Neural Network#

SAClient.download_model(model, output_dir)#

Downloads the neural network and related files which are the <model_name>.pth/pkl. <model_name>.json, <model_name>.yaml, classes_mapper.json

Parameters:

  • model (dict) – the model that needs to be downloaded

  • output_dir (str) – the directory in which the files will be saved

Returns:

the metadata of the model

Return type:

dict

SAClient.run_prediction(project, images_list, model)#
This function runs smart prediction on given list of images from a given project

using the neural network of your choice

Parameters:

  • project (str or dict) – the project in which the target images are uploaded.

  • images_list (list of str) – the list of image names on which smart prediction has to be run

  • model (str or dict) – the name of the model that should be used for running smart prediction

Returns:

tuple of two lists, list of images on which the prediction has succeeded and failed respectively

Return type:

tuple

SAClient.search_models(name=None, type_=None, project_id=None, task=None, include_global=True)#

Search for ML models.

Parameters:

  • name (str) – search string

  • type_ (str) – ml model type string

  • project_id (int) – project id

  • task (str) – training task

  • include_global (bool) – include global ml models

Returns:

ml model metadata

Return type:

list of dicts