nlpaug
nlpaug copied to clipboard
makcedward
Data augmentation for NLP
1 aug = naw.WordEmbsAug(model_type='word2vec', model_path=model_path, action="insert") 2 augmented_text = aug.augment(text) 3 print("Original:") 4 print(text) 5 print("Augmented Text:") [/usr/local/lib/python3.7/dist-packages/nlpaug/augmenter/word/word_embs.py](https://localhost:8080/#) in __init__(self, model_type, model_path, model, action, name, aug_min, aug_max, aug_p, top_k, n_gram_separator,...
In the notebook with examples https://github.com/makcedward/nlpaug/blob/master/example/textual_augmenter.ipynb nas.AbstSummAug used with num_beam parameter which doesn't exist
I am trying to implement the word2vec embedding but I get 'Word2VecKeyedVectors' object has no attribute 'index_to_key' error, I implemented the code just as it is in the repositorty, how...
Hi, I had some problems when using word embeddings models from other languages due to the file format. Currently, we have only three options of models with fixed format parameters...
error/warning :Input length of input_ids is 1836, but ``max_length`` is set to 1200. This can lead to unexpected behavior. You should consider increasing ``config.max_length`` or ``max_length``. I am getting this...
Hello I have a similar issue that someone else asked about. I have a dataframe with text column, and classes column. I would like to augment the text column based...
I think the token [UNK] used for tokens unknown to model interferes with the use of the unknown token to temporarily replace provided stopwords. In example 1, there is one...
How to generate more than 1 augmented samples using Back-Translation method? Current implementation of `BackTranslationAug` can only support generating a single text, but actually we can change the decoding strategy...
I am trying to use `ContextualWordEmbsAug` with a custom BERT model. There are two problems when trying to use a custom transformer that I have trained using the HuggingFace API....
Hi, I added Electra to the list of models that can be used for Contextual Word Embeddings Augmentation. Electra has the same special tokens as BERT, so really just copied...
