run_experiment_grid

run_experiment_grid(
    specs=None,
    param_grid=None,
    task_root=None,
    output_root='outputs',
    metrics_output_root=None,
    pull_models=True,
    start_ollama_if_needed=True,
)

Run a whole parameter sweep from a grid or a prebuilt spec list.

Parameters

Name	Type	Description	Default
specs	list[ExperimentSpec] \| None	Optional pre-expanded list of `ExperimentSpec` objects.	`None`
param_grid	dict \| None	Optional in-memory parameter grid dictionary.	`None`
task_root	str \| Path \| None	Optional directory containing user-defined task folders.	`None`
output_root	str \| Path	Root directory for per-run outputs.	`'outputs'`
metrics_output_root	str \| Path \| None	Directory for aggregate metrics CSV files.	`None`
pull_models	bool	If `True`, pull each Ollama model before its first run. Defaults to `True` so sweep runs ensure requested models are available.	`True`
start_ollama_if_needed	bool	If `True`, try to auto-start a local Ollama server before each run when needed. Defaults to `True`.	`True`

Returns

Name	Type	Description
	list[ExperimentRunResult]	List of `ExperimentRunResult` objects, one per completed run.