feat: Qdrant support (#1228)

* feat: Qdrant support

* Update private_gpt/components/vector_store/vector_store_component.py
This commit is contained in:
Anush 2023-11-14 01:53:26 +05:30 committed by GitHub
parent 86fc4781d8
commit 03d1ae6d70
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
6 changed files with 320 additions and 18 deletions

View File

@ -248,6 +248,34 @@ llama_new_context_with_model: total VRAM used: 4857.93 MB (model: 4095.05 MB, co
AVX = 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 |
```
#### Vectorstores
PrivateGPT supports [Chroma](https://www.trychroma.com/), [Qdrant](https://qdrant.tech/) as vectorstore providers. Chroma being the default.
To enable Qdrant, set the `vectorstore.database` property in the `settings.yaml` file to `qdrant` and install the `qdrant` extra.
```bash
poetry install --extras qdrant
```
By default Qdrant tries to connect to an instance at `http://localhost:3000`.
Qdrant settings can be configured by setting values to the `qdrant` propery in the `settings.yaml` file.
The available configuration options are:
| Field | Description |
|--------------|-------------|
| location | If `:memory:` - use in-memory Qdrant instance.<br>If `str` - use it as a `url` parameter.|
| url | Either host or str of 'Optional[scheme], host, Optional[port], Optional[prefix]'.<br> Eg. `http://localhost:6333` |
| port | Port of the REST API interface. Default: `6333` |
| grpc_port | Port of the gRPC interface. Default: `6334` |
| prefer_grpc | If `true` - use gRPC interface whenever possible in custom methods. |
| https | If `true` - use HTTPS(SSL) protocol.|
| api_key | API key for authentication in Qdrant Cloud.|
| 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.|
| timeout | Timeout for REST and gRPC API requests.<br>Default: 5.0 seconds for REST and unlimited for gRPC |
| host | Host name of Qdrant service. If url and host are not set, defaults to 'localhost'.|
| path | Persistence path for QdrantLocal. Eg. `local_data/private_gpt/qdrant`|
#### Known issues and Troubleshooting
Execution of LLMs locally still has a lot of sharp edges, specially when running on non Linux platforms.

178
poetry.lock generated
View File

@ -1,4 +1,4 @@
# This file is automatically @generated by Poetry 1.7.0 and should not be changed by hand.
# This file is automatically @generated by Poetry 1.6.1 and should not be changed by hand.
[[package]]
name = "aiofiles"
@ -1339,6 +1339,74 @@ files = [
[package.extras]
protobuf = ["grpcio-tools (>=1.59.2)"]
[[package]]
name = "grpcio-tools"
version = "1.59.2"
description = "Protobuf code generator for gRPC"
optional = true
python-versions = ">=3.7"
files = [
{file = "grpcio-tools-1.59.2.tar.gz", hash = "sha256:75905266cf90f1866b322575c2edcd4b36532c33fc512bb1b380dc58d84b1030"},
{file = "grpcio_tools-1.59.2-cp310-cp310-linux_armv7l.whl", hash = "sha256:9b2885c0e2c9a97bde33497a919032afbd8b5c6dc2f8d4dd4198e77226e0de05"},
{file = "grpcio_tools-1.59.2-cp310-cp310-macosx_12_0_universal2.whl", hash = "sha256:2f410375830a9bb7140a07da4d75bf380e0958377bed50d77d1dae302de4314e"},
{file = "grpcio_tools-1.59.2-cp310-cp310-manylinux_2_17_aarch64.whl", hash = "sha256:e21fc172522d2dda815223a359b2aca9bc317a1b5e5dea5a58cd5079333af133"},
{file = "grpcio_tools-1.59.2-cp310-cp310-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:072a7ce979ea4f7579c3c99fcbde3d1882c3d1942a3b51d159f67af83b714cd8"},
{file = "grpcio_tools-1.59.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:b38f8edb2909702c2478b52f6213982c21e4f66f739ac953b91f97863ba2c06a"},
{file = "grpcio_tools-1.59.2-cp310-cp310-musllinux_1_1_i686.whl", hash = "sha256:12fdee2de80d83eadb1294e0f8a0cb6cefcd2e4988ed680038ab09cd04361ee4"},
{file = "grpcio_tools-1.59.2-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:a3cb707da722a0b6c4021fc2cc1c005a8d4037d8ad0252f93df318b9b8a6b4f3"},
{file = "grpcio_tools-1.59.2-cp310-cp310-win32.whl", hash = "sha256:ec2fbb02ebb9f2ae1b1c69cccf913dee8c41f5acad94014d3ce11b53720376e3"},
{file = "grpcio_tools-1.59.2-cp310-cp310-win_amd64.whl", hash = "sha256:b0dc271a200dbab6547b2c73fcbdb7efe94c31cb633aa20d073f7cf4493493e1"},
{file = "grpcio_tools-1.59.2-cp311-cp311-linux_armv7l.whl", hash = "sha256:d634b65cc8ee769edccf1647d8a16861a27e0d8cbd787c711168d2c5e9bddbd1"},
{file = "grpcio_tools-1.59.2-cp311-cp311-macosx_10_10_universal2.whl", hash = "sha256:b0b712acec00a9cbc2204c271d638062a2cb8ce74f25d158b023ff6e93182659"},
{file = "grpcio_tools-1.59.2-cp311-cp311-manylinux_2_17_aarch64.whl", hash = "sha256:dd5c78f8e7c6e721b9009c92481a0e3b30a9926ef721120723a03b8a34a34fb9"},
{file = "grpcio_tools-1.59.2-cp311-cp311-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:724f4f0eecc17fa66216eebfff145631070f04ed7fb4ddf7a7d1c4f954ecc2a1"},
{file = "grpcio_tools-1.59.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:77ec33ddee691e60511e2a7c793aad4cf172ae20e08d95c786cbba395f6203a7"},
{file = "grpcio_tools-1.59.2-cp311-cp311-musllinux_1_1_i686.whl", hash = "sha256:fa1b9dee7811fad081816e884d063c4dd4946dba61aa54243b4c76c311090c48"},
{file = "grpcio_tools-1.59.2-cp311-cp311-musllinux_1_1_x86_64.whl", hash = "sha256:ba8dba19e7b2b6f7369004533866f222ba483b9e14d2d152ecf9339c0df1283a"},
{file = "grpcio_tools-1.59.2-cp311-cp311-win32.whl", hash = "sha256:df35d145bc2f6e5f57b74cb69f66526675a5f2dcf7d54617ce0deff0c82cca0a"},
{file = "grpcio_tools-1.59.2-cp311-cp311-win_amd64.whl", hash = "sha256:99ddc0f5304071a355c261ae49ea5d29b9e9b6dcf422dfc55ada70a243e27e8f"},
{file = "grpcio_tools-1.59.2-cp312-cp312-linux_armv7l.whl", hash = "sha256:670f5889853215999eb3511a623dd7dff01b1ce1a64610d13366e0fd337f8c79"},
{file = "grpcio_tools-1.59.2-cp312-cp312-macosx_10_10_universal2.whl", hash = "sha256:1e949e66d4555ce319fd7acef90df625138078d8729c4dc6f6a9f05925034433"},
{file = "grpcio_tools-1.59.2-cp312-cp312-manylinux_2_17_aarch64.whl", hash = "sha256:09d809ca88999b2578119683f9f0f6a9b42de95ea21550852114a1540b6a642c"},
{file = "grpcio_tools-1.59.2-cp312-cp312-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:db0925545180223fabd6da9b34513efac83aa16673ef8b1cb0cc678e8cf0923c"},
{file = "grpcio_tools-1.59.2-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:a2ccb59dfbf2ebd668a5a7c4b7bb2b859859641d2b199114b557cd045aac6102"},
{file = "grpcio_tools-1.59.2-cp312-cp312-musllinux_1_1_i686.whl", hash = "sha256:12cc7698fad48866f68fdef831685cb31ef5814ac605d248c4e5fc964a6fb3f6"},
{file = "grpcio_tools-1.59.2-cp312-cp312-musllinux_1_1_x86_64.whl", hash = "sha256:55c401599d5093c4cfa83b8f0ee9757b4d6d3029b10bd67be2cffeada7a44961"},
{file = "grpcio_tools-1.59.2-cp312-cp312-win32.whl", hash = "sha256:896f5cdf58f658025a4f7e4ea96c81183b4b6a4b1b4d92ae66d112ac91f062f1"},
{file = "grpcio_tools-1.59.2-cp312-cp312-win_amd64.whl", hash = "sha256:b53db1523015a3acda75722357df6c94afae37f6023800c608e09a5c05393804"},
{file = "grpcio_tools-1.59.2-cp37-cp37m-linux_armv7l.whl", hash = "sha256:d08b398509ea4d544bcecddd9a21f59dc556396916c3915904cac206af2db72b"},
{file = "grpcio_tools-1.59.2-cp37-cp37m-macosx_10_10_universal2.whl", hash = "sha256:09749e832e06493841000275248b031f7154665900d1e1b0e42fc17a64bf904d"},
{file = "grpcio_tools-1.59.2-cp37-cp37m-manylinux_2_17_aarch64.whl", hash = "sha256:e972746000aa192521715f776fab617a3437bed29e90fe0e0fd0d0d6f498d7d4"},
{file = "grpcio_tools-1.59.2-cp37-cp37m-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:cbeeb3d8ec4cb25c92e17bfbdcef3c3669e85c5ee787a6e581cb942bc0ae2b88"},
{file = "grpcio_tools-1.59.2-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:ed8e6632d8d839456332d97b96db10bd2dbf3078e728d063394ac2d54597ad80"},
{file = "grpcio_tools-1.59.2-cp37-cp37m-musllinux_1_1_i686.whl", hash = "sha256:531f87c8e884c6a2e58f040039dfbfe997a4e33baa58f7c7d9993db37b1f5ad0"},
{file = "grpcio_tools-1.59.2-cp37-cp37m-musllinux_1_1_x86_64.whl", hash = "sha256:feca316e17cfead823af6eae0fc20c0d5299a94d71cfb7531a0e92d050a5fb2f"},
{file = "grpcio_tools-1.59.2-cp37-cp37m-win_amd64.whl", hash = "sha256:41b5dd6a06c2563ac3b3adda6d875b15e63eb7b1629e85fc9af608c3a76c4c82"},
{file = "grpcio_tools-1.59.2-cp38-cp38-linux_armv7l.whl", hash = "sha256:7ec536cdae870a74080c665cfb1dca8d0784a931aa3c26376ef971a3a51b59d4"},
{file = "grpcio_tools-1.59.2-cp38-cp38-macosx_10_10_universal2.whl", hash = "sha256:9c106ebbed0db446f59f0efe5c3fce33a0a21bf75b392966585e4b5934891b92"},
{file = "grpcio_tools-1.59.2-cp38-cp38-manylinux_2_17_aarch64.whl", hash = "sha256:32141ef309543a446337e934f0b7a2565a6fca890ff4e543630a09ef72c8d00b"},
{file = "grpcio_tools-1.59.2-cp38-cp38-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:5f2ce5ecd63c492949b03af73b1dd6d502c567cc2f9c2057137e518b0c702a01"},
{file = "grpcio_tools-1.59.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:2a9ce2a209871ed1c5ae2229e6f4f5a3ea96d83b7871df5d9773d72a72545683"},
{file = "grpcio_tools-1.59.2-cp38-cp38-musllinux_1_1_i686.whl", hash = "sha256:7f0e26af7c07bfa906c91ca9f5932514928a7f032f5f20aecad6b5541037de7e"},
{file = "grpcio_tools-1.59.2-cp38-cp38-musllinux_1_1_x86_64.whl", hash = "sha256:48782727c5cff8b8c96e028a8a58614ff6a37eadc0db85866516210c7aafe9ae"},
{file = "grpcio_tools-1.59.2-cp38-cp38-win32.whl", hash = "sha256:4a1810bc5de51cc162a19ed3c11da8ddc64d8cfcba049ef337c20fcb397f048b"},
{file = "grpcio_tools-1.59.2-cp38-cp38-win_amd64.whl", hash = "sha256:3cf9949a2aadcece3c1e0dd59249aea53dbfc8cc94f7d707797acd67cf6cf931"},
{file = "grpcio_tools-1.59.2-cp39-cp39-linux_armv7l.whl", hash = "sha256:f52e0ce8f2dcf1f160c847304016c446075a83ab925d98933d4681bfa8af2962"},
{file = "grpcio_tools-1.59.2-cp39-cp39-macosx_10_10_universal2.whl", hash = "sha256:eb597d6bf9f5bfa54d00546e828f0d4e2c69250d1bc17c27903c0c7b66372135"},
{file = "grpcio_tools-1.59.2-cp39-cp39-manylinux_2_17_aarch64.whl", hash = "sha256:17ef468836d7cf0b2419f4d5c7ac84ec2d598a1ae410773585313edacf7c393e"},
{file = "grpcio_tools-1.59.2-cp39-cp39-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:dee5f7e7a56177234e61a483c70ca2ae34e73128372c801bb7039993870889f1"},
{file = "grpcio_tools-1.59.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:f50ff312b88918c5a6461e45c5e03869749a066b1c24a7327e8e13e117efe4fc"},
{file = "grpcio_tools-1.59.2-cp39-cp39-musllinux_1_1_i686.whl", hash = "sha256:a85da4200295ee17e3c1ae068189a43844420ed7e9d531a042440f52de486dfb"},
{file = "grpcio_tools-1.59.2-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:f518f22a3082de00f0d7a216e96366a87e6973111085ba1603c3bfa7dba2e728"},
{file = "grpcio_tools-1.59.2-cp39-cp39-win32.whl", hash = "sha256:6e735a26e8ea8bb89dc69343d1d00ea607449c6d81e21f339ee118562f3d1931"},
{file = "grpcio_tools-1.59.2-cp39-cp39-win_amd64.whl", hash = "sha256:3491cb69c909d586c23d7e6d0ac87844ca22f496f505ce429c0d3301234f2cf3"},
]
[package.dependencies]
grpcio = ">=1.59.2"
protobuf = ">=4.21.6,<5.0dev"
setuptools = "*"
[[package]]
name = "h11"
version = "0.14.0"
@ -1350,6 +1418,32 @@ files = [
{file = "h11-0.14.0.tar.gz", hash = "sha256:8f19fbbe99e72420ff35c00b27a34cb9937e902a8b810e2c88300c6f0a3b699d"},
]
[[package]]
name = "h2"
version = "4.1.0"
description = "HTTP/2 State-Machine based protocol implementation"
optional = true
python-versions = ">=3.6.1"
files = [
{file = "h2-4.1.0-py3-none-any.whl", hash = "sha256:03a46bcf682256c95b5fd9e9a99c1323584c3eec6440d379b9903d709476bc6d"},
{file = "h2-4.1.0.tar.gz", hash = "sha256:a83aca08fbe7aacb79fec788c9c0bac936343560ed9ec18b82a13a12c28d2abb"},
]
[package.dependencies]
hpack = ">=4.0,<5"
hyperframe = ">=6.0,<7"
[[package]]
name = "hpack"
version = "4.0.0"
description = "Pure-Python HPACK header compression"
optional = true
python-versions = ">=3.6.1"
files = [
{file = "hpack-4.0.0-py3-none-any.whl", hash = "sha256:84a076fad3dc9a9f8063ccb8041ef100867b1878b25ef0ee63847a5d53818a6c"},
{file = "hpack-4.0.0.tar.gz", hash = "sha256:fc41de0c63e687ebffde81187a948221294896f6bdc0ae2312708df339430095"},
]
[[package]]
name = "httpcore"
version = "1.0.2"
@ -1433,6 +1527,7 @@ files = [
[package.dependencies]
anyio = "*"
certifi = "*"
h2 = {version = ">=3,<5", optional = true, markers = "extra == \"http2\""}
httpcore = "*"
idna = "*"
sniffio = "*"
@ -1490,6 +1585,17 @@ files = [
[package.dependencies]
pyreadline3 = {version = "*", markers = "sys_platform == \"win32\" and python_version >= \"3.8\""}
[[package]]
name = "hyperframe"
version = "6.0.1"
description = "HTTP/2 framing layer for Python"
optional = true
python-versions = ">=3.6.1"
files = [
{file = "hyperframe-6.0.1-py3-none-any.whl", hash = "sha256:0ec6bafd80d8ad2195c4f03aacba3a8265e57bc4cff261e802bf39970ed02a15"},
{file = "hyperframe-6.0.1.tar.gz", hash = "sha256:ae510046231dc8e9ecb1a6586f63d2347bf4c8905914aa84ba585ae85f28a914"},
]
[[package]]
name = "identify"
version = "2.5.31"
@ -2941,6 +3047,25 @@ files = [
dev = ["pre-commit", "tox"]
testing = ["pytest", "pytest-benchmark"]
[[package]]
name = "portalocker"
version = "2.8.2"
description = "Wraps the portalocker recipe for easy usage"
optional = true
python-versions = ">=3.8"
files = [
{file = "portalocker-2.8.2-py3-none-any.whl", hash = "sha256:cfb86acc09b9aa7c3b43594e19be1345b9d16af3feb08bf92f23d4dce513a28e"},
{file = "portalocker-2.8.2.tar.gz", hash = "sha256:2b035aa7828e46c58e9b31390ee1f169b98e1066ab10b9a6a861fe7e25ee4f33"},
]
[package.dependencies]
pywin32 = {version = ">=226", markers = "platform_system == \"Windows\""}
[package.extras]
docs = ["sphinx (>=1.7.1)"]
redis = ["redis"]
tests = ["pytest (>=5.4.1)", "pytest-cov (>=2.8.1)", "pytest-mypy (>=0.8.0)", "pytest-timeout (>=2.1.0)", "redis", "sphinx (>=6.0.0)", "types-redis"]
[[package]]
name = "posthog"
version = "3.0.2"
@ -3416,6 +3541,29 @@ files = [
{file = "pytz-2023.3.post1.tar.gz", hash = "sha256:7b4fddbeb94a1eba4b557da24f19fdf9db575192544270a9101d8509f9f43d7b"},
]
[[package]]
name = "pywin32"
version = "306"
description = "Python for Window Extensions"
optional = true
python-versions = "*"
files = [
{file = "pywin32-306-cp310-cp310-win32.whl", hash = "sha256:06d3420a5155ba65f0b72f2699b5bacf3109f36acbe8923765c22938a69dfc8d"},
{file = "pywin32-306-cp310-cp310-win_amd64.whl", hash = "sha256:84f4471dbca1887ea3803d8848a1616429ac94a4a8d05f4bc9c5dcfd42ca99c8"},
{file = "pywin32-306-cp311-cp311-win32.whl", hash = "sha256:e65028133d15b64d2ed8f06dd9fbc268352478d4f9289e69c190ecd6818b6407"},
{file = "pywin32-306-cp311-cp311-win_amd64.whl", hash = "sha256:a7639f51c184c0272e93f244eb24dafca9b1855707d94c192d4a0b4c01e1100e"},
{file = "pywin32-306-cp311-cp311-win_arm64.whl", hash = "sha256:70dba0c913d19f942a2db25217d9a1b726c278f483a919f1abfed79c9cf64d3a"},
{file = "pywin32-306-cp312-cp312-win32.whl", hash = "sha256:383229d515657f4e3ed1343da8be101000562bf514591ff383ae940cad65458b"},
{file = "pywin32-306-cp312-cp312-win_amd64.whl", hash = "sha256:37257794c1ad39ee9be652da0462dc2e394c8159dfd913a8a4e8eb6fd346da0e"},
{file = "pywin32-306-cp312-cp312-win_arm64.whl", hash = "sha256:5821ec52f6d321aa59e2db7e0a35b997de60c201943557d108af9d4ae1ec7040"},
{file = "pywin32-306-cp37-cp37m-win32.whl", hash = "sha256:1c73ea9a0d2283d889001998059f5eaaba3b6238f767c9cf2833b13e6a685f65"},
{file = "pywin32-306-cp37-cp37m-win_amd64.whl", hash = "sha256:72c5f621542d7bdd4fdb716227be0dd3f8565c11b280be6315b06ace35487d36"},
{file = "pywin32-306-cp38-cp38-win32.whl", hash = "sha256:e4c092e2589b5cf0d365849e73e02c391c1349958c5ac3e9d5ccb9a28e017b3a"},
{file = "pywin32-306-cp38-cp38-win_amd64.whl", hash = "sha256:e8ac1ae3601bee6ca9f7cb4b5363bf1c0badb935ef243c4733ff9a393b1690c0"},
{file = "pywin32-306-cp39-cp39-win32.whl", hash = "sha256:e25fd5b485b55ac9c057f67d94bc203f3f6595078d1fb3b458c9c28b7153a802"},
{file = "pywin32-306-cp39-cp39-win_amd64.whl", hash = "sha256:39b61c15272833b5c329a2989999dcae836b1eed650252ab1b7bfbe1d59f30f4"},
]
[[package]]
name = "pyyaml"
version = "6.0.1"
@ -3475,6 +3623,29 @@ files = [
{file = "PyYAML-6.0.1.tar.gz", hash = "sha256:bfdf460b1736c775f2ba9f6a92bca30bc2095067b8a9d77876d1fad6cc3b4a43"},
]
[[package]]
name = "qdrant-client"
version = "1.6.4"
description = "Client library for the Qdrant vector search engine"
optional = true
python-versions = ">=3.8,<3.13"
files = [
{file = "qdrant_client-1.6.4-py3-none-any.whl", hash = "sha256:db4696978d6a62d78ff60f70b912383f1e467bda3053f732b01ddb5f93281b10"},
{file = "qdrant_client-1.6.4.tar.gz", hash = "sha256:bbd65f383b6a55a9ccf4e301250fa925179340dd90cfde9b93ce4230fd68867b"},
]
[package.dependencies]
grpcio = ">=1.41.0"
grpcio-tools = ">=1.41.0"
httpx = {version = ">=0.14.0", extras = ["http2"]}
numpy = {version = ">=1.21", markers = "python_version >= \"3.8\" and python_version < \"3.12\""}
portalocker = ">=2.7.0,<3.0.0"
pydantic = ">=1.10.8"
urllib3 = ">=1.26.14,<2.0.0"
[package.extras]
fastembed = ["fastembed (==0.1.1)"]
[[package]]
name = "referencing"
version = "0.30.2"
@ -5320,7 +5491,10 @@ files = [
docs = ["furo", "jaraco.packaging (>=9.3)", "jaraco.tidelift (>=1.4)", "rst.linker (>=1.9)", "sphinx (<7.2.5)", "sphinx (>=3.5)", "sphinx-lint"]
testing = ["big-O", "jaraco.functools", "jaraco.itertools", "more-itertools", "pytest (>=6)", "pytest-black (>=0.3.7)", "pytest-checkdocs (>=2.4)", "pytest-cov", "pytest-enabler (>=2.2)", "pytest-ignore-flaky", "pytest-mypy (>=0.9.1)", "pytest-ruff"]
[extras]
qdrant = ["qdrant-client"]
[metadata]
lock-version = "2.0"
python-versions = ">=3.11,<3.12"
content-hash = "3ea28113b04a06b1dfcbe6f43cdce086b596c714a53c10eb6501ab0c5d464412"
content-hash = "a72d664bdd24534d353deb2ad98cdebb5caa1647315405260c9c9f6e64d52f5a"

View File

@ -1,7 +1,8 @@
import logging
import typing
import chromadb
from chromadb.config import Settings
from chromadb.config import Settings as ChromaSettings
from injector import inject, singleton
from llama_index import VectorStoreIndex
from llama_index.indices.vector_store import VectorIndexRetriever
@ -10,6 +11,9 @@ from llama_index.vector_stores.types import VectorStore
from private_gpt.components.vector_store.batched_chroma import BatchedChromaVectorStore
from private_gpt.open_ai.extensions.context_filter import ContextFilter
from private_gpt.paths import local_data_path
from private_gpt.settings.settings import Settings
logger = logging.getLogger(__name__)
@typing.no_type_check
@ -36,8 +40,10 @@ class VectorStoreComponent:
vector_store: VectorStore
@inject
def __init__(self) -> None:
chroma_settings = Settings(anonymized_telemetry=False)
def __init__(self, settings: Settings) -> None:
match settings.vectorstore.database:
case "chroma":
chroma_settings = ChromaSettings(anonymized_telemetry=False)
chroma_client = chromadb.PersistentClient(
path=str((local_data_path / "chroma_db").absolute()),
settings=chroma_settings,
@ -53,6 +59,40 @@ class VectorStoreComponent:
),
)
case "qdrant":
try:
from llama_index.vector_stores.qdrant import QdrantVectorStore
from qdrant_client import QdrantClient # type: ignore
except ImportError as e:
raise ImportError(
"'qdrant_client' is not installed."
"To use PrivateGPT with Qdrant, install the 'qdrant' extra."
"`poetry install --extras qdrant`"
) from e
if settings.qdrant is None:
logger.info(
"Qdrant config not found. Using default settings."
"Trying to connect to Qdrant at localhost:6333."
)
client = QdrantClient()
else:
client = QdrantClient(
**settings.qdrant.model_dump(exclude_none=True)
)
self.vector_store = typing.cast(
VectorStore,
QdrantVectorStore(
client=client,
collection_name="make_this_parameterizable_per_api_call",
), # TODO
)
case _:
# Should be unreachable
# The settings validator should have caught this
raise ValueError(
f"Vectorstore database {settings.vectorstore.database} not supported"
)
@staticmethod
def get_retriever(
index: VectorStoreIndex,

View File

@ -84,6 +84,10 @@ class LLMSettings(BaseModel):
mode: Literal["local", "openai", "sagemaker", "mock"]
class VectorstoreSettings(BaseModel):
database: Literal["chroma", "qdrant"]
class LocalSettings(BaseModel):
llm_hf_repo_id: str
llm_hf_model_file: str
@ -104,6 +108,53 @@ class UISettings(BaseModel):
path: str
class QdrantSettings(BaseModel):
location: str | None = Field(
None,
description=(
"If `:memory:` - use in-memory Qdrant instance.\n"
"If `str` - use it as a `url` parameter.\n"
),
)
url: str | None = Field(
None,
description=(
"Either host or str of 'Optional[scheme], host, Optional[port], Optional[prefix]'."
),
)
port: int | None = Field(6333, description="Port of the REST API interface.")
grpc_port: int | None = Field(6334, description="Port of the gRPC interface.")
prefer_grpc: bool | None = Field(
False,
description="If `true` - use gRPC interface whenever possible in custom methods.",
)
https: bool | None = Field(
None,
description="If `true` - use HTTPS(SSL) protocol.",
)
api_key: str | None = Field(
None,
description="API key for authentication in Qdrant Cloud.",
)
prefix: str | None = Field(
None,
description=(
"Prefix to add to the REST URL path."
"Example: `service/v1` will result in "
"'http://localhost:6333/service/v1/{qdrant-endpoint}' for REST API."
),
)
timeout: float | None = Field(
None,
description="Timeout for REST and gRPC API requests.",
)
host: str | None = Field(
None,
description="Host name of Qdrant service. If url and host are None, set to 'localhost'.",
)
path: str | None = Field(None, description="Persistence path for QdrantLocal.")
class Settings(BaseModel):
server: ServerSettings
data: DataSettings
@ -112,6 +163,8 @@ class Settings(BaseModel):
local: LocalSettings
sagemaker: SagemakerSettings
openai: OpenAISettings
vectorstore: VectorstoreSettings
qdrant: QdrantSettings | None = None
"""

View File

@ -15,6 +15,7 @@ pypdf = "^3.16.2"
llama-index = "0.8.67"
chromadb = "^0.4.13"
watchdog = "^3.0.0"
qdrant-client = {version = "^1.6.4", optional = true}
[tool.poetry.group.dev.dependencies]
black = "^22"
@ -42,6 +43,9 @@ sentence-transformers = "^2.2.2"
torch = ">=2.0.0, !=2.0.1, !=2.1.0"
transformers = "^4.34.0"
[tool.poetry.extras]
qdrant = ["qdrant-client"]
[build-system]
requires = ["poetry-core>=1.0.0"]
build-backend = "poetry.core.masonry.api"

View File

@ -23,6 +23,9 @@ ui:
llm:
mode: local
vectorstore:
database: chroma
local:
llm_hf_repo_id: TheBloke/Mistral-7B-Instruct-v0.1-GGUF
llm_hf_model_file: mistral-7b-instruct-v0.1.Q4_K_M.gguf