Do fuzzy matching (something like fuzzywuzzy maybe) to see if the the words line up (allowing for wrong words). You'll need to work out how to use scoring to work out how well aligned the two lists are. Source: over 1 year ago

Convert the original lines to full furigana and do a fuzzy match. (For reference, the original line is 貴方がこれまでに得てきた力、存分に発揮してくださいね。) You can do a regional search using the initial scene data (E60) first, and if the confidence is low, go for a slower full search. Source: over 1 year ago

It's now known as "thefuzz", see https://github.com/seatgeek/fuzzywuzzy. Source: about 2 years ago

You can have a look at this library to use fuzzy search instead of looking for plaintext muck: https://github.com/seatgeek/fuzzywuzzy. Source: over 2 years ago

To deal with comparing the string, I found FuzzyWuzzy ratio function that is returning a score of how much the strings are similar from 0-100. Source: almost 3 years ago

Many settings affect the outputs in interesting ways, but that's half the fun. These LoRAs are very lightly trained; more training may or may not help. The competitions are also performed using zero-shot text guessing, and if Facebook said it, you can bet that's actually Meta AI saying it, and they are leaders in the field. Source: over 1 year ago

You should look at the entire niche of MAE-related papers, that's quite exciting, and the neuroscience-inspired stream of stuff like Barlow Twins. As well, the official Facebook AI blog is surprisingly good coverage of much of the interesting un/semi-supervised DL research FAIR does, and worth going through. Source: almost 2 years ago