summaryrefslogtreecommitdiffstats
path: root/dreambooth.py
Commit message (Collapse)AuthorAgeFilesLines
* Euler_a: Re-introduce generator arg for reproducible outputVolpeon2022-10-271-2/+2
|
* New Euler_a schedulerVolpeon2022-10-261-3/+3
|
* Advanced datasetsVolpeon2022-10-241-26/+42
|
* UpdateVolpeon2022-10-241-0/+34
|
* Revert lat; fix skip attribute in datasetVolpeon2022-10-221-75/+86
|
* Training updateVolpeon2022-10-221-86/+75
|
* Add optional TI functionality to DreamboothVolpeon2022-10-211-35/+66
|
* Dreambooth: Added option to insert a new input token; removed Dreambooth PlusVolpeon2022-10-191-6/+42
|
* Updated Dreambooth trainingVolpeon2022-10-191-33/+71
|
* Adapted other scripts for new prompt processingVolpeon2022-10-181-7/+10
|
* Upstream updates; better handling of textual embeddingVolpeon2022-10-171-10/+16
|
* UpdateVolpeon2022-10-161-4/+6
|
* UpdateVolpeon2022-10-161-9/+32
|
* Removed aesthetic gradients; training improvementsVolpeon2022-10-141-5/+5
|
* Added support for Aesthetic GradientsVolpeon2022-10-131-10/+0
|
* Added TI+Dreambooth trainingVolpeon2022-10-131-5/+10
|
* Various updatesVolpeon2022-10-121-30/+23
|
* Added EMA support to Textual InversionVolpeon2022-10-101-11/+12
|
* Dreambooth: Add EMA supportVolpeon2022-10-101-6/+44
|
* Updated default paramsVolpeon2022-10-101-4/+4
|
* UpdateVolpeon2022-10-091-5/+2
|
* UpdateVolpeon2022-10-081-30/+45
|
* Fix small detailsVolpeon2022-10-071-22/+23
|
* Training: Create multiple class images per training imageVolpeon2022-10-071-6/+22
|
* UpdateVolpeon2022-10-061-1/+1
|
* Add Textual Inversion with class dataset (a la Dreambooth)Volpeon2022-10-041-15/+10
|
* Default sample steps 30 -> 40Volpeon2022-10-041-1/+1
|
* Fix: PromptsVolpeon2022-10-041-2/+6
|
* BugfixVolpeon2022-10-041-4/+4
|
* FixVolpeon2022-10-031-8/+2
|
* Dreambooth: Generate specialized class images from input promptsVolpeon2022-10-031-98/+70
|
* Small perf improvementsVolpeon2022-10-031-43/+46
|
* Added negative prompt support for training scriptsVolpeon2022-10-031-60/+21
|
* Fixed euler_a generator argumentVolpeon2022-10-031-8/+0
|
* Assign unused images in validation dataset to train datasetVolpeon2022-10-031-1/+2
|
* Use euler_a for samples in learning scripts; backported improvement from ↵Volpeon2022-10-031-6/+20
| | | | Dreambooth to Textual Inversion
* Added custom SD pipeline + euler_a schedulerVolpeon2022-09-301-4/+8
|
* Better sample file structureVolpeon2022-09-281-5/+5
|
* Batches of size 1 cause error: Expected query.is_contiguous() to be true, ↵Volpeon2022-09-281-20/+7
| | | | but got false
* Improved sample output and progress barsVolpeon2022-09-281-60/+44
|
* Dreamooth script: Fixed step calculation, added file logVolpeon2022-09-281-24/+24
|
* Better output dir structureVolpeon2022-09-271-1/+1
|
* Various improvements, added inference scriptVolpeon2022-09-271-46/+52
|
* Incorporate upstream changesVolpeon2022-09-271-8/+65
|
* Supply dataset CSV file instead of dir with hardcoded CSV filenameVolpeon2022-09-271-11/+11
|
* Freeze models that aren't trainedVolpeon2022-09-271-8/+8
|
* Added Dreambooth training scriptVolpeon2022-09-271-0/+825
generated by cgit v1.2.3-70-g09d2 (git 2.47.0) at 2024-11-20 06:56:13 +0000