The Fine-Tuning Index / PEFT & LoRA / #53
wuwangzhang1216/abliterix
by wuwangzhang1216 · PEFT & LoRA · updated 28d ago
Automated alignment adjustment for LLMs — direct steering, LoRA, and MoE expert-granular abliteration, optimized via multi-objective Optuna TPE.
54
momentum
264
stars
44
forks
#53
rank
abliterationalignmentdecensoringgemmallmloramixture-of-expertsmodel-editingmoeoptunapeftpytorch
View on GitHub →