deep-residual-networks icon indicating copy to clipboard operation
deep-residual-networks copied to clipboard

Published 20 hours ago • verified publisher icon KaimingHe

Problem when finetuning with ResNet caffemodel

Open gtlim opened this issue 9 years ago • 2 comments

Hi, I was trying to finetune residual network using ResNet caffemodel But it does not work when it goes to UpgradeNetInput(), and keep tells me segment fault Is there some thing wrong with my solver.prototxt? or my train_val.sh?

// this is my solver net: "/v8/eccv/residual/train_val_50.prototxt" test_iter: 1000 test_interval: 1000 base_lr: 0.1 lr_policy: "step" average_loss: 40 stepsize: 320000 gamma: 0.96 display: 20 max_iter: 450000 momentum: 0.9 weight_decay: 0.0001 snapshot: 10000 snapshot_prefix: "/v8/dummy" solver_mode: GPU

// this is my train_val.sh /v8/eccv/caffe/build/tools/caffe train --gpu=all
--solver=/v8/eccv/residual/solver.prototxt
--weights=/ResNet-50.caffemodel

sorry about poor english

gtlim avatar Mar 01 '16 09:03 gtlim

can you tell me where can i download the resnet caffemodel for finetuning

swordlidev avatar Jul 27 '16 07:07 swordlidev

https://github.com/KaimingHe/deep-residual-networks

On this page you can find download links for model files

gtlim avatar Jul 27 '16 07:07 gtlim