kaggle projects
Name | url | Repo name |
---|---|---|
imdb-dataset-of-50k-movie-reviews |
https://www.kaggle.com/lakshmi25npathi/imdb-dataset-of-50k-movie-reviews | Deep_IMDB |
Jigsaw Multilingual Toxic Comment Classification |
https://www.kaggle.com/c/jigsaw-multilingual-toxic-comment-classification | Multilingual_Toxic_Comment_Classification |
iMaterialist (Fashion) 2020 at FGVC7 |
https://www.kaggle.com/c/imaterialist-fashion-2020-fgvc7 | imaterialist_fashion_2020 |
Tweet Sentiment Extraction |
https://www.kaggle.com/c/tweet-sentiment-extraction | Tweet_Sentiment_Extraction |
Reddit selfposts classification |
https://www.kaggle.com/mswarbrickjones/reddit-selfposts | Reddit_selfposts_classification |
Contradictory, My Dear Watson |
https://www.kaggle.com/c/contradictory-my-dear-watson | Contradictory_classification |
google-quest-challenge |
https://www.kaggle.com/c/google-quest-challenge | google_quest_challenge |
utils description
- setup.py
- set_TPU: config TPU for tensorflow
- methdispatch: Adjustment of @singledispatchmethod usage to a python version lower than 3.8.
- text.py
- clean_text: Clean the text
- create_tfds_dataset: create tfds dataset
- textAugmentation: perform text augmentation
- get_most_similar: get most similar words
- get_MLM: return mask language model output (for example 'This is cool' -> 'This is pretty cool')
- get_synonym: return the synonym of a word
- get_translation: translate text using google API
- text_models.py
- BertInputs: create a BERT inputs
- XLNetInputs: create a XLNet inputs and build a model
- vision.py
- refine_masks: refine the mask to avoid overlapping
- resize_image: resize the image by the config settigs
apps description
- Text augmentation app
- Most Similar Words
- Synonyms
- Translation
- Mask Word Prediction