v1.41.0 - 2023-09-22
- Update diffusers to 0.21.2
v1.40.0 - 2023-08-23
- Add test for stable diffusion xl
- Add stable diffusion xl
- Update diffusers to 0.20.0
v1.39.0 - 2023-07-19
- Update diffusers to 0.18.2
- Update python to 3.11
v1.38.0 - 2023-06-20
- Update diffusers to 0.17.1
v1.37.0 - 2023-05-29
- Update xformers to 0.0.20
v1.36.0 - 2023-05-19
- Update diffusers to 0.16.1
v1.35.0 - 2023-04-15
- Add test for stable unclip variations
- Add stable unclip variations pipeline
- Add vae slicing to existing test
- Add vae slicing for image batches
- Add filename normalization to existing test
- Update torch to v2.0.0
- Temporarily suppress pipeline UserWarning
- Sanitize and truncate filenames
v1.34.0 - 2023-03-17
- Add test for custom model and vae tiling
- Add vae tiling for high resolution images
- Update diffusers, onnx, and safetensors
v1.33.0 - 2023-02-25
- Update diffusers to 0.13.1
- Ensure booleans do not consume other args
v1.32.0 - 2023-02-20
- Ensure onnx option is tested
- Split onnx into separate option
v1.31.0 - 2023-02-17
- Add test for instruct pix2pix
- Add instruct pix2pix pipeline
- Update transformers to 4.26.1
- Update onnxruntime to 1.14.0
- Add full-sized image for testing
v1.30.1 - 2023-02-03
- Switch xformers to 0.0.16 stable
v1.30.0 - 2023-01-28
- Update diffusers to 0.12.1
- Reorder arguments alphabetically
- Rename option names in tests
- Rename certain options for ease-of-use
- Download test image since img folder is gone
- Rename option names in depth diffusion test
v1.29.0 - 2023-01-26
- Add test for depth-guided diffusion
- Add depth-guided stable diffusion
v1.28.0 - 2023-01-26
- Upgrade diffusers to 0.12.0
- Suppress CLIPFeatureExtractor warning
v1.27.0 - 2023-01-16
- Automatically publish new versions
- Update unstable xformers to 0.0.16rc425
v1.26.0 - 2023-01-13
- Support pulling image from ghcr
- Publish docker pipeline to ghcr
- Simplify ghcr publish action
- Move ghcr url into variable
v1.25.1 - 2023-01-10
- Update safetensors and unstable xformers
- Add numpy 1.23.5 to fix float errors
v1.25.0 - 2022-12-26
- Add test for scheduler option
- Allow different schedulers to be used
- Update torch to 1.13.1 and unstable xformers
v1.24.0 - 2022-12-20
- Add tests for stable diffusion 2.0 and 2.1
- Update diffusers to 0.11.1
- Switch to python slim and halve image size
v1.23.0 - 2022-12-14
- Add memory efficient transformers
- Add test for image upscaling
- Ensure default test produces output
v1.22.0 - 2022-12-12
- Add upscaler pipeline
- Update diffusers to 0.10.2
v1.21.0 - 2022-12-09
- Update diffusers to 0.10.1
- Update diffusers to 0.10.0
- Update transformers to 4.25.1
- Remove use of autocast
- Remove unused arguments from pipeline
- Return after setting gpu arg
- Removed unused imports
v1.20.0 - 2022-11-29
- Update diffusers to 0.9.0
- Update tensorflow to 2.11.0
- Ensure skip option works on onnx pipeline
v1.19.1 - 2022-11-22
- Test image to image pipeline
- Do not pass gpu arg if using cpu pipeline
v1.19.0 - 2022-11-17
- Add test for cpu pipeline
- Add cpu pipeline using onnx
- Test image with different dimensions
v1.18.0 - 2022-11-11
- Add diffusion inpainting
v1.17.0 - 2022-11-08
- Add image-to-image diffusion
- Add input folder for image-to-image diffusion
- Call run command directly from tests
- Ensure output folder is always created
v1.16.0 - 2022-11-07
- Add standard tests for all options
- Upgrade diffusers to 0.7.2
v1.15.0 - 2022-11-04
- Add negative prompts
- Switch to using num_images_per_prompt
v1.14.0 - 2022-11-03
- Upgrade diffusers to 0.7.0
v1.13.0 - 2022-11-02
- Upgrade torch to 1.13.0 and cuda to 11.7
v1.12.0 - 2022-11-01
- Update transformers to v4.24.0
- Always use torch instead of tensorflow
v1.11.1 - 2022-10-26
- Split pipeline and inference
v1.11.0 - 2022-10-21
- Make model configurable
v1.10.0 - 2022-10-19
- Upgrade to diffusers 0.6.0
- Switch output and image names
v1.9.0 - 2022-10-14
- Upgrade to latest huggingface releases
v1.8.1 - 2022-10-13
- Remove cuda stubs to prevent errors
v1.8.0 - 2022-10-12
- Upgrade to diffusers 0.4.2
v1.7.0 - 2022-10-07
- Upgrade to diffusers 0.4.1
- Pin dependencies using requirements.txt
v1.6.0 - 2022-09-14
- Add attention slicing
- Use tagged version of tensorflow
v1.5.0 - 2022-09-08
- Specify user access token at command line
v1.4.0 - 2022-09-05
- Add pipeline iteration
- Move image name out of loop
v1.3.0 - 2022-09-02
- Support skipping safety checker
v1.2.2 - 2022-09-01
- Rename iso date time function
- Prevent errors when file name is too long
v1.2.1 - 2022-08-29
- Allow full range of random seeds
v1.2.0 - 2022-08-26
- Support half-sized (float16) tensors
- Update scale argument description
- Remove unused sys import
v1.1.1 - 2022-08-25
- Double quote to prevent globbing
- Only output two digits of precision
v1.1.0 - 2022-08-24
- Add a subset of txt2img.py options
- Include model params in image filename
- Use enumerate instead of index
- Remove unnecessary image library import
v1.0.0 - 2022-08-22
- Initial release