This section provides guidance on running demonstration files located in the knowledge_graph
directory of
the legacy_code_assistant
project.
- Location:
.demo/
-
github_graph_DEMO_1_GIF.py
: Visualizes GitHub graphs. -
repo_code_graph_DEMO_2_GIF.py
: Demonstrates repository code graphs.- Note for DEMO 2:
- To integrate an LLM into this demo, locate the
TODO
comment in the code where additional processing should be added. - Currently, the template selection feature is not operational. This needs to be fixed to enable proper functionality.
- To integrate an LLM into this demo, locate the
- Note for DEMO 2:
To run the Streamlit demonstrations for each script, use the following commands in PowerShell:
- For GitHub Graph Visualization:
streamlit run github_graph_DEMO_1_GIF.py
- For Repository Code Graph:
streamlit run repo_code_graph_DEMO_2_GIF.py
Ensure the repository path is set to tests/test_repo
. To clone the required repository, execute the repo_cloner
script:
python legacy_code_assistant/data_extraction/repo_cloner.py
You must provide credentials in the demo folder. A file credentials.yaml should be placed there. It should have the following structure:
'AZURE_OPENAI_ENDPOINT': <URL OF AZURE ENDPOINT>
'AZURE_OPENAI_API_KEY': <KEY>
'Deployment_completion': <AZURE DEPLOYMENT FOR COMPLETION MODEL (e.g. gpt4)>
'Deployment_embeddings': <AZURE DEPLOYMENT FOR EMBEDDING MODEL (e.g. ada)>