| Component name | Description |
|---|---|
| str_lower | component lowercasing inputs |
| nltk_moses_tokenizer | For tokenization of texts (inputs for prediction, correct texts for training) |
| spelling_error_model | Model for generation of hypotheses |
| KenlmElector | For selection of the best hypothesis |
Last active
July 1, 2019 09:32
-
-
Save acriptis/56a4471413de6e4504165ec022831603 to your computer and use it in GitHub Desktop.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment