DORA
Weight-decomposed low-rank adaptation
Episodes
-
#6160: Where Do Tiny Speech Rules Live? A Ladder From Regex to LoRAA regex notebook, a WFST grammar, or a fine-tune? Where do tiny speech-rewrite rules actually belong in a 600M-parameter model? -
#5520: Keeping a Fine-Tune Alive Across Model ReleasesDaniel wants to fine-tune DeepSeek Flash 4.1 on edited podcast scripts. The hard part isn't training — it's surviving the next release. -
#5410: Adapters: 102KB That Reshapes a 403GB ModelA 102KB adapter file changes how a 403GB base model behaves — without ever merging into it. Here's how model adapters actually work. -
#2525: Who Actually Reads Academic Journals?Half of all papers are read by nobody but the author and reviewers. So why do 300,000 journals exist? -
#1700: Can LLMs Learn Continuously Without Forgetting?We explore a new approach: micro-training updates every few days to keep AI knowledge fresh without constant web searches.