v0.1.7

What's Changed

• Faster Training by @cloneofsimo in #178
• v0.1.7 by @cloneofsimo in #180

Full Changelog: v0.1.6...v0.1.7

Release v0.1.6

What's Changed

• accept training filenames with capitalized extensions by @zeke in #169
• Add inpainting support to pivotal tuning by @levi in #152
• fix "local variable 'unet_path' referenced before assignment" by @arceus-jia in #175
• LoRA Manager for managing multiple loras with different scaling parameters & tokens by @cloneofsimo in #172
• v0.1.6 by @cloneofsimo in #171

New Contributors

• @zeke made their first contribution in #169
• @arceus-jia made their first contribution in #175

Full Changelog: v0.1.5...v0.1.6

I've worked with team @replicate to improve this repo capable of various productions! Thank you so much!

Release v0.1.5

What's Changed

• Support prefix text for blip captioning by @levi in #166
• allow jpeg files in addition to jpg files by @anotherjesse in #165
• v0.1.5 by @cloneofsimo in #168

New Contributors

• @anotherjesse made their first contribution in #165

Full Changelog: v0.1.4...v0.1.5

Release v0.1.4

What's Changed

• Fix dates in readme by @levi in #162
• v0.1.4 by @cloneofsimo in #164

Now output CLIP alignment scores during lora tuning step, mitigated center-crop by default error, saves lora with extra argument name.

Full Changelog: v0.1.3...v0.1.4

Release v0.1.3

What's Changed

• Include mediapipe package in requirements.txt by @levi in #158
• Update web demo link by @hysts in #157
• LoRA Join operation by @cloneofsimo in #160
• v0.1.3 by @cloneofsimo in #161

New Contributors

• @levi made their first contribution in #158
• @hysts made their first contribution in #157

Full Changelog: v0.1.2...v0.1.3

Release v0.1.2

What's Changed

• Feature/ Better LoRA : Dropout, Conv2d by @cloneofsimo in #133
• Remove useless model_id declaration in svd_distill by @wsxiaoys in #136
• Avoid using python 3.9 feature ( |= for dict ) so cli_pt_to_safetenso… by @wsxiaoys in #137
• Basic dataset pipelines by @cloneofsimo in #139
• Update README.md by @JackLangerman in #144
• SVD update with conv support + LoRA add, weight update following recent updates by @cloneofsimo in #140
• v0.1.2 by @cloneofsimo in #135

New Contributors

• @wsxiaoys made their first contribution in #136
• @JackLangerman made their first contribution in #144

Full Changelog: v0.1.1...v0.1.2

Release v0.1.1

What's Changed

• Hotfixes/docs and minor adj by @cloneofsimo in #122
• v0.1.1 by @cloneofsimo in #130

Full Changelog: v0.1.0-alpha...v0.1.1

Release v0.1.0

Better Pivotal Tuning, CLIP evaluation, SVD distillation!

Many tricks / features went into this release :

1. Extended Latent textual inversion
2. Pivotal tuning with stage-learning rate schedule
3. Norm-prior Bayesian learning
4. Mask conditioned score estimation (+ in-pipeline face detector)
5. Sub-1 denoising
6. CLIP score (text alignment score, image alignment score) evaluater
7. xformer support (not in lorpt CLI yet)
8. wandb logging
9. Better Caption dataset
10. Distilling fully trained model into LoRA : SVD distillation CLI

What's Changed

• fix(missing-var): h_flip by @oscarnevarezleal in #86
• BetterPTI: Support pivotal tuning with multiword latent by @cloneofsimo in #91
• Other Acceleration tricks by @cloneofsimo in #93
• fix : save lora 16 precision by @cloneofsimo in #97
• feat : svd distillation with CLI by @cloneofsimo in #98
• Add xformers to training scripts by @hafriedlander in #103
• Make safetensors properly optional, and support storing TI by @hafriedlander in #101
• Feat/inversion pp by @cloneofsimo in #104
• image transforms by @ethansmith2000 in #115
• Feat/allsave by @cloneofsimo in #119
• v 0.1.0 by @cloneofsimo in #88

New Contributors

• @timh made their first contribution in #11
• @AK391 made their first contribution in #20
• @2kpr made their first contribution in #25
• @DavidePaglieri made their first contribution in #39
• @hdeezy made their first contribution in #48
• @hafriedlander made their first contribution in #73
• @milyiyo made their first contribution in #75
• @laksjdjf made their first contribution in #76
• @oscarnevarezleal made their first contribution in #86
• @ethansmith2000 made their first contribution in #115

Full Changelog: https://github.com/cloneofsimo/lora/commits/v0.1.0-alpha