Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
|
|
|
|
|
|
for OSX
|
|
|
|
|
|
added support for extensions
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
test_resolve_conflicts
|
|
|
|
|
|
|
|
embedding before the generation using text
|
|
|
|
|
|
|
|
|
|
* Add cross-attention optimization from InvokeAI (~30% speed improvement on MPS)
* Add command line option for it
* Make it default when CUDA is unavailable
|
|
filename that people who use zip instead of git clone will have
|
|
|
|
Comma backtrack padding
|
|
|
|
|
|
|
|
Unlimited tokens actually work now. Works with textual inversion too. Replaces the previous not-so-much-working implementation.
|
|
|
|
|
|
|
|
|
|
regarding cross attention optimizations
|
|
|
|
|
|
|
|
|
|
|
|
xformers attention
|
|
|
|
|
|
before the final AND will be dropped
|
|
|
|
|