search

Sha-Lab / FEAT

The code repository for "Few-Shot Learning via Embedding Adaptation with Set-to-Set Functions"
MIT License
418 stars 84 forks source link

Command lines for training Protonet model #27

Closed gabrielhuang closed 4 years ago

gabrielhuang commented 4 years ago

Hi, thanks for sharing your code.

Could you provide command lines for training the Protonet models ? Or are the hyperparameters exacty the same as for FEAT ?

Thanks !

Han-Jia commented 4 years ago

Hi,

For ProtoNet w/ ResNet could be learned using the learning rate 0.0001, the other parameters are the same as FEAT.

gabrielhuang commented 4 years ago

Thank you, I will try that !

On Tue, May 12, 2020 at 8:40 AM Han-Jia Ye notifications@github.com wrote:

Hi,

For ProtoNet w/ ResNet could be learned using the learning rate 0.0001, the other parameters are the same as FEAT.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Sha-Lab/FEAT/issues/27#issuecomment-627317118, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB3P5RGYMY7L2VRCFZLRXCLRRE7TRANCNFSM4M6K3MMA .