automl / automl/ConfigurableOptimizer

Add option to finetune with full rank for LoRA

Open
#62 0 comments 0 reactions 1 assignee Claimed by @Neonkraft View on GitHub
Low priority Medium priority
Dominant language
Python
Stars
12
Forks
2
PR merge metrics
No merged PRs in 30d

Description

Instead of using lora_A and lora_B weights to approximate W, have a full rank W_new.

Contributor guide

No contributing guide indexed for this repository

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.