V1.0.0-DEV0

What's Changed

• fix apostrophe for convert dataset by @Jean-Baptiste-Camps in #13
• fix segmentation in convert by @Jean-Baptiste-Camps in #18
• Avoid rich missing by @Jean-Baptiste-Camps in #22
• 1.0.0/vgsl by @PonteIneptique in #24

Full Changelog: v0.1.0...v1.0.0-dev0

Clean release

• Removed part of code that were dependent on the first approach that was taken.
• Better Scorer

Article release

0.3.0

Article update

Article draft

Accuracy model epochs' score in CSVs are wrong except for NoPos (was fixed at this point).

0.0.1a

Train Loss: 0.004 | Perplexity:   1.004 |  Acc.: 0.566 |  Lev.: 0.037 |  Lev. / char: 0.001
 Val. Loss: 0.066 | Perplexity:   1.069 |  Acc.: 0.585 |  Lev.: 0.272 |  Lev. / char: 0.009
 Test Loss: 0.057 | Perplexity:   1.059 |  Acc,: 0.586 |  Lev.: 0.235 |  Lev. / char: 0.008