enamoria
/ unicode_to_ascii.py
Last active
December 16, 2018 10:39
Convert unicode character to the nearest (removed tone, accent, ... that feature a specific language) ASCII character. For ex: "Тгцмр" will be converted to "Tgtsmr", not Trump
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| # Source: https://stackoverflow.com/a/518232/4102479 | |
| # This is better approach, compared to unidecode, though unidecode provide acceptable result in almost every cases | |
| import unicodedata | |
| import string | |
| all_letters = string.ascii_letters + " .,;'" | |
| nb_letter = len(all_letters) | |
| print(all_letters, len(all_letters)) |
enamoria
/ keras_weighted_categorical_crossentropy.py
Created
August 24, 2018 04:05
— forked from wassname/keras_weighted_categorical_crossentropy.py
Keras weighted categorical_crossentropy
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| """ | |
| A weighted version of categorical_crossentropy for keras (2.0.6). This lets you apply a weight to unbalanced classes. | |
| @url: https://gist.github.com/wassname/ce364fddfc8a025bfab4348cf5de852d | |
| @author: wassname | |
| """ | |
| from keras import backend as K | |
| def weighted_categorical_crossentropy(weights): | |
| """ | |
| A weighted version of keras.objectives.categorical_crossentropy | |
NewerOlder