Spaces:
Sleeping
Sleeping
File size: 48,366 Bytes
bf6d237 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 |
{
"openapi": "3.1.0",
"info": {
"title": "PrivateGPT",
"summary": "PrivateGPT is a production-ready AI project that allows you to ask questions to your documents using the power of Large Language Models (LLMs), even in scenarios without Internet connection. 100% private, no data leaves your execution environment at any point.",
"description": "## Introduction\n\nPrivateGPT provides an **API** containing all the building blocks required to build\n**private, context-aware AI applications**. The API follows and extends OpenAI API standard, and supports\nboth normal and streaming responses.\n\nThe API is divided in two logical blocks:\n\n- High-level API, abstracting all the complexity of a RAG (Retrieval Augmented Generation) pipeline implementation:\n - Ingestion of documents: internally managing document parsing, splitting, metadata extraction,\n embedding generation and storage.\n - Chat & Completions using context from ingested documents: abstracting the retrieval of context, the prompt\n engineering and the response generation.\n- Low-level API, allowing advanced users to implement their own complex pipelines:\n - Embeddings generation: based on a piece of text.\n - Contextual chunks retrieval: given a query, returns the most relevant chunks of text from the ingested\n documents.\n\n> A working **Gradio UI client** is provided to test the API, together with a set of\n> useful tools such as bulk model download script, ingestion script, documents folder\n> watch, etc.\n\n## Quick Local Installation steps\n\nThe steps in `Installation and Settings` section are better explained and cover more\nsetup scenarios. But if you are looking for a quick setup guide, here it is:\n\n```\n# Clone the repo\ngit clone https://github.com/imartinez/privateGPT\ncd privateGPT\n\n# Install Python 3.11\npyenv install 3.11\npyenv local 3.11\n\n# Install dependencies\npoetry install --with ui,local\n\n# Download Embedding and LLM models\npoetry run python scripts/setup\n\n# (Optional) For Mac with Metal GPU, enable it. Check Installation and Settings section \nto know how to enable GPU on other platforms\nCMAKE_ARGS=\"-DLLAMA_METAL=on\" pip install --force-reinstall --no-cache-dir llama-cpp-python\n\n# Run the local server \nPGPT_PROFILES=local make run\n\n# Note: on Mac with Metal you should see a ggml_metal_add_buffer log, stating GPU is \nbeing used\n\n# Navigate to the UI and try it out! \nhttp://localhost:8001/\n```\n\n## Installation and Settings\n\n### Base requirements to run PrivateGPT\n\n* Git clone PrivateGPT repository, and navigate to it:\n\n```\n git clone https://github.com/imartinez/privateGPT\n cd privateGPT\n```\n\n* Install Python 3.11. Ideally through a python version manager like `pyenv`.\n Python 3.12\n should work too. Earlier python versions are not supported.\n * osx/linux: [pyenv](https://github.com/pyenv/pyenv)\n * windows: [pyenv-win](https://github.com/pyenv-win/pyenv-win)\n\n``` \npyenv install 3.11\npyenv local 3.11\n```\n\n* Install [Poetry](https://python-poetry.org/docs/#installing-with-the-official-installer) for dependency management:\n\n* Have a valid C++ compiler like gcc. See [Troubleshooting: C++ Compiler](#troubleshooting-c-compiler) for more details.\n\n* Install `make` for scripts:\n * osx: (Using homebrew): `brew install make`\n * windows: (Using chocolatey) `choco install make`\n\n### Install dependencies\n\nInstall the dependencies:\n\n```bash\npoetry install --with ui\n```\n\nVerify everything is working by running `make run` (or `poetry run python -m private_gpt`) and navigate to\nhttp://localhost:8001. You should see a [Gradio UI](https://gradio.app/) **configured with a mock LLM** that will\necho back the input. Later we'll see how to configure a real LLM.\n\n### Settings\n\n> Note: the default settings of PrivateGPT work out-of-the-box for a 100% local setup. Skip this section if you just\n> want to test PrivateGPT locally, and come back later to learn about more configuration options.\n\nPrivateGPT is configured through *profiles* that are defined using yaml files, and selected through env variables.\nThe full list of properties configurable can be found in `settings.yaml`\n\n#### env var `PGPT_SETTINGS_FOLDER`\n\nThe location of the settings folder. Defaults to the root of the project.\nShould contain the default `settings.yaml` and any other `settings-{profile}.yaml`.\n\n#### env var `PGPT_PROFILES`\n\nBy default, the profile definition in `settings.yaml` is loaded.\nUsing this env var you can load additional profiles; format is a comma separated list of profile names.\nThis will merge `settings-{profile}.yaml` on top of the base settings file.\n\nFor example:\n`PGPT_PROFILES=local,cuda` will load `settings-local.yaml`\nand `settings-cuda.yaml`, their contents will be merged with\nlater profiles properties overriding values of earlier ones like `settings.yaml`.\n\nDuring testing, the `test` profile will be active along with the default, therefore `settings-test.yaml`\nfile is required.\n\n#### Environment variables expansion\n\nConfiguration files can contain environment variables,\nthey will be expanded at runtime.\n\nExpansion must follow the pattern `${VARIABLE_NAME:default_value}`.\n\nFor example, the following configuration will use the value of the `PORT`\nenvironment variable or `8001` if it's not set.\nMissing variables with no default will produce an error.\n\n```yaml\nserver:\n port: ${PORT:8001}\n```\n\n### Local LLM requirements\n\nInstall extra dependencies for local execution:\n\n```bash\npoetry install --with local\n```\n\nFor PrivateGPT to run fully locally GPU acceleration is required\n(CPU execution is possible, but very slow), however,\ntypical Macbook laptops or window desktops with mid-range GPUs lack VRAM to run\neven the smallest LLMs. For that reason\n**local execution is only supported for models compatible with [llama.cpp](https://github.com/ggerganov/llama.cpp)**\n\nThese two models are known to work well:\n\n* https://huggingface.co/TheBloke/Llama-2-7B-chat-GGUF\n* https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.1-GGUF (recommended)\n\nTo ease the installation process, use the `setup` script that will download both\nthe embedding and the LLM model and place them in the correct location (under `models` folder):\n\n```bash\npoetry run python scripts/setup\n```\n\nIf you are ok with CPU execution, you can skip the rest of this section.\n\nAs stated before, llama.cpp is required and in\nparticular [llama-cpp-python](https://github.com/abetlen/llama-cpp-python)\nis used.\n\n> It's highly encouraged that you fully read llama-cpp and llama-cpp-python documentation relevant to your platform.\n> Running into installation issues is very likely, and you'll need to troubleshoot them yourself.\n\n#### Customizing low level parameters\n\nCurrently not all the parameters of llama-cpp and llama-cpp-python are available at PrivateGPT's `settings.yaml` file. In case you need to customize parameters such as the number of layers loaded into the GPU, you might change these at the `llm_component.py` file under the `private_gpt/components/llm/llm_component.py`. If you are getting an out of memory error, you might also try a smaller model or stick to the proposed recommended models, instead of custom tuning the parameters.\n\n#### OSX GPU support\n\nYou will need to build [llama.cpp](https://github.com/ggerganov/llama.cpp) with\nmetal support. To do that run:\n\n```bash\nCMAKE_ARGS=\"-DLLAMA_METAL=on\" pip install --force-reinstall --no-cache-dir llama-cpp-python\n```\n\n#### Windows NVIDIA GPU support\n\nWindows GPU support is done through CUDA.\nFollow the instructions on the original [llama.cpp](https://github.com/ggerganov/llama.cpp) repo to install the required\ndependencies.\n\nSome tips to get it working with an NVIDIA card and CUDA (Tested on Windows 10 with CUDA 11.5 RTX 3070):\n\n* Install latest VS2022 (and build tools) https://visualstudio.microsoft.com/vs/community/\n* Install CUDA toolkit https://developer.nvidia.com/cuda-downloads\n* Verify your installation is correct by running `nvcc --version` and `nvidia-smi`, ensure your CUDA version is up to\n date and your GPU is detected.\n* [Optional] Install CMake to troubleshoot building issues by compiling llama.cpp directly https://cmake.org/download/\n\nIf you have all required dependencies properly configured running the\nfollowing powershell command should succeed.\n\n```powershell\n$env:CMAKE_ARGS='-DLLAMA_CUBLAS=on'; poetry run pip install --force-reinstall --no-cache-dir llama-cpp-python\n```\n\nIf your installation was correct, you should see a message similar to the following next\ntime you start the server `BLAS = 1`.\n\n```\nllama_new_context_with_model: total VRAM used: 4857.93 MB (model: 4095.05 MB, context: 762.87 MB)\nAVX = 1 | AVX2 = 1 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | FMA = 1 | NEON = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 0 | VSX = 0 | \n```\n\nNote that llama.cpp offloads matrix calculations to the GPU but the performance is\nstill hit heavily due to latency between CPU and GPU communication. You might need to tweak\nbatch sizes and other parameters to get the best performance for your particular system.\n\n#### Linux NVIDIA GPU support and Windows-WSL\n\nLinux GPU support is done through CUDA.\nFollow the instructions on the original [llama.cpp](https://github.com/ggerganov/llama.cpp) repo to install the required\nexternal\ndependencies.\n\nSome tips:\n\n* Make sure you have an up-to-date C++ compiler\n* Install CUDA toolkit https://developer.nvidia.com/cuda-downloads\n* Verify your installation is correct by running `nvcc --version` and `nvidia-smi`, ensure your CUDA version is up to\n date and your GPU is detected.\n\nAfter that running the following command in the repository will install llama.cpp with GPU support:\n\n`\nCMAKE_ARGS='-DLLAMA_CUBLAS=on' poetry run pip install --force-reinstall --no-cache-dir llama-cpp-python\n`\n\nIf your installation was correct, you should see a message similar to the following next\ntime you start the server `BLAS = 1`.\n\n```\nllama_new_context_with_model: total VRAM used: 4857.93 MB (model: 4095.05 MB, context: 762.87 MB)\nAVX = 1 | AVX2 = 1 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | FMA = 1 | NEON = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 0 | VSX = 0 | \n```\n\n#### Vectorstores\nPrivateGPT supports [Chroma](https://www.trychroma.com/), [Qdrant](https://qdrant.tech/) as vectorstore providers. Chroma being the default.\n\nTo enable Qdrant, set the `vectorstore.database` property in the `settings.yaml` file to `qdrant` and install the `qdrant` extra.\n\n```bash\npoetry install --extras qdrant\n```\n\nBy default Qdrant tries to connect to an instance at `http://localhost:3000`.\n\nQdrant settings can be configured by setting values to the `qdrant` propery in the `settings.yaml` file.\n\nThe available configuration options are:\n| Field | Description |\n|--------------|-------------|\n| location | If `:memory:` - use in-memory Qdrant instance.<br>If `str` - use it as a `url` parameter.|\n| url | Either host or str of 'Optional[scheme], host, Optional[port], Optional[prefix]'.<br> Eg. `http://localhost:6333` |\n| port | Port of the REST API interface. Default: `6333` |\n| grpc_port | Port of the gRPC interface. Default: `6334` |\n| prefer_grpc | If `true` - use gRPC interface whenever possible in custom methods. |\n| https | If `true` - use HTTPS(SSL) protocol.|\n| api_key | API key for authentication in Qdrant Cloud.|\n| prefix | If set, add `prefix` to the REST URL path.<br>Example: `service/v1` will result in `http://localhost:6333/service/v1/{qdrant-endpoint}` for REST API.|\n| timeout | Timeout for REST and gRPC API requests.<br>Default: 5.0 seconds for REST and unlimited for gRPC |\n| host | Host name of Qdrant service. If url and host are not set, defaults to 'localhost'.|\n| path | Persistence path for QdrantLocal. Eg. `local_data/private_gpt/qdrant`|\n| force_disable_check_same_thread | Force disable check_same_thread for QdrantLocal sqlite connection.|\n\n#### Known issues and Troubleshooting\n\nExecution of LLMs locally still has a lot of sharp edges, specially when running on non Linux platforms.\nYou might encounter several issues:\n\n* Performance: RAM or VRAM usage is very high, your computer might experience slowdowns or even crashes.\n* GPU Virtualization on Windows and OSX: Simply not possible with docker desktop, you have to run the server directly on\n the host.\n* Building errors: Some of PrivateGPT dependencies need to build native code, and they might fail on some platforms.\n Most likely you are missing some dev tools in your machine (updated C++ compiler, CUDA is not on PATH, etc.).\n If you encounter any of these issues, please open an issue and we'll try to help.\n\n#### Troubleshooting: C++ Compiler\n\nIf you encounter an error while building a wheel during the `pip install` process, you may need to install a C++\ncompiler on your computer.\n\n**For Windows 10/11**\n\nTo install a C++ compiler on Windows 10/11, follow these steps:\n\n1. Install Visual Studio 2022.\n2. Make sure the following components are selected:\n * Universal Windows Platform development\n * C++ CMake tools for Windows\n3. Download the MinGW installer from the [MinGW website](https://sourceforge.net/projects/mingw/).\n4. Run the installer and select the `gcc` component.\n\n** For OSX **\n\n1. Check if you have a C++ compiler installed, Xcode might have done it for you. for example running `gcc`.\n2. If not, you can install clang or gcc with homebrew `brew install gcc`\n\n#### Troubleshooting: Mac Running Intel\n\nWhen running a Mac with Intel hardware (not M1), you may run into _clang: error: the clang compiler does not support '\n-march=native'_ during pip install.\n\nIf so set your archflags during pip install. eg: _ARCHFLAGS=\"-arch x86_64\" pip3 install -r requirements.txt_\n\n## Running the Server\n\nAfter following the installation steps you should be ready to go. Here are some common run setups:\n\n### Running 100% locally\n\nMake sure you have followed the *Local LLM requirements* section before moving on.\n\nThis command will start PrivateGPT using the `settings.yaml` (default profile) together with the `settings-local.yaml`\nconfiguration files. By default, it will enable both the API and the Gradio UI. Run:\n\n```\nPGPT_PROFILES=local make run\n``` \n\nor\n\n```\nPGPT_PROFILES=local poetry run python -m private_gpt\n```\n\nWhen the server is started it will print a log *Application startup complete*.\nNavigate to http://localhost:8001 to use the Gradio UI or to http://localhost:8001/docs (API section) to try the API\nusing Swagger UI.\n\n### Local server using OpenAI as LLM\n\nIf you cannot run a local model (because you don't have a GPU, for example) or for testing purposes, you may\ndecide to run PrivateGPT using OpenAI as the LLM.\n\nIn order to do so, create a profile `settings-openai.yaml` with the following contents:\n\n```yaml\nllm:\n mode: openai\n\nopenai:\n api_key: <your_openai_api_key> # You could skip this configuration and use the OPENAI_API_KEY env var instead\n```\n\nAnd run PrivateGPT loading that profile you just created:\n\n```PGPT_PROFILES=openai make run```\n\nor\n\n```PGPT_PROFILES=openai poetry run python -m private_gpt```\n\n> Note this will still use the local Embeddings model, as it is ok to use it on a CPU.\n> We'll support using OpenAI embeddings in a future release.\n\nWhen the server is started it will print a log *Application startup complete*.\nNavigate to http://localhost:8001 to use the Gradio UI or to http://localhost:8001/docs (API section) to try the API.\nYou'll notice the speed and quality of response is higher, given you are using OpenAI's servers for the heavy\ncomputations.\n\n### Use AWS's Sagemaker\n\n\ud83d\udea7 Under construction \ud83d\udea7\n\n## Gradio UI user manual\n\nGradio UI is a ready to use way of testing most of PrivateGPT API functionalities.\n\n\n\n### Execution Modes\n\nIt has 3 modes of execution (you can select in the top-left):\n\n* Query Docs: uses the context from the\n ingested documents to answer the questions posted in the chat. It also takes\n into account previous chat messages as context.\n * Makes use of `/chat/completions` API with `use_context=true` and no\n `context_filter`.\n* Search in Docs: fast search that returns the 4 most related text\n chunks, together with their source document and page.\n * Makes use of `/chunks` API with no `context_filter`, `limit=4` and\n `prev_next_chunks=0`.\n* LLM Chat: simple, non-contextual chat with the LLM. The ingested documents won't\n be taken into account, only the previous messages.\n * Makes use of `/chat/completions` API with `use_context=false`.\n\n### Document Ingestion\n\nIngest documents by using the `Upload a File` button. You can check the progress of\nthe ingestion in the console logs of the server.\n\nThe list of ingested files is shown below the button.\n\nIf you want to delete the ingested documents, refer to *Reset Local documents\ndatabase* section in the documentation.\n\n### Chat\n\nNormal chat interface, self-explanatory ;)\n\nYou can check the actual prompt being passed to the LLM by looking at the logs of\nthe server. We'll add better observability in future releases.\n\n## Deployment options\n\n\ud83d\udea7 We are working on Dockerized deployment guidelines \ud83d\udea7\n\n## Observability\n\nBasic logs are enabled using LlamaIndex\nbasic logging (for example ingestion progress or LLM prompts and answers).\n\n\ud83d\udea7 We are working on improved Observability. \ud83d\udea7\n\n## Ingesting & Managing Documents\n\n\ud83d\udea7 Document Update and Delete are still WIP. \ud83d\udea7\n\nThe ingestion of documents can be done in different ways:\n\n* Using the `/ingest` API\n* Using the Gradio UI\n* Using the Bulk Local Ingestion functionality (check next section)\n\n### Bulk Local Ingestion\n\nWhen you are running PrivateGPT in a fully local setup, you can ingest a complete folder for convenience (containing\npdf, text files, etc.)\nand optionally watch changes on it with the command:\n\n```bash\nmake ingest /path/to/folder -- --watch\n```\n\nTo log the processed and failed files to an additional file, use:\n\n```bash\nmake ingest /path/to/folder -- --watch --log-file /path/to/log/file.log\n```\n\nAfter ingestion is complete, you should be able to chat with your documents\nby navigating to http://localhost:8001 and using the option `Query documents`,\nor using the completions / chat API.\n\n### Reset Local documents database\n\nWhen running in a local setup, you can remove all ingested documents by simply\ndeleting all contents of `local_data` folder (except .gitignore).\n\n## API\n\nAs explained in the introduction, the API contains high level APIs (ingestion and chat/completions) and low level APIs\n(embeddings and chunk retrieval). In this section the different specific API calls are explained.\n",
"contact": {
"url": "https://github.com/imartinez/privateGPT"
},
"license": {
"name": "Apache 2.0",
"url": "https://www.apache.org/licenses/LICENSE-2.0.html"
},
"version": "0.1.0",
"x-logo": {
"url": "https://lh3.googleusercontent.com/drive-viewer/AK7aPaD_iNlMoTquOBsw4boh4tIYxyEuhz6EtEs8nzq3yNkNAK00xGjE1KUCmPJSk3TYOjcs6tReG6w_cLu1S7L_gPgT9z52iw=s2560"
}
},
"paths": {
"/v1/completions": {
"post": {
"tags": [
"Contextual Completions"
],
"summary": "Completion",
"description": "We recommend most users use our Chat completions API.\n\nGiven a prompt, the model will return one predicted completion. If `use_context`\nis set to `true`, the model will use context coming from the ingested documents\nto create the response. The documents being used can be filtered using the\n`context_filter` and passing the document IDs to be used. Ingested documents IDs\ncan be found using `/ingest/list` endpoint. If you want all ingested documents to\nbe used, remove `context_filter` altogether.\n\nWhen using `'include_sources': true`, the API will return the source Chunks used\nto create the response, which come from the context provided.\n\nWhen using `'stream': true`, the API will return data chunks following [OpenAI's\nstreaming model](https://platform.openai.com/docs/api-reference/chat/streaming):\n```\n{\"id\":\"12345\",\"object\":\"completion.chunk\",\"created\":1694268190,\n\"model\":\"private-gpt\",\"choices\":[{\"index\":0,\"delta\":{\"content\":\"Hello\"},\n\"finish_reason\":null}]}\n```",
"operationId": "prompt_completion_v1_completions_post",
"requestBody": {
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/CompletionsBody"
}
}
},
"required": true
},
"responses": {
"200": {
"description": "Successful Response",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/OpenAICompletion"
}
}
}
},
"422": {
"description": "Validation Error",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/HTTPValidationError"
}
}
}
}
}
}
},
"/v1/chat/completions": {
"post": {
"tags": [
"Contextual Completions"
],
"summary": "Chat Completion",
"description": "Given a list of messages comprising a conversation, return a response.\n\nIf `use_context` is set to `true`, the model will use context coming\nfrom the ingested documents to create the response. The documents being used can\nbe filtered using the `context_filter` and passing the document IDs to be used.\nIngested documents IDs can be found using `/ingest/list` endpoint. If you want\nall ingested documents to be used, remove `context_filter` altogether.\n\nWhen using `'include_sources': true`, the API will return the source Chunks used\nto create the response, which come from the context provided.\n\nWhen using `'stream': true`, the API will return data chunks following [OpenAI's\nstreaming model](https://platform.openai.com/docs/api-reference/chat/streaming):\n```\n{\"id\":\"12345\",\"object\":\"completion.chunk\",\"created\":1694268190,\n\"model\":\"private-gpt\",\"choices\":[{\"index\":0,\"delta\":{\"content\":\"Hello\"},\n\"finish_reason\":null}]}\n```",
"operationId": "chat_completion_v1_chat_completions_post",
"requestBody": {
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/ChatBody"
}
}
},
"required": true
},
"responses": {
"200": {
"description": "Successful Response",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/OpenAICompletion"
}
}
}
},
"422": {
"description": "Validation Error",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/HTTPValidationError"
}
}
}
}
}
}
},
"/v1/chunks": {
"post": {
"tags": [
"Context Chunks"
],
"summary": "Chunks Retrieval",
"description": "Given a `text`, returns the most relevant chunks from the ingested documents.\n\nThe returned information can be used to generate prompts that can be\npassed to `/completions` or `/chat/completions` APIs. Note: it is usually a very\nfast API, because only the Embeddings model is involved, not the LLM. The\nreturned information contains the relevant chunk `text` together with the source\n`document` it is coming from. It also contains a score that can be used to\ncompare different results.\n\nThe max number of chunks to be returned is set using the `limit` param.\n\nPrevious and next chunks (pieces of text that appear right before or after in the\ndocument) can be fetched by using the `prev_next_chunks` field.\n\nThe documents being used can be filtered using the `context_filter` and passing\nthe document IDs to be used. Ingested documents IDs can be found using\n`/ingest/list` endpoint. If you want all ingested documents to be used,\nremove `context_filter` altogether.",
"operationId": "chunks_retrieval_v1_chunks_post",
"requestBody": {
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/ChunksBody"
}
}
},
"required": true
},
"responses": {
"200": {
"description": "Successful Response",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/ChunksResponse"
}
}
}
},
"422": {
"description": "Validation Error",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/HTTPValidationError"
}
}
}
}
}
}
},
"/v1/ingest": {
"post": {
"tags": [
"Ingestion"
],
"summary": "Ingest",
"description": "Ingests and processes a file, storing its chunks to be used as context.\n\nThe context obtained from files is later used in\n`/chat/completions`, `/completions`, and `/chunks` APIs.\n\nMost common document\nformats are supported, but you may be prompted to install an extra dependency to\nmanage a specific file type.\n\nA file can generate different Documents (for example a PDF generates one Document\nper page). All Documents IDs are returned in the response, together with the\nextracted Metadata (which is later used to improve context retrieval). Those IDs\ncan be used to filter the context used to create responses in\n`/chat/completions`, `/completions`, and `/chunks` APIs.",
"operationId": "ingest_v1_ingest_post",
"requestBody": {
"content": {
"multipart/form-data": {
"schema": {
"$ref": "#/components/schemas/Body_ingest_v1_ingest_post"
}
}
},
"required": true
},
"responses": {
"200": {
"description": "Successful Response",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/IngestResponse"
}
}
}
},
"422": {
"description": "Validation Error",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/HTTPValidationError"
}
}
}
}
}
}
},
"/v1/ingest/list": {
"get": {
"tags": [
"Ingestion"
],
"summary": "List Ingested",
"description": "Lists already ingested Documents including their Document ID and metadata.\n\nThose IDs can be used to filter the context used to create responses\nin `/chat/completions`, `/completions`, and `/chunks` APIs.",
"operationId": "list_ingested_v1_ingest_list_get",
"responses": {
"200": {
"description": "Successful Response",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/IngestResponse"
}
}
}
}
}
}
},
"/v1/ingest/{doc_id}": {
"delete": {
"tags": [
"Ingestion"
],
"summary": "Delete Ingested",
"description": "Delete the specified ingested Document.\n\nThe `doc_id` can be obtained from the `GET /ingest/list` endpoint.\nThe document will be effectively deleted from your storage context.",
"operationId": "delete_ingested_v1_ingest__doc_id__delete",
"parameters": [
{
"name": "doc_id",
"in": "path",
"required": true,
"schema": {
"type": "string",
"title": "Doc Id"
}
}
],
"responses": {
"200": {
"description": "Successful Response",
"content": {
"application/json": {
"schema": {}
}
}
},
"422": {
"description": "Validation Error",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/HTTPValidationError"
}
}
}
}
}
}
},
"/v1/embeddings": {
"post": {
"tags": [
"Embeddings"
],
"summary": "Embeddings Generation",
"description": "Get a vector representation of a given input.\n\nThat vector representation can be easily consumed\nby machine learning models and algorithms.",
"operationId": "embeddings_generation_v1_embeddings_post",
"requestBody": {
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/EmbeddingsBody"
}
}
},
"required": true
},
"responses": {
"200": {
"description": "Successful Response",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/EmbeddingsResponse"
}
}
}
},
"422": {
"description": "Validation Error",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/HTTPValidationError"
}
}
}
}
}
}
},
"/health": {
"get": {
"tags": [
"Health"
],
"summary": "Health",
"description": "Return ok if the system is up.",
"operationId": "health_health_get",
"responses": {
"200": {
"description": "Successful Response",
"content": {
"application/json": {
"schema": {
"$ref": "#/components/schemas/HealthResponse"
}
}
}
}
}
}
}
},
"components": {
"schemas": {
"Body_ingest_v1_ingest_post": {
"properties": {
"file": {
"type": "string",
"format": "binary",
"title": "File"
}
},
"type": "object",
"required": [
"file"
],
"title": "Body_ingest_v1_ingest_post"
},
"ChatBody": {
"properties": {
"messages": {
"items": {
"$ref": "#/components/schemas/OpenAIMessage"
},
"type": "array",
"title": "Messages"
},
"use_context": {
"type": "boolean",
"title": "Use Context",
"default": false
},
"context_filter": {
"anyOf": [
{
"$ref": "#/components/schemas/ContextFilter"
},
{
"type": "null"
}
]
},
"include_sources": {
"type": "boolean",
"title": "Include Sources",
"default": true
},
"stream": {
"type": "boolean",
"title": "Stream",
"default": false
}
},
"type": "object",
"required": [
"messages"
],
"title": "ChatBody",
"examples": [
{
"context_filter": {
"docs_ids": [
"c202d5e6-7b69-4869-81cc-dd574ee8ee11"
]
},
"include_sources": true,
"messages": [
{
"content": "How do you fry an egg?",
"role": "user"
}
],
"stream": false,
"use_context": true
}
]
},
"Chunk": {
"properties": {
"object": {
"const": "context.chunk",
"title": "Object"
},
"score": {
"type": "number",
"title": "Score",
"examples": [
0.023
]
},
"document": {
"$ref": "#/components/schemas/IngestedDoc"
},
"text": {
"type": "string",
"title": "Text",
"examples": [
"Outbound sales increased 20%, driven by new leads."
]
},
"previous_texts": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"title": "Previous Texts",
"examples": [
[
"SALES REPORT 2023",
"Inbound didn't show major changes."
]
]
},
"next_texts": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"title": "Next Texts",
"examples": [
[
"New leads came from Google Ads campaign.",
"The campaign was run by the Marketing Department"
]
]
}
},
"type": "object",
"required": [
"object",
"score",
"document",
"text"
],
"title": "Chunk"
},
"ChunksBody": {
"properties": {
"text": {
"type": "string",
"title": "Text",
"examples": [
"Q3 2023 sales"
]
},
"context_filter": {
"anyOf": [
{
"$ref": "#/components/schemas/ContextFilter"
},
{
"type": "null"
}
]
},
"limit": {
"type": "integer",
"title": "Limit",
"default": 10
},
"prev_next_chunks": {
"type": "integer",
"title": "Prev Next Chunks",
"default": 0,
"examples": [
2
]
}
},
"type": "object",
"required": [
"text"
],
"title": "ChunksBody"
},
"ChunksResponse": {
"properties": {
"object": {
"const": "list",
"title": "Object"
},
"model": {
"const": "private-gpt",
"title": "Model"
},
"data": {
"items": {
"$ref": "#/components/schemas/Chunk"
},
"type": "array",
"title": "Data"
}
},
"type": "object",
"required": [
"object",
"model",
"data"
],
"title": "ChunksResponse"
},
"CompletionsBody": {
"properties": {
"prompt": {
"type": "string",
"title": "Prompt"
},
"use_context": {
"type": "boolean",
"title": "Use Context",
"default": false
},
"context_filter": {
"anyOf": [
{
"$ref": "#/components/schemas/ContextFilter"
},
{
"type": "null"
}
]
},
"include_sources": {
"type": "boolean",
"title": "Include Sources",
"default": true
},
"stream": {
"type": "boolean",
"title": "Stream",
"default": false
}
},
"type": "object",
"required": [
"prompt"
],
"title": "CompletionsBody",
"examples": [
{
"include_sources": false,
"prompt": "How do you fry an egg?",
"stream": false,
"use_context": false
}
]
},
"ContextFilter": {
"properties": {
"docs_ids": {
"anyOf": [
{
"items": {
"type": "string"
},
"type": "array"
},
{
"type": "null"
}
],
"title": "Docs Ids",
"examples": [
[
"c202d5e6-7b69-4869-81cc-dd574ee8ee11"
]
]
}
},
"type": "object",
"required": [
"docs_ids"
],
"title": "ContextFilter"
},
"Embedding": {
"properties": {
"index": {
"type": "integer",
"title": "Index"
},
"object": {
"const": "embedding",
"title": "Object"
},
"embedding": {
"items": {
"type": "number"
},
"type": "array",
"title": "Embedding",
"examples": [
[
0.0023064255,
-0.009327292
]
]
}
},
"type": "object",
"required": [
"index",
"object",
"embedding"
],
"title": "Embedding"
},
"EmbeddingsBody": {
"properties": {
"input": {
"anyOf": [
{
"type": "string"
},
{
"items": {
"type": "string"
},
"type": "array"
}
],
"title": "Input"
}
},
"type": "object",
"required": [
"input"
],
"title": "EmbeddingsBody"
},
"EmbeddingsResponse": {
"properties": {
"object": {
"const": "list",
"title": "Object"
},
"model": {
"const": "private-gpt",
"title": "Model"
},
"data": {
"items": {
"$ref": "#/components/schemas/Embedding"
},
"type": "array",
"title": "Data"
}
},
"type": "object",
"required": [
"object",
"model",
"data"
],
"title": "EmbeddingsResponse"
},
"HTTPValidationError": {
"properties": {
"detail": {
"items": {
"$ref": "#/components/schemas/ValidationError"
},
"type": "array",
"title": "Detail"
}
},
"type": "object",
"title": "HTTPValidationError"
},
"HealthResponse": {
"properties": {
"status": {
"const": "ok",
"title": "Status",
"default": "ok"
}
},
"type": "object",
"title": "HealthResponse"
},
"IngestResponse": {
"properties": {
"object": {
"const": "list",
"title": "Object"
},
"model": {
"const": "private-gpt",
"title": "Model"
},
"data": {
"items": {
"$ref": "#/components/schemas/IngestedDoc"
},
"type": "array",
"title": "Data"
}
},
"type": "object",
"required": [
"object",
"model",
"data"
],
"title": "IngestResponse"
},
"IngestedDoc": {
"properties": {
"object": {
"const": "ingest.document",
"title": "Object"
},
"doc_id": {
"type": "string",
"title": "Doc Id",
"examples": [
"c202d5e6-7b69-4869-81cc-dd574ee8ee11"
]
},
"doc_metadata": {
"anyOf": [
{
"type": "object"
},
{
"type": "null"
}
],
"title": "Doc Metadata",
"examples": [
{
"file_name": "Sales Report Q3 2023.pdf",
"page_label": "2"
}
]
}
},
"type": "object",
"required": [
"object",
"doc_id",
"doc_metadata"
],
"title": "IngestedDoc"
},
"OpenAIChoice": {
"properties": {
"finish_reason": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Finish Reason",
"examples": [
"stop"
]
},
"delta": {
"anyOf": [
{
"$ref": "#/components/schemas/OpenAIDelta"
},
{
"type": "null"
}
]
},
"message": {
"anyOf": [
{
"$ref": "#/components/schemas/OpenAIMessage"
},
{
"type": "null"
}
]
},
"sources": {
"anyOf": [
{
"items": {
"$ref": "#/components/schemas/Chunk"
},
"type": "array"
},
{
"type": "null"
}
],
"title": "Sources"
},
"index": {
"type": "integer",
"title": "Index",
"default": 0
}
},
"type": "object",
"required": [
"finish_reason"
],
"title": "OpenAIChoice",
"description": "Response from AI.\n\nEither the delta or the message will be present, but never both.\nSources used will be returned in case context retrieval was enabled."
},
"OpenAICompletion": {
"properties": {
"id": {
"type": "string",
"title": "Id"
},
"object": {
"type": "string",
"enum": [
"completion",
"completion.chunk"
],
"title": "Object",
"default": "completion"
},
"created": {
"type": "integer",
"title": "Created",
"examples": [
1623340000
]
},
"model": {
"const": "private-gpt",
"title": "Model"
},
"choices": {
"items": {
"$ref": "#/components/schemas/OpenAIChoice"
},
"type": "array",
"title": "Choices"
}
},
"type": "object",
"required": [
"id",
"created",
"model",
"choices"
],
"title": "OpenAICompletion",
"description": "Clone of OpenAI Completion model.\n\nFor more information see: https://platform.openai.com/docs/api-reference/chat/object"
},
"OpenAIDelta": {
"properties": {
"content": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Content"
}
},
"type": "object",
"required": [
"content"
],
"title": "OpenAIDelta",
"description": "A piece of completion that needs to be concatenated to get the full message."
},
"OpenAIMessage": {
"properties": {
"role": {
"type": "string",
"enum": [
"assistant",
"system",
"user"
],
"title": "Role",
"default": "user"
},
"content": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"title": "Content"
}
},
"type": "object",
"required": [
"content"
],
"title": "OpenAIMessage",
"description": "Inference result, with the source of the message.\n\nRole could be the assistant or system\n(providing a default response, not AI generated)."
},
"ValidationError": {
"properties": {
"loc": {
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "integer"
}
]
},
"type": "array",
"title": "Location"
},
"msg": {
"type": "string",
"title": "Message"
},
"type": {
"type": "string",
"title": "Error Type"
}
},
"type": "object",
"required": [
"loc",
"msg",
"type"
],
"title": "ValidationError"
}
}
},
"tags": [
{
"name": "Ingestion",
"description": "High-level APIs covering document ingestion -internally managing document parsing, splitting,metadata extraction, embedding generation and storage- and ingested documents CRUD.Each ingested document is identified by an ID that can be used to filter the contextused in *Contextual Completions* and *Context Chunks* APIs."
},
{
"name": "Contextual Completions",
"description": "High-level APIs covering contextual Chat and Completions. They follow OpenAI's format, extending it to allow using the context coming from ingested documents to create the response. Internallymanage context retrieval, prompt engineering and the response generation."
},
{
"name": "Context Chunks",
"description": "Low-level API that given a query return relevant chunks of text coming from the ingesteddocuments."
},
{
"name": "Embeddings",
"description": "Low-level API to obtain the vector representation of a given text, using an Embeddings model.Follows OpenAI's embeddings API format."
},
{
"name": "Health",
"description": "Simple health API to make sure the server is up and running."
}
]
} |