CHANGE.md

NLPAUG Change Log

1.1.10 Dec 23, 2021

• KeywordAug supports Turkish
• Fix FrequencyMasking time range
• Remove unnecessary printout
• [Rollback ContextualWordEmbsForSentenceAug and AbstSummAug to use custom transformers API to reduce execution time]

1.1.9 Dec 1, 2021

• ReservedAug supports generating all combinations
• Rollback to use native HuggingFace API from Huggingface pipeline to solve slow performance issue
• Added description to explain the model of WordEmbsAug is custom class
• Change random behavior to increase more augmentation samples
• Fix SpeedAug random factor issue

1.1.8, Oct 18, 2021

• Added RandomSentAug
• Added skip_check parameter for WordEmbsAug
• OCRAug support customer mapping/ json file
• Improve slow loading word2vec issue
• Solve transformers comparability issue

1.1.7, Jul 20, 2021

• Fixed mising document bug

1.1.6, Jul 16, 2021

• Fixed the missing library dependency issue

1.1.5, Jul 15, 2021

• Added LambadaAug(https://arxiv.org/pdf/1911.03118.pdf) under sentencen augmenter group
• ContextualWordEmbsAug, BackTranslationAug, ContextualWordEmbsForSentenceAug and AbstSummAug support batch model generation.

1.1.4, Jun 20, 2021

• Fix performance issue when using single thread
• Fix wordnet interface change issue
• Adopt HuggingFace API for ContextualWordEmbsAug
• Change model soruce from Fairseq to HuggingFace for BackTranslationAug
• Adopt HuggingFace API for AbstSummAug

1.1.3, Mar 7, 2021

• Add multi language (DE, ES, FR, HE, IT, NL, PL and UK) support to KeyboardAug (Special thanks to Binoy Dalal)

1.1.2, Jan 4, 2021

• Add NormalizeAug (audio) and PolarityInverseAug (audio)
• Fix #191, #192, #194, Fix #196

1.1.1, Dec 12, 2020

• Fix #182, #184, #185, #187

1.1.0, Nov 13, 2020

• Fix audio augmenter's documentation error #158
• Introduced ReservedAug.
• Fix #161, #166, #167, #168, #175

1.0.1 Sep 25, 2020

• Added Spectrogram's Loudness augmenter #156

1.0.0 Sep 24, 2020

• Upgraded to use AutoModel and AutoTokeizer for ContextualWordEmbsAug, ContextualWordEmbsForSentenceAug and AbstSummAug. Fix #133, #105
• Refactoring audio and spectrogram augmenters
• Added LoudnessAug into spectrogram augmenters
• Support single forward data input for deep learning models (i.e. ContextualWordEmbsAug, BackTranslationAug, ContextualWordEmbsForSentenceAug, AbstSummAug). #146
• Fix missing re-assing model paramters (e.g. device) for deep learning model
• BackTranslation supports to load model from local #149
• Fix device parameter bug #150
• Deprecated include_detail feature

0.0.20 Aug 22, 2020

• Update MANIFECT file to include txt resource

0.0.19 Aug 22, 2020

• Add back English mispelling dictionary

0.0.18 Aug 21, 2020

• Fix PPDB model misloaded nltk module#144

0.0.17 Aug 20, 2020

• Enhance default tokenizer and reverse tokenizer#143
• Introduce Abstractive Summarization in sentence ausgmenter (Check out example from here)

0.0.16 Aug 10, 2020

• Fix #142

0.0.15 Aug 10, 2020

• Support crop action in RandomWordAug #126
• Fix #130
• Fix #132
• Fix #134
• Upgraded and verified torch (1.6.0) and transformers (3.0.2) libraies
• Add new Back Translation Augmenter #75 #102 #131

0.0.14 Apr 24, 2020

• Remove QWERTAug example (Replaced by KeyboardAug) [#110] (makcedward#110)
• Fix [#117] (makcedward#117), [#114] (makcedward#114), [#111] (makcedward#111), #105
• Support Change Log [#116] (makcedward#117)
• Fix typo [#123] (makcedward#123)
• Support accepting candidates in RandomCharAug [#125] (makcedward#125)

0.0.13 Feb 25, 2020

• Fix spectrogram tutorial notebook [#98] (makcedward#98)
• Fix RandomWordAug missed aug_max parameter [#100] (makcedward#100)
• Fix loading KeyboardAug model problem [#101] (makcedward#101)
• Fix performance issue when sampling candidate in ContextualWordEmbsAug and ContextualWordEmbsForSentenceAug #107

0.0.12 Feb 5, 2020

• ContextualWordEmbsAug supports bert-base-multilingual-uncased (for non English inputs)
• Fix missing library dependency #74
• Fix single token error when using RandomWordAug #76
• Fix replacing character in RandomCharAug error #77
• Enhance word's augmenter to support regular expression stopwords #81
• Enhance char's augmenter to support regular expression stopwords #86
• KeyboardAug supports Thai language #92
• Fix word casing issue #82

0.0.11 Dec 6, 2019

• Support color noise (pink, blue, red and violet noise) in audio's NoiseAug
• Support given background noise in audio's NoiseAug
• Support inject noise to portion of audio only in audio's NoiseAug
• Introduce zone, coverage to all audio augmenter. Support only augmented portion of audio input
• Add VTLP augmentation methods (Audio's augmenter)
• Adopt latest transformer's interface #59
• Support RoBERTa (including DistilRoBERTa) and DistilBERT (ContextualWordEmbsAug)
• Support DistilGPT2 (ContextualWordEmbsForSentenceAug)
• Fix librosa hard dependency #62
• Introduce optimize attribute ContextualWordEmbsForSentenceAug #63
• Optimize word selection for ContextualWordEmbsAug and ContextualWordEmbsForSentenceAug (Speed up around 30%)
• Add retry mechanism into ContextualWordEmbsAug insert action #68

0.0.10 Nov, 2019

• Add aug_max to control maximum number of augmented item
• Fix ContextualWordEmbsAug (for BERT) error when input is longer than max sequence length
• Add RandomWordAug Substitute action
• Fix ContextualWordEmbsAug error when no augmented data
• Support multi thread processing (for CPU only) to speed up the augmentation
• Fix KeyboardAug error #55

0.0.9### Sep 30, 2019

• Added Swap Mode (adjacent, middle and random) for RandomAug (character level)
• Added SynonymAug (WordNet/ PPDB) and AntonymAug (WordNet)
• WordNetAug is deprecated. Uses SynonymAug instead
• Introduce parameter n. Returning more than 1 augmented data. Changing output format from text (or numpy) to list of text (or numpy) if n > 1
• Introduce parameter temperature in ContextualWordEmbsAug and ContextualWordEmbsForSentenceAug to control the randomness
• aug_n parameter is deprecated. This parameter will be replaced by top_k parameter
• Fixed tokenization issue #48
• Upgraded transformers dependency (or pytorch_transformer) to 2.0.0
• Upgraded PyTorch dependency to 1.2.0
• Added SplitAug

0.0.8### Sep 4, 2019

• BertAug is replaced by ContextualWordEmbsAug
• Support GPU (for ContextualWordEmbsAug and ContextualWordEmbsForSentenceAug only) #26
• Upgraded pytorch_transformer to 1.1.0 version #33
• ContextualWordEmbsAug suuports both BERT and XLNet model
• Removed librosa dependency
• Add ContextualWordEmbsForSentenceAug for generating next sentence
• Fix sampling issue #38

0.0.7### Aug 21, 2019

• Add new augmenter (CropAug, LoudnessAug, MaskAug)
• QwertyAug is deprecated. It will be replaced by KeyboardAug
• Remove StopWordsAug. It will be replaced by RandomWordAug
• Code refactoring
• Added model download function for word2vec, GloVe and fasttext

0.0.6### Jul 29, 2019:

• Added new augmenter TF-IDF based word replacement augmenter(TfIdfAug)
• Added new augmenter Spelling mistake simulation augmenter(SpellingAug)
• Added new augmenter Stopword Dropout augmenter(StopWordsAug)
• Fixed #14

0.0.5### Jul 2, 2019:

• Fixed #3, #4, #5, #7, #10

0.0.4### Jun 7, 2019:

• Added stopwords feature in character and word augmenter.
• Added character's swap augmenter.
• Added word's swap augmenter.
• Added validation rule for #1.
• Fixed BERT reverse tokenization for #2.

0.0.3### May 23, 2019:

• Added Speed, Noise, Shift and Pitch augmenters for Audio

0.0.2### Apr 30, 2019:

• Added Frequency Masking and Time Masking for Speech Recognition (Spectrogram).
• Added librosa library dependency for converting wav to spectrogram.

0.0.1### Mar 20, 2019: Project initialization