What's Changed
- fix: multiword broken by @stephantul in #317
- feat: refactor training by @stephantul in #316
- fix: upgrade skeletoken by @stephantul in #318
- fix: Incorrect added token can cause issues when adding token as multiword token by @stephantul in #319
- fix: don't set weights for non-quantized models by @stephantul in #321
- fix: correctly set weights when w is initialized by @stephantul in #322
- feat: convert output to float in distill by @stephantul in #320
- clean up post-processing code by @stephantul in #323
- chore: fix pre-commit by @stephantul in #326
- chore: add exclude-newer by @stephantul in #328
- feat: add frozen-weights option to training by @stephantul in #327
- fix: small typing issue by @stephantul in #329
- chore: Update email by @Pringled in #331
- feat: add trusted publisher by @stephantul in #332
Full Changelog: v0.8.1...v0.8.2