730c38f2f25505573a5888a94b62b920f8"}, + {file = "lxml-5.2.2-cp36-cp36m-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:38b67afb0a06b8575948641c1d6d68e41b83a3abeae2ca9eed2ac59892b36706"}, + {file = "lxml-5.2.2-cp36-cp36m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:c689d0d5381f56de7bd6966a4541bff6e08bf8d3871bbd89a0c6ab18aa699573"}, + {file = "lxml-5.2.2-cp36-cp36m-manylinux_2_28_x86_64.whl", hash = "sha256:cf2a978c795b54c539f47964ec05e35c05bd045db5ca1e8366988c7f2fe6b3ce"}, + {file = "lxml-5.2.2-cp36-cp36m-manylinux_2_5_x86_64.manylinux1_x86_64.whl", hash = "sha256:739e36ef7412b2bd940f75b278749106e6d025e40027c0b94a17ef7968d55d56"}, + {file = "lxml-5.2.2-cp36-cp36m-musllinux_1_1_x86_64.whl", hash = "sha256:d8bbcd21769594dbba9c37d3c819e2d5847656ca99c747ddb31ac1701d0c0ed9"}, + {file = "lxml-5.2.2-cp36-cp36m-musllinux_1_2_x86_64.whl", hash = "sha256:2304d3c93f2258ccf2cf7a6ba8c761d76ef84948d87bf9664e14d203da2cd264"}, + {file = "lxml-5.2.2-cp36-cp36m-win32.whl", hash = "sha256:02437fb7308386867c8b7b0e5bc4cd4b04548b1c5d089ffb8e7b31009b961dc3"}, + {file = "lxml-5.2.2-cp36-cp36m-win_amd64.whl", hash = "sha256:edcfa83e03370032a489430215c1e7783128808fd3e2e0a3225deee278585196"}, + {file = "lxml-5.2.2-cp37-cp37m-macosx_10_9_x86_64.whl", hash = "sha256:28bf95177400066596cdbcfc933312493799382879da504633d16cf60bba735b"}, + {file = "lxml-5.2.2-cp37-cp37m-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:3a745cc98d504d5bd2c19b10c79c61c7c3df9222629f1b6210c0368177589fb8"}, + {file = "lxml-5.2.2-cp37-cp37m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:1b590b39ef90c6b22ec0be925b211298e810b4856909c8ca60d27ffbca6c12e6"}, + {file = "lxml-5.2.2-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:b336b0416828022bfd5a2e3083e7f5ba54b96242159f83c7e3eebaec752f1716"}, + {file = "lxml-5.2.2-cp37-cp37m-manylinux_2_28_aarch64.whl", hash = "sha256:c2faf60c583af0d135e853c86ac2735ce178f0e338a3c7f9ae8f622fd2eb788c"}, + {file = "lxml-5.2.2-cp37-cp37m-manylinux_2_28_x86_64.whl", hash = "sha256:4bc6cb140a7a0ad1f7bc37e018d0ed690b7b6520ade518285dc3171f7a117905"}, + {file = "lxml-5.2.2-cp37-cp37m-musllinux_1_1_aarch64.whl", hash = "sha256:7ff762670cada8e05b32bf1e4dc50b140790909caa8303cfddc4d702b71ea184"}, + {file = "lxml-5.2.2-cp37-cp37m-musllinux_1_1_x86_64.whl", hash = "sha256:57f0a0bbc9868e10ebe874e9f129d2917750adf008fe7b9c1598c0fbbfdde6a6"}, + {file = "lxml-5.2.2-cp37-cp37m-musllinux_1_2_aarch64.whl", hash = "sha256:a6d2092797b388342c1bc932077ad232f914351932353e2e8706851c870bca1f"}, + {file = "lxml-5.2.2-cp37-cp37m-musllinux_1_2_x86_64.whl", hash = "sha256:60499fe961b21264e17a471ec296dcbf4365fbea611bf9e303ab69db7159ce61"}, + {file = "lxml-5.2.2-cp37-cp37m-win32.whl", hash = "sha256:d9b342c76003c6b9336a80efcc766748a333573abf9350f4094ee46b006ec18f"}, + {file = "lxml-5.2.2-cp37-cp37m-win_amd64.whl", hash = "sha256:b16db2770517b8799c79aa80f4053cd6f8b716f21f8aca962725a9565ce3ee40"}, + {file = "lxml-5.2.2-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:7ed07b3062b055d7a7f9d6557a251cc655eed0b3152b76de619516621c56f5d3"}, + {file = "lxml-5.2.2-cp38-cp38-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:f60fdd125d85bf9c279ffb8e94c78c51b3b6a37711464e1f5f31078b45002421"}, + {file = "lxml-5.2.2-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:8a7e24cb69ee5f32e003f50e016d5fde438010c1022c96738b04fc2423e61706"}, + {file = "lxml-5.2.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:23cfafd56887eaed93d07bc4547abd5e09d837a002b791e9767765492a75883f"}, + {file = "lxml-5.2.2-cp38-cp38-manylinux_2_28_aarch64.whl", hash = "sha256:19b4e485cd07b7d83e3fe3b72132e7df70bfac22b14fe4bf7a23822c3a35bff5"}, + {file = "lxml-5.2.2-cp38-cp38-manylinux_2_28_x86_64.whl", hash = "sha256:7ce7ad8abebe737ad6143d9d3bf94b88b93365ea30a5b81f6877ec9c0dee0a48"}, + {file = "lxml-5.2.2-cp38-cp38-musllinux_1_1_aarch64.whl", hash = "sha256:e49b052b768bb74f58c7dda4e0bdf7b79d43a9204ca584ffe1fb48a6f3c84c66"}, + {file = "lxml-5.2.2-cp38-cp38-musllinux_1_1_x86_64.whl", hash = "sha256:d14a0d029a4e176795cef99c056d58067c06195e0c7e2dbb293bf95c08f772a3"}, + {file = "lxml-5.2.2-cp38-cp38-musllinux_1_2_aarch64.whl", hash = "sha256:be49ad33819d7dcc28a309b86d4ed98e1a65f3075c6acd3cd4fe32103235222b"}, + {file = "lxml-5.2.2-cp38-cp38-musllinux_1_2_x86_64.whl", hash = "sha256:a6d17e0370d2516d5bb9062c7b4cb731cff921fc875644c3d751ad857ba9c5b1"}, + {file = "lxml-5.2.2-cp38-cp38-win32.whl", hash = "sha256:5b8c041b6265e08eac8a724b74b655404070b636a8dd6d7a13c3adc07882ef30"}, + {file = "lxml-5.2.2-cp38-cp38-win_amd64.whl", hash = "sha256:f61efaf4bed1cc0860e567d2ecb2363974d414f7f1f124b1df368bbf183453a6"}, + {file = "lxml-5.2.2-cp39-cp39-macosx_10_9_universal2.whl", hash = "sha256:fb91819461b1b56d06fa4bcf86617fac795f6a99d12239fb0c68dbeba41a0a30"}, + {file = "lxml-5.2.2-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:d4ed0c7cbecde7194cd3228c044e86bf73e30a23505af852857c09c24e77ec5d"}, + {file = "lxml-5.2.2-cp39-cp39-manylinux_2_12_i686.manylinux2010_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:54401c77a63cc7d6dc4b4e173bb484f28a5607f3df71484709fe037c92d4f0ed"}, + {file = "lxml-5.2.2-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:625e3ef310e7fa3a761d48ca7ea1f9d8718a32b1542e727d584d82f4453d5eeb"}, + {file = "lxml-5.2.2-cp39-cp39-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:519895c99c815a1a24a926d5b60627ce5ea48e9f639a5cd328bda0515ea0f10c"}, + {file = "lxml-5.2.2-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:c7079d5eb1c1315a858bbf180000757db8ad904a89476653232db835c3114001"}, + {file = "lxml-5.2.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:343ab62e9ca78094f2306aefed67dcfad61c4683f87eee48ff2fd74902447726"}, + {file = "lxml-5.2.2-cp39-cp39-manylinux_2_28_aarch64.whl", hash = "sha256:cd9e78285da6c9ba2d5c769628f43ef66d96ac3085e59b10ad4f3707980710d3"}, + {file = "lxml-5.2.2-cp39-cp39-manylinux_2_28_ppc64le.whl", hash = "sha256:546cf886f6242dff9ec206331209db9c8e1643ae642dea5fdbecae2453cb50fd"}, + {file = "lxml-5.2.2-cp39-cp39-manylinux_2_28_s390x.whl", hash = "sha256:02f6a8eb6512fdc2fd4ca10a49c341c4e109aa6e9448cc4859af5b949622715a"}, + {file = "lxml-5.2.2-cp39-cp39-manylinux_2_28_x86_64.whl", hash = "sha256:339ee4a4704bc724757cd5dd9dc8cf4d00980f5d3e6e06d5847c1b594ace68ab"}, + {file = "lxml-5.2.2-cp39-cp39-musllinux_1_1_aarch64.whl", hash = "sha256:0a028b61a2e357ace98b1615fc03f76eb517cc028993964fe08ad514b1e8892d"}, + {file = "lxml-5.2.2-cp39-cp39-musllinux_1_1_ppc64le.whl", hash = "sha256:f90e552ecbad426eab352e7b2933091f2be77115bb16f09f78404861c8322981"}, + {file = "lxml-5.2.2-cp39-cp39-musllinux_1_1_s390x.whl", hash = "sha256:d83e2d94b69bf31ead2fa45f0acdef0757fa0458a129734f59f67f3d2eb7ef32"}, + {file = "lxml-5.2.2-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:a02d3c48f9bb1e10c7788d92c0c7db6f2002d024ab6e74d6f45ae33e3d0288a3"}, + {file = "lxml-5.2.2-cp39-cp39-musllinux_1_2_aarch64.whl", hash = "sha256:6d68ce8e7b2075390e8ac1e1d3a99e8b6372c694bbe612632606d1d546794207"}, + {file = "lxml-5.2.2-cp39-cp39-musllinux_1_2_ppc64le.whl", hash = "sha256:453d037e09a5176d92ec0fd282e934ed26d806331a8b70ab431a81e2fbabf56d"}, + {file = "lxml-5.2.2-cp39-cp39-musllinux_1_2_s390x.whl", hash = "sha256:3b019d4ee84b683342af793b56bb35034bd749e4cbdd3d33f7d1107790f8c472"}, + {file = "lxml-5.2.2-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:cb3942960f0beb9f46e2a71a3aca220d1ca32feb5a398656be934320804c0df9"}, + {file = "lxml-5.2.2-cp39-cp39-win32.whl", hash = "sha256:ac6540c9fff6e3813d29d0403ee7a81897f1d8ecc09a8ff84d2eea70ede1cdbf"}, + {file = "lxml-5.2.2-cp39-cp39-win_amd64.whl", hash = "sha256:610b5c77428a50269f38a534057444c249976433f40f53e3b47e68349cca1425"}, + {file = "lxml-5.2.2-pp310-pypy310_pp73-macosx_10_9_x86_64.whl", hash = "sha256:b537bd04d7ccd7c6350cdaaaad911f6312cbd61e6e6045542f781c7f8b2e99d2"}, + {file = "lxml-5.2.2-pp310-pypy310_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:4820c02195d6dfb7b8508ff276752f6b2ff8b64ae5d13ebe02e7667e035000b9"}, + {file = "lxml-5.2.2-pp310-pypy310_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f2a09f6184f17a80897172863a655467da2b11151ec98ba8d7af89f17bf63dae"}, + {file = "lxml-5.2.2-pp310-pypy310_pp73-manylinux_2_28_aarch64.whl", hash = "sha256:76acba4c66c47d27c8365e7c10b3d8016a7da83d3191d053a58382311a8bf4e1"}, + {file = "lxml-5.2.2-pp310-pypy310_pp73-manylinux_2_28_x86_64.whl", hash = "sha256:b128092c927eaf485928cec0c28f6b8bead277e28acf56800e972aa2c2abd7a2"}, + {file = "lxml-5.2.2-pp310-pypy310_pp73-win_amd64.whl", hash = "sha256:ae791f6bd43305aade8c0e22f816b34f3b72b6c820477aab4d18473a37e8090b"}, + {file = "lxml-5.2.2-pp37-pypy37_pp73-macosx_10_9_x86_64.whl", hash = "sha256:a2f6a1bc2460e643785a2cde17293bd7a8f990884b822f7bca47bee0a82fc66b"}, + {file = "lxml-5.2.2-pp37-pypy37_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:8e8d351ff44c1638cb6e980623d517abd9f580d2e53bfcd18d8941c052a5a009"}, + {file = "lxml-5.2.2-pp37-pypy37_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:bec4bd9133420c5c52d562469c754f27c5c9e36ee06abc169612c959bd7dbb07"}, + {file = "lxml-5.2.2-pp37-pypy37_pp73-manylinux_2_28_aarch64.whl", hash = "sha256:55ce6b6d803890bd3cc89975fca9de1dff39729b43b73cb15ddd933b8bc20484"}, + {file = "lxml-5.2.2-pp37-pypy37_pp73-manylinux_2_28_x86_64.whl", hash = "sha256:8ab6a358d1286498d80fe67bd3d69fcbc7d1359b45b41e74c4a26964ca99c3f8"}, + {file = "lxml-5.2.2-pp37-pypy37_pp73-win_amd64.whl", hash = "sha256:06668e39e1f3c065349c51ac27ae430719d7806c026fec462e5693b08b95696b"}, + {file = "lxml-5.2.2-pp38-pypy38_pp73-macosx_10_9_x86_64.whl", hash = "sha256:9cd5323344d8ebb9fb5e96da5de5ad4ebab993bbf51674259dbe9d7a18049525"}, + {file = "lxml-5.2.2-pp38-pypy38_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:89feb82ca055af0fe797a2323ec9043b26bc371365847dbe83c7fd2e2f181c34"}, + {file = "lxml-5.2.2-pp38-pypy38_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:e481bba1e11ba585fb06db666bfc23dbe181dbafc7b25776156120bf12e0d5a6"}, + {file = "lxml-5.2.2-pp38-pypy38_pp73-manylinux_2_28_aarch64.whl", hash = "sha256:9d6c6ea6a11ca0ff9cd0390b885984ed31157c168565702959c25e2191674a14"}, + {file = "lxml-5.2.2-pp38-pypy38_pp73-manylinux_2_28_x86_64.whl", hash = "sha256:3d98de734abee23e61f6b8c2e08a88453ada7d6486dc7cdc82922a03968928db"}, + {file = "lxml-5.2.2-pp38-pypy38_pp73-win_amd64.whl", hash = "sha256:69ab77a1373f1e7563e0fb5a29a8440367dec051da6c7405333699d07444f511"}, + {file = "lxml-5.2.2-pp39-pypy39_pp73-macosx_10_9_x86_64.whl", hash = "sha256:34e17913c431f5ae01d8658dbf792fdc457073dcdfbb31dc0cc6ab256e664a8d"}, + {file = "lxml-5.2.2-pp39-pypy39_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:05f8757b03208c3f50097761be2dea0aba02e94f0dc7023ed73a7bb14ff11eb0"}, + {file = "lxml-5.2.2-pp39-pypy39_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:6a520b4f9974b0a0a6ed73c2154de57cdfd0c8800f4f15ab2b73238ffed0b36e"}, + {file = "lxml-5.2.2-pp39-pypy39_pp73-manylinux_2_28_aarch64.whl", hash = "sha256:5e097646944b66207023bc3c634827de858aebc226d5d4d6d16f0b77566ea182"}, + {file = "lxml-5.2.2-pp39-pypy39_pp73-manylinux_2_28_x86_64.whl", hash = "sha256:b5e4ef22ff25bfd4ede5f8fb30f7b24446345f3e79d9b7455aef2836437bc38a"}, + {file = "lxml-5.2.2-pp39-pypy39_pp73-win_amd64.whl", hash = "sha256:ff69a9a0b4b17d78170c73abe2ab12084bdf1691550c5629ad1fe7849433f324"}, + {file = "lxml-5.2.2.tar.gz", hash = "sha256:bb2dc4898180bea79863d5487e5f9c7c34297414bad54bcd0f0852aee9cfdb87"}, +] + +[package.extras] +cssselect = ["cssselect (>=0.7)"] +html-clean = ["lxml-html-clean"] +html5 = ["html5lib"] +htmlsoup = ["BeautifulSoup4"] +source = ["Cython (>=3.0.10)"] + [[package]] name = "lz4" version = "4.3.3" @@ -4754,6 +4941,23 @@ files = [ [package.extras] dev = ["atomicwrites (==1.4.1)", "attrs (==23.2.0)", "coverage (==7.4.1)", "hatch", "invoke (==2.2.0)", "more-itertools (==10.2.0)", "pbr (==6.0.0)", "pluggy (==1.4.0)", "py (==1.11.0)", "pytest (==8.0.0)", "pytest-cov (==4.1.0)", "pytest-timeout (==2.2.0)", "pyyaml (==6.0.1)", "ruff (==0.2.1)"] +[[package]] +name = "python-pptx" +version = "1.0.2" +description = "Create, read, and update PowerPoint 2007+ (.pptx) files." +optional = false +python-versions = ">=3.8" +files = [ + {file = "python_pptx-1.0.2-py3-none-any.whl", hash = "sha256:160838e0b8565a8b1f67947675886e9fea18aa5e795db7ae531606d68e785cba"}, + {file = "python_pptx-1.0.2.tar.gz", hash = "sha256:479a8af0eaf0f0d76b6f00b0887732874ad2e3188230315290cd1f9dd9cc7095"}, +] + +[package.dependencies] +lxml = ">=3.1.0" +Pillow = ">=3.3.2" +typing-extensions = ">=4.9.0" +XlsxWriter = ">=0.5.7" + [[package]] name = "pytz" version = "2024.1" @@ -6477,6 +6681,17 @@ files = [ {file = "wrapt-1.16.0.tar.gz", hash = "sha256:5f370f952971e7d17c7d1ead40e49f32345a7f7a5373571ef44d800d06b1899d"}, ] +[[package]] +name = "xlsxwriter" +version = "3.2.0" +description = "A Python module for creating Excel XLSX files." +optional = false +python-versions = ">=3.6" +files = [ + {file = "XlsxWriter-3.2.0-py3-none-any.whl", hash = "sha256:ecfd5405b3e0e228219bcaf24c2ca0915e012ca9464a14048021d21a995d490e"}, + {file = "XlsxWriter-3.2.0.tar.gz", hash = "sha256:9977d0c661a72866a61f9f7a809e25ebbb0fb7036baa3b9fe74afcfca6b3cb8c"}, +] + [[package]] name = "yarl" version = "1.9.4" @@ -6733,4 +6948,4 @@ vector-stores-qdrant = ["llama-index-vector-stores-qdrant"] [metadata] lock-version = "2.0" python-versions = ">=3.11,<3.12" -content-hash = "7ec5345f582d22af25859e961a76dddb92efe5c58dab21a2a3ba0403f20e44f5" +content-hash = "205a43fecaf7521b9673728e72c043d1ce878ce3e53ebc48292bd45f64090471" diff --git a/private_gpt/__init__.py b/private_gpt/__init__.py index 360b53a4..6fcfd76a 100644 --- a/private_gpt/__init__.py +++ b/private_gpt/__init__.py @@ -4,7 +4,7 @@ import logging import os # Set to 'DEBUG' to have extensive logging turned on, even for libraries -ROOT_LOG_LEVEL = "INFO" +ROOT_LOG_LEVEL = "DEBUG" PRETTY_LOG_FORMAT = ( "%(asctime)s.%(msecs)03d [%(levelname)-8s] %(name)+25s - %(message)s" diff --git a/private_gpt/__main__.py b/private_gpt/__main__.py index 18b42fd7..818aa3b2 100644 --- a/private_gpt/__main__.py +++ b/private_gpt/__main__.py @@ -8,4 +8,7 @@ from private_gpt.settings.settings import settings # Set log_config=None to do not use the uvicorn logging configuration, and # use ours instead. For reference, see below: # https://github.com/tiangolo/fastapi/discussions/7457#discussioncomment-5141108 -uvicorn.run(app, host="0.0.0.0", port=settings().server.port, log_config=None) + + +if __name__ == "__main__": + uvicorn.run(app, host="0.0.0.0", port=settings().server.port, log_config=None) diff --git a/private_gpt/components/ingest/ingest_helper.py b/private_gpt/components/ingest/ingest_helper.py index a1109070..5ac8ce23 100644 --- a/private_gpt/components/ingest/ingest_helper.py +++ b/private_gpt/components/ingest/ingest_helper.py @@ -10,7 +10,23 @@ logger = logging.getLogger(__name__) # Inspired by the `llama_index.core.readers.file.base` module -def _try_loading_included_file_formats() -> dict[str, type[BaseReader]]: +def _try_loading_included_file_formats( + llmsherpa_api_url: str = None, +) -> dict[str, type[BaseReader]]: + simple_pdf_extractor = None + if llmsherpa_api_url is not None: + try: + from llama_index.readers.smart_pdf_loader import SmartPDFLoader + + # llmsherpa_api_url = "http://localhost:5010/api/parseDocument?renderFormat=all&useNewIndentParser=yes" + simple_pdf_extractor = SmartPDFLoader( + llmsherpa_api_url=llmsherpa_api_url, + ) + except ImportError as e: + raise ImportError( + "`llama-index-readers-smart-pdf-loader` package not found" + ) from e + try: from llama_index.readers.file.docs import ( # type: ignore DocxReader, @@ -31,31 +47,36 @@ def _try_loading_included_file_formats() -> dict[str, type[BaseReader]]: raise ImportError("`llama-index-readers-file` package not found") from e default_file_reader_cls: dict[str, type[BaseReader]] = { - ".hwp": HWPReader, - ".pdf": PDFReader, - ".docx": DocxReader, - ".pptx": PptxReader, - ".ppt": PptxReader, - ".pptm": PptxReader, - ".jpg": ImageReader, - ".png": ImageReader, - ".jpeg": ImageReader, - ".mp3": VideoAudioReader, - ".mp4": VideoAudioReader, - ".csv": PandasCSVReader, - ".epub": EpubReader, - ".md": MarkdownReader, - ".mbox": MboxReader, - ".ipynb": IPYNBReader, + ".hwp": HWPReader(), + # ".pdf": simple_pdf_extractor if simple_pdf_extractor else PDFReader, + ".pdf": PDFReader(), + ".docx": simple_pdf_extractor if simple_pdf_extractor else DocxReader(), + ".pptx": PptxReader(), + ".ppt": PptxReader(), + ".pptm": PptxReader(), + ".jpg": ImageReader(), + ".png": ImageReader(), + ".jpeg": ImageReader(), + # ".mp3": VideoAudioReader(), + # ".mp4": VideoAudioReader(), + ".csv": simple_pdf_extractor if simple_pdf_extractor else PandasCSVReader(), + ".xls": simple_pdf_extractor if simple_pdf_extractor else None, + ".xlsx": simple_pdf_extractor if simple_pdf_extractor else None, + ".epub": EpubReader(), + ".md": MarkdownReader(), + ".mbox": MboxReader(), + ".ipynb": IPYNBReader(), } return default_file_reader_cls # Patching the default file reader to support other file types -FILE_READER_CLS = _try_loading_included_file_formats() +FILE_READER_CLS = _try_loading_included_file_formats( + "http://localhost:5010/api/parseDocument?renderFormat=all&useNewIndentParser=yes" +) FILE_READER_CLS.update( { - ".json": JSONReader, + ".json": JSONReader(), } ) @@ -91,8 +112,10 @@ class IngestionHelper: string_reader = StringIterableReader() return string_reader.load_data([file_data.read_text()]) - logger.debug("Specific reader found for extension=%s", extension) - return reader_cls().load_data(file_data) + logger.debug( + f"Specific reader found for extension=%s, {reader_cls=}", extension + ) + return reader_cls.load_data(file_data.as_posix()) @staticmethod def _exclude_metadata(documents: list[Document]) -> None: diff --git a/pyproject.toml b/pyproject.toml index a27aea80..24880cfe 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -67,6 +67,9 @@ ollama = {version ="^0.3.0", optional = true} # Optional HF Transformers einops = {version = "^0.8.0", optional = true} +llama-index-readers-smart-pdf-loader = "^0.1.4" +python-pptx = "^1.0.2" +pillow = "^10.4.0" [tool.poetry.extras] ui = ["gradio", "ffmpy"] diff --git a/settings-huglama.yaml b/settings-huglama.yaml index 40a4c7db..4c26ea59 100644 --- a/settings-huglama.yaml +++ b/settings-huglama.yaml @@ -1,18 +1,30 @@ server: env_name: ${APP_ENV:huglama} +data: + local_data_folder: local_data/private_gpt + local_ingestion: + enabled: ${LOCAL_INGESTION_ENABLED:true} + allow_ingest_from: ["local_data/input_raw"] + llm: mode: ollama max_new_tokens: 10000 - context_window: 100000 + context_window: 12800 temperature: 0.1 embedding: mode: huggingface - ingest_mode: parallel + ingest_mode: simple +# count_workers: 2 vectorstore: - database: qdrant + database: postgres -qdrant: - path: local_data/private_gpt/qdrant +postgres: + host: localhost + port: 5432 + database: vector_db + user: user + password: pass31415 + schema_name: private_gpt diff --git a/settings-mock.yaml b/settings-mock.yaml index 3c63f7fe..c13639fe 100644 --- a/settings-mock.yaml +++ b/settings-mock.yaml @@ -1,7 +1,6 @@ server: env_name: ${APP_ENV:mock} -# This configuration allows you to use GPU for creating embeddings while avoiding loading LLM into vRAM llm: mode: mock diff --git a/settings.yaml b/settings.yaml index cc823250..2c1d281e 100644 --- a/settings.yaml +++ b/settings.yaml @@ -18,7 +18,7 @@ server: data: local_ingestion: - enabled: ${LOCAL_INGESTION_ENABLED:false} + enabled: ${LOCAL_INGESTION_ENABLED:true} allow_ingest_from: ["*"] local_data_folder: local_data/private_gpt @@ -50,16 +50,16 @@ llm: mode: llamacpp prompt_style: "llama3" # Should be matching the selected model - max_new_tokens: 512 - context_window: 3900 + max_new_tokens: 10000 + context_window: 100000 # Select your tokenizer. Llama-index tokenizer is the default. - # tokenizer: meta-llama/Meta-Llama-3.1-8B-Instruct +# tokenizer: meta-llama/Meta-Llama-3.1-8B-Instruct temperature: 0.1 # The temperature of the model. Increasing the temperature will make the model answer more creatively. A value of 0.1 would be more factual. (Default: 0.1) rag: - similarity_top_k: 20 + similarity_top_k: 30 #This value controls how many "top" documents the RAG returns to use in the context. - #similarity_value: 0.45 + similarity_value: 0.45 #This value is disabled by default. If you enable this settings, the RAG will only use articles that meet a certain percentage score. rerank: enabled: false