View source on GitHub |
Run model analysis for a single model on multiple data sets.
tfma.make_eval_results(
results: List[tfma.EvalResult
],
mode: str
) -> view_types.EvalResults
Args | |
---|---|
results
|
A list of TFMA evaluation results. |
mode
|
The mode of the evaluation. Currently, tfma.DATA_CENTRIC_MODE and tfma.MODEL_CENTRIC_MODE are supported. |
Returns | |
---|---|
An tfma.view.EvalResults object containing all evaluation results. This
can be used to construct a time series view.
|