summaryrefslogtreecommitdiffstats
path: root/pipelines
Commit message (Expand)AuthorAgeFilesLines
* UpdateVolpeon2023-03-251-28/+33
* Fixed Lora training perf issueVolpeon2023-03-241-2/+1
* UpdateVolpeon2023-03-231-1/+3
* UpdateVolpeon2023-03-071-1/+1
* UpdateVolpeon2023-03-061-5/+9
* More flexible pipeline wrt init noiseVolpeon2023-03-041-13/+44
* Added Perlin noise to trainingVolpeon2023-03-041-2/+2
* UpdateVolpeon2023-03-041-9/+2
* Pipeline: Perlin noise for init imageVolpeon2023-03-041-14/+9
* Pipeline: Improved initial image generationVolpeon2023-03-041-23/+26
* Changed init noise algorithmVolpeon2023-03-031-3/+11
* Removed offset noise from training, added init offset to pipelineVolpeon2023-03-031-56/+39
* UpdateVolpeon2023-02-131-32/+37
* Integrate Self-Attention-Guided (SAG) Stable Diffusion in my custom pipelineVolpeon2023-02-081-7/+162
* Add LoraVolpeon2023-02-071-2/+12
* Restored functional trainerVolpeon2023-01-151-8/+8
* WIP: Modularization ("free(): invalid pointer" my ass)Volpeon2023-01-141-1/+0
* Removed PromptProcessor, modularized training loopVolpeon2023-01-131-8/+12
* Code deduplicationVolpeon2023-01-131-23/+9
* Fixed aspect ratio bucketing; allow passing token IDs to pipelineVolpeon2023-01-081-10/+30
* Set default dimensions to 768; add config inheritanceVolpeon2022-12-261-4/+4
* Fix trainingVolpeon2022-12-201-9/+0
* Upstream patchesVolpeon2022-12-191-5/+17
* Add support for resume in Textual InversionVolpeon2022-12-131-19/+23
* Dreambooth: Support loading Textual Inversion embeddingsVolpeon2022-12-121-5/+8
* Support attention_mask of text encoderVolpeon2022-12-101-4/+12
* UpdateVolpeon2022-12-042-16/+9
* UpdateVolpeon2022-12-011-16/+11
* UpdateVolpeon2022-11-301-1/+26
* UpdateVolpeon2022-11-141-6/+1
* RefactoringVolpeon2022-11-141-104/+214
* UpdateVolpeon2022-11-141-11/+22
* UpdateVolpeon2022-11-031-2/+34
* UpdateVolpeon2022-11-011-2/+3
* New Euler_a schedulerVolpeon2022-10-261-14/+13
* UpdateVolpeon2022-10-241-0/+3
* Improved prompt handlingVolpeon2022-10-181-55/+17
* Implemented extended prompt limitVolpeon2022-10-171-41/+55
* Upstream updates; better handling of textual embeddingVolpeon2022-10-171-1/+1
* UpdateVolpeon2022-10-151-6/+2
* Removed aesthetic gradients; training improvementsVolpeon2022-10-141-46/+4
* Added support for Aesthetic GradientsVolpeon2022-10-131-3/+49
* Various updatesVolpeon2022-10-121-6/+5
* UpdateVolpeon2022-10-081-2/+3
* UpdateVolpeon2022-10-061-30/+5
* Small performance improvementsVolpeon2022-10-021-8/+10
* Fix img2imgVolpeon2022-10-021-35/+36
* More consistent euler_aVolpeon2022-10-021-11/+6
* WIP: img2imgVolpeon2022-10-021-16/+64
* Fix seed, better progress bar, fix euler_a for batch size > 1Volpeon2022-10-011-2/+2