Y
Hacker News
new
|
ask
|
show
|
jobs
by
RC_ITR
206 days ago
When the new pre-trained parameters come out in a new model generation, your old fine tuning doesn't apply to them.