summaryrefslogtreecommitdiffstats
path: root/training/strategy/dreambooth.py
Commit message (Collapse)AuthorAgeFilesLines
* Improved automation capsVolpeon2023-04-161-2/+2
|
* UpdateVolpeon2023-04-131-5/+2
|
* UpdateVolpeon2023-04-101-1/+1
|
* Fix sample gen: models sometimes weren't in eval modeVolpeon2023-04-101-14/+13
|
* UpdateVolpeon2023-04-101-1/+1
|
* UpdateVolpeon2023-04-081-1/+1
|
* FixVolpeon2023-04-071-4/+2
|
* Refactoring, fixed Lora trainingVolpeon2023-03-241-10/+7
|
* UpdateVolpeon2023-03-231-4/+4
|
* UpdateVolpeon2023-03-011-3/+3
|
* Fixed TI normalization orderVolpeon2023-02-211-6/+5
|
* Don't rely on Accelerate for gradient accumulationVolpeon2023-02-211-6/+0
|
* Remove xformers, switch to Pytorch NightlyVolpeon2023-02-171-4/+4
|
* Fix Lora memory usageVolpeon2023-02-071-2/+0
|
* Add LoraVolpeon2023-02-071-8/+27
|
* Restored LR finderVolpeon2023-01-201-3/+1
|
* Move Accelerator preparation into strategyVolpeon2023-01-191-1/+13
|
* UpdateVolpeon2023-01-171-3/+2
|
* UpdateVolpeon2023-01-171-4/+4
|
* UpdateVolpeon2023-01-171-2/+8
|
* Training updateVolpeon2023-01-161-3/+7
|
* Handle empty validation datasetVolpeon2023-01-161-1/+1
|
* Added Dreambooth strategyVolpeon2023-01-151-0/+183