Age | Commit message (Collapse) | Author |
|
Changelog minor correction
|
|
|
|
|
|
Add PR refs to changelog
|
|
|
|
|
|
Add extra noise param for img2img operations
|
|
|
|
|
|
many prompts/negative prompts there are #12509
|
|
rather than in settings
|
|
xyz_grid: support refiner_checkpoint and refiner_switch_at
|
|
Update SD XL commit hash
|
|
0.0.
|
|
resolution
|
|
|
|
Fix typo in shared_options.py
|
|
separate Extra options
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
unperdictable -> unpredictable
|
|
|
|
Add RES sampler and reorder the sampler list
|
|
Fix MHA error with ex_bias and support ex_bias for layers which don't have bias
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Fixes for `git checkout`, MPS/macOS fixes and optimizations
|
|
Fix typo in launch_utils.py
|
|
|
|
existance -> existence
|
|
add new callback for scripts to be used before processing
|
|
|
|
|
|
|
|
|
|
|
|
For MPS, using a tensor created with `torch.empty()` can cause `torch.baddbmm()` to include NaNs in the tensor it returns, even though `beta=0`. However, with a tensor of shape [1,1,1], there should be a negligible performance difference between `torch.empty()` and `torch.zeros()` anyway, so it's better to just use `torch.zeros()` for this and avoid unnecessarily creating issues.
|