tk_nn_classifier package

Subpackages

• tk_nn_classifier.classifiers package
  • Submodules
  • tk_nn_classifier.classifiers.graph_selector module
  • tk_nn_classifier.classifiers.spacy_classifier module
  • tk_nn_classifier.classifiers.tf_best_export module
  • tk_nn_classifier.classifiers.tf_classifier module
  • tk_nn_classifier.classifiers.tf_multi_feat_classifier module
  • tk_nn_classifier.classifiers.utils module
  • Module contents
• tk_nn_classifier.data_loader package
  • Submodules
  • tk_nn_classifier.data_loader.common_loader module
  • tk_nn_classifier.data_loader.csv_loader module
  • tk_nn_classifier.data_loader.data_reader module
  • tk_nn_classifier.data_loader.label_class_mapper module
  • tk_nn_classifier.data_loader.spacy_data_reader module
  • tk_nn_classifier.data_loader.tf_data_reader module
  • tk_nn_classifier.data_loader.tokenizer module
  • tk_nn_classifier.data_loader.trxml_loader module
  • tk_nn_classifier.data_loader.word_vector module
  • Module contents

Submodules

tk_nn_classifier.config module

The module to load the config file

tk_nn_classifier.config.get_default_config()

get the default config

tk_nn_classifier.config.load_config(config_file: str)

load the config file

params:

• config_file: the full path of the config (type: string)

output:

• dictionary object contains config key and config value pairs

tk_nn_classifier.config.load_config_from_dikt(config_dikt)

load the config from dikt

params:

• config_dikt: config options saved in dictionary

output:

• dictionary object contains config key and config value pairs

tk_nn_classifier.config.spacy_lang_model_consistency(config)

check the consistency of the language and pretrained model:

params:

• config: dict

output:

• updated input config or die on error

tk_nn_classifier.model module

Model class:

hub to integrate methods from both spaCy and Tensorflow frame

class tk_nn_classifier.model.Model(config)

Bases: object

build_and_train()

build_graph()

evaluate(test_data_path)

load(model_path=None)

predict_on_text(text)

process_with_saved_model(input)

save(output_path)

train(train_data=None, eval_data=None)

Module contents

Top-level package for tk_nn_classifier

tk_nn_classifier.define_logger(mod_name)

Set the default logging configuration

tk_nn_classifier.set_logging_level(level=30)

Change logging level