aymara_ai.lib.images_utils#
Functions#
|
Display a grid of image eval responses with their eval questions as captions. |
Module Contents#
- aymara_ai.lib.images_utils.as_eval_response_dict(obj)#
- Parameters:
obj (Any)
- Return type:
dict
- aymara_ai.lib.images_utils.display_image_responses(evals, eval_prompts, eval_responses, n_images_per_eval=5, figsize=None)#
Display a grid of image eval responses with their eval questions as captions. If eval runs are included, display their eval evals as captions instead and add a red border to failed images.
- Parameters:
evals (List of Eval objects.) – Evals corresponding to the eval responses.
eval_responses (Dictionary of eval UUIDs to lists of EvalResponseParam objects.) – Eval responses.
eval_runs (List of EvalRunResponse objects, optional) – Eval runs corresponding to the eval responses.
n_images_per_eval (int, optional) – Number of images to display per eval.
figsize (integer tuple, optional) – Figure size. Defaults to (n_images_per_eval * 3, n_evals * 2 * 4).
eval_prompts (Dict[str, List[aymara_ai.types.eval_prompt.EvalPrompt]])
- Return type:
None