qkeras icon indicating copy to clipboard operation
qkeras copied to clipboard

Published 20 hours ago • verified publisher icon google

(Re)quantize existing QKeras model with model_quantize

Open thesps opened this issue 4 years ago • 3 comments

Add to model_quantize the capability to accept an existing qkeras model and return a model with a different quantizer_config.

This could be used, for example, to train at higher precision then initialize a lower precision model with transfer_weights=True to speedup convergence of the lower precision model.

thesps avatar Apr 01 '20 14:04 thesps

Hello @thesps, thank you for the PR! Could you also add file to demonstrate your usage and tests? I want more information to review this PR.

zhuangh avatar Apr 02 '20 20:04 zhuangh

Hello @thesps, thank you for the PR! Could you also add file to demonstrate your usage and tests? I want more information to review this PR.

zhuangh avatar Apr 02 '20 20:04 zhuangh

Hi, there are conflicts in utils.py. Please solve if you want to proceed. thanks!

zhuangh avatar Apr 25 '20 22:04 zhuangh