Report Generation: Adding notebooks to better explain the usage#61
Report Generation: Adding notebooks to better explain the usage#61
Conversation
amrit110
left a comment
amrit110
left a comment
There was a problem hiding this comment.
Just the one comment. The code itself looks good!
There was a problem hiding this comment.
I'm generally not in favour of committing the outputs of cells. It just adds a lot more clutter to the git history, and usually it can change between runs as well. So consider clearing the outputs and only commit the code.
amrit110
left a comment
amrit110
left a comment
There was a problem hiding this comment.
Just minor typos. Actually you could try adding this pre-commit hook and run ot once for all the notebooks:
https://github.com/VectorInstitute/aieng-template-uv/blob/main/.pre-commit-config.yaml#L45C1-L49C17
| "\n", | ||
| "Offline evaluations are evaluations run against a **pre-defined dataset**. It performs **detailed evaluations** of the **outputs** of the agentic system and the **steps** it has taken to produce those evaluations.\n", | ||
| "\n", | ||
| "This dataset is called the **expected results** or the **ground-truth** dataset, and on this case it's a **handcrafted** dataset with **inputs, oputputs and trajectory** for a few known use cases.\n", |
| "source": [ | ||
| "## Running the Evaluations\n", | ||
| "\n", | ||
| "To run those two evaluatoirs against all of the ground-truth dataset samples, run the function below:" |
|
@amrit110 the precommit didn't work very well. It's picking up |
2 participants
Summary
Adding notebooks to better explain the usage and features of the Report Generation Agent. Also performing minor bug fixes.
Clickup Ticket(s): NA
Type of Change
Changes Made
Adding three notebooks to:
Also making a few adjustments and fixing a few bugs found while making the notebooks.
Testing
uv run pytest tests/)uv run mypy <src_dir>)uv run ruff check src_dir/)Manual testing details:
Ran the notebooks and made sure they work.
Checklist