v0.2.1

[0.2.1]

• Fixed bug when decoding small texts

v0.2.0

[0.2.0]

• [Bugfix] Fix math in sentence filtering. This should address what behaved as overaggressive summarization when filtering around k50

v0.1.9

[0.1.9]

• Remove dependency on NLTK stopwords download

v0.1.8

[0.1.8]

• remove separate utils file
• improve compatibility for both 3.7 and 3.8
• flexible scipy version

v0.1.7

[0.1.7]

• Adds explicit scipy dependency

v0.1.6

[0.1.6]

• Adds optional intra-sentence summarization
• Patches max possible percentile for document summary

v0.1.5

[0.1.5] Adds single sentence summarization.

Default percentile remains 99, but the max percentile allowable for
a single sentence is derived from the mean of the token values.

v0.1.2

Fix relative import