aboutsummaryrefslogtreecommitdiff
path: root/modules/sd_models.py
AgeCommit message (Collapse)Author
2023-05-02do not load wait for shared.sd_model to load at startupAUTOMATIC
2023-04-02Embed model merge metadata in .safetensors filespace-nuko
2023-03-28sort hypernetworks and checkpoints by nameAUTOMATIC
2023-03-28Merge pull request #8958 from MrCheeze/variations-modelAUTOMATIC1111
Add support for the unclip (Variations) models, unclip-h and unclip-l
2023-03-26overwrite xformers in the unclip model config if not availableMrCheeze
2023-03-25Add support for the Variations models (unclip-h and unclip-l)MrCheeze
2023-03-25Merge pull request #8780 from Brawlence/masterAUTOMATIC1111
Unload and re-load checkpoint to VRAM on request (API & Manual)
2023-03-23fix variable typocarat-johyun
2023-03-21Unload checkpoints on RequestΦφ
…to free VRAM. New Action buttons in the settings to manually free and reload checkpoints, essentially juggling models between RAM and VRAM.
2023-03-14fix an error loading Lora with empty values in metadataAUTOMATIC
2023-03-14Add view metadata button for Lora cards.AUTOMATIC
2023-02-19when existsw-e-w
2023-02-19fix auto sd download issuew-e-w
2023-02-15Add ".vae.ckpt" to ext_blacklistmissionfloyd
2023-02-14Download model if none are foundmissionfloyd
2023-02-05make it possible to load SD1 checkpoints without CLIPAUTOMATIC
2023-02-04fix issue with switching back to checkpoint that had its checksum calculated ↵AUTOMATIC
during runtime mentioned in #7506
2023-02-04Merge pull request #7470 from ↵AUTOMATIC1111
cbrownstein-lambda/update-error-message-no-checkpoint Update error message WRT missing checkpoint file
2023-02-04add --no-hashingAUTOMATIC
2023-02-01Update error message WRT missing checkpoint fileCody Brownstein
The Safetensors format is also supported.
2023-01-29support for searching subdirectory names for extra networksAUTOMATIC
2023-01-28fixed a bug where after switching to a checkpoint with unknown hash, you'd ↵AUTOMATIC
get empty space instead of checkpoint name in UI fixed a bug where if you update a selected checkpoint on disk and then restart the program, a different checkpoint loads, but the name is shown for the the old one.
2023-01-27add data-dir flag and set all user data directories based on itMax Audron
2023-01-27support detecting midas modelAUTOMATIC
fix broken api for checkpoint list
2023-01-27remove the need to place configs near modelsAUTOMATIC
2023-01-25Merge pull request #6510 from brkirch/unet16-upcast-precisionAUTOMATIC1111
Add upcast options, full precision sampling from float16 UNet and upcasting attention for inference using SD 2.1 models without --no-half
2023-01-25Add instruct-pix2pix hijackKyle
Allows loading instruct-pix2pix models via same method as inpainting models in sd_models.py and sd_hijack_ip2p.py Adds ddpm_edit.py necessary for instruct-pix2pix
2023-01-25Add option for float32 sampling with float16 UNetbrkirch
This also handles type casting so that ROCm and MPS torch devices work correctly without --no-half. One cast is required for deepbooru in deepbooru_model.py, some explicit casting is required for img2img and inpainting. depth_model can't be converted to float16 or it won't work correctly on some systems (it's known to have issues on MPS) so in sd_models.py model.depth_model is removed for model.half().
2023-01-19bring back short hashes to sd checkpoint selectionAUTOMATIC
2023-01-14fix bug with "Ignore selected VAE for..." option completely disabling VAE ↵AUTOMATIC
election rework VAE resolving code to be more simple
2023-01-14load hashes from cache for checkpoints that have themAUTOMATIC
add checkpoint hash to footer
2023-01-14update key to use with checkpoints' sha256 in cacheAUTOMATIC
2023-01-14change hypernets to use sha256 hashesAUTOMATIC
2023-01-14change hash to sha256AUTOMATIC
2023-01-11fix for an error caused by skipping initialization, for realsies this time: ↵AUTOMATIC
TypeError: expected str, bytes or os.PathLike object, not NoneType
2023-01-11possible fix for fallback for fast model creation from config, attempt 2AUTOMATIC
2023-01-11possible fix for fallback for fast model creation from configAUTOMATIC
2023-01-10add support for transformers==4.25.1AUTOMATIC
add fallback for when quick model creation fails
2023-01-10add more stuff to ignore when creating model from configAUTOMATIC
prevent .vae.safetensors files from being listed as stable diffusion models
2023-01-10disable torch weight initialization and CLIP downloading/reading checkpoint ↵AUTOMATIC
to speedup creating sd model from config
2023-01-09allow model load if previous model failedVladimir Mandic
2023-01-04use commandline-supplied cuda device name instead of cuda:0 for safetensors ↵AUTOMATIC
PR that doesn't fix anything
2023-01-04Merge remote-tracking branch 'Narsil/fix_safetensors_load_speed'AUTOMATIC
2023-01-04fix broken inpainting modelAUTOMATIC
2023-01-04find configs for models at runtime rather than when startingAUTOMATIC
2023-01-04helpful error message when trying to load 2.0 without configAUTOMATIC
failing to load model weights from settings won't break generation for currently loaded model anymore
2023-01-03call script callbacks for reloaded model after loading embeddingsAUTOMATIC
2023-01-02fix the issue with training on SD2.0AUTOMATIC
2022-12-31validate textual inversion embeddingsVladimir Mandic
2022-12-27Attempting to solve slow loads for `safetensors`.Nicolas Patry
Fixes #5893
generated by cgit v1.2.1 (git 2.18.0) at 2024-09-20 12:25:30 +0000