|
|
|
|
|
|
by gjm11
1239 days ago
|
|
|
Cranking up the "legibility" I see plenty of perfectly respectable "x"s. Are you sure this is "not enough training data" rather than "actually, people writing quickly write pretty bad 'x's"? [EDITED to add:] Ah, oops, I see. You were talking specifically about _capital_ X. You are right and I was wrong: the model produces disastrous capital Xs even at high legibility settings. |
|
|