kanshan/graphrag
1
0
Fork 0
mirror of https://github.com/microsoft/graphrag.git synced 2026-02-14 23:15:51 +08:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

Merge branch 'main' into feature/drift_search_v0

Browse Source
This commit is contained in:
Alonso Guevara 2024-10-14 17:06:12 -06:00
commit 5afae11a0e
6 changed files with 606 additions and 556 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
.semversioner/next-release/patch-20241011223119275846.json Normal file
View File

@ -0,0 +1,4 @@
{
"type": "patch",
"description": "Dependency updates"
}

4
.semversioner/next-release/patch-20241014140818103774.json Normal file
View File

@ -0,0 +1,4 @@
{
"type": "patch",
"description": "fix docs image path"
}

2
docs/prompt_tuning/auto_prompt_tuning.md
View File

@ -5,7 +5,7 @@ GraphRAG provides the ability to create domain adapted prompts for the generatio
These are generated by loading the inputs, splitting them into chunks (text units) and then running a series of LLM invocations and template substitutions to generate the final prompts. We suggest using the default values provided by the script, but in this page you'll find the detail of each in case you want to further explore and tweak the prompt tuning algorithm.
<p align="center">
<img src="../../../img/auto-tune-diagram.png" alt="Figure 1: Auto Tuning Conceptual Diagram." width="850" align="center" />
<img src="../../img/auto-tune-diagram.png" alt="Figure 1: Auto Tuning Conceptual Diagram." width="850" align="center" />
</p>
<p align="center">
Figure 1: Auto Tuning Conceptual Diagram.

18
graphrag/vector_stores/azure_ai_search.py
View File

@ -41,9 +41,9 @@ class AzureAISearch(BaseVectorStore):
def connect(self, **kwargs: Any) -> Any:
"""Connect to the AzureAI vector store."""
url = kwargs.get("url", None)
api_key = kwargs.get("api_key", None)
audience = kwargs.get("audience", None)
url = kwargs.get("url")
api_key = kwargs.get("api_key")
audience = kwargs.get("audience")
self.vector_size = kwargs.get("vector_size", DEFAULT_VECTOR_SIZE)
self.vector_search_profile_name = kwargs.get(
@ -55,16 +55,16 @@ class AzureAISearch(BaseVectorStore):
self.db_connection = SearchClient(
endpoint=url,
index_name=self.collection_name,
credential=AzureKeyCredential(api_key)
if api_key
else DefaultAzureCredential(),
credential=(
AzureKeyCredential(api_key) if api_key else DefaultAzureCredential()
),
**audience_arg,
)
self.index_client = SearchIndexClient(
endpoint=url,
credential=AzureKeyCredential(api_key)
if api_key
else DefaultAzureCredential(),
credential=(
AzureKeyCredential(api_key) if api_key else DefaultAzureCredential()
),
**audience_arg,
)
else:

1122
poetry.lock generated
View File

File diff suppressed because it is too large Load Diff

12
pyproject.toml
View File

@ -55,7 +55,7 @@ aiolimiter = "^1.1.0"
aiofiles = "^24.1.0"
# LLM
openai = "^1.46.1"
openai = "^1.51.2"
nltk = "3.9.1"
tiktoken = "^0.7.0"
@ -63,7 +63,7 @@ tiktoken = "^0.7.0"
numpy = "^1.25.2"
graspologic = "^3.4.1"
networkx = "^3"
pandas = "^2.2.2"
pandas = "^2.2.3"
matplotlib = "^3.9.0"
pyarrow = "^15.0.0"
umap-learn = "^0.5.6"
@ -84,7 +84,7 @@ typing-extensions = "^4.12.2"
#Azure
azure-storage-blob = "^12.22.0"
azure-identity = "^1.17.1"
json-repair = "^0.28.4"
json-repair = "^0.30.0"
future = "^1.0.0" # Needed until graspologic fixes their dependency
@ -94,14 +94,14 @@ ipykernel = "^6.29.4"
jupyter = "^1.1.1"
nbconvert = "^7.16.3"
poethepoet = "^0.28.0"
pyright = "^1.1.371"
pyright = "^1.1.384"
pytest = "^8.3.2"
pytest-asyncio = "^0.24.0"
pytest-timeout = "^2.3.1"
ruff = "^0.6.5"
ruff = "^0.6.9"
semversioner = "^2.0.3"
update-toml = "^0.2.1"
deptry = "^0.16.1"
deptry = "^0.20.0"
mkdocs-material = "^9.5.39"
mkdocs-jupyter = "^0.25.0"
mkdocs-exclude-search = "^0.6.6"