fine-tuning — Taking a trained model and training it a bit more on your own examples so it gets better at one specific job.
Why it matters
Selective layer training captures much of full fine-tuningTaking a trained model and training it a bit more on your own examples so it gets better at one specific job.Full definition →'s benefit at a fraction of the compute, and the frozen low-signal layers preserve knowledge that full-weight updates tend to erase.
