[HN Gopher] Show HN: WinkNLP delivers 600k tokens/second speed o... ___________________________________________________________________ Show HN: WinkNLP delivers 600k tokens/second speed on browsers (MBP M1) Author : sanjayaksaxena3 Score : 32 points Date : 2022-11-24 17:51 UTC (5 hours ago) (HTM) web link (github.com) (TXT) w3m dump (github.com) | ardkor wrote: | I remember using it. Pretty fast and accurate. My only concern | was its failure to deal with verbs used as adjectives, like | "bleeding edge" as VRB NN | Oras wrote: | The documentation is amazingly clear and thorough. | | Two questions: | | Is there a way to train custom models? | | Is there a way to convert models trained with spacy or PyTorch to | WinkNLP? ___________________________________________________________________ (page generated 2022-11-24 23:00 UTC)