Initial Release
More Flexibility than the lora version
Different dataset
More consistent
Trained 1024 instead of 768 like the lora
Initial Release
More Flexibility than the lora version
Different dataset
More consistent
Trained 1024 instead of 768 like the lora