You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In Abstrat part of this Paper, it wrotes "with no parameter updates or task-specific tem-plates"
I thought this project a new method for "Prompt Tuning" through meta-learning, aiming at providing a better Prompt/Instruction than regular In-Context Learning.
But in the code, in "model.do_train()", it updates the model's parameters by BP (loss.backward()) ? Is is still a type of FT (fine-tune) ?
If I change my base-LM to something HUGE like GPT-3 176B,it costs too much.
In Abstrat part of this Paper, it wrotes "with no parameter updates or task-specific tem-plates"
I thought this project a new method for "Prompt Tuning" through meta-learning, aiming at providing a better Prompt/Instruction than regular In-Context Learning.
But in the code, in "model.do_train()", it updates the model's parameters by BP (loss.backward()) ? Is is still a type of FT (fine-tune) ?
If I change my base-LM to something HUGE like GPT-3 176B,it costs too much.