rethinking-bnn-optimization

rethinking-bnn-optimization copied to clipboard

kernel_initializer="glorot_normal"

4

In birealnet.py, I observe that kernel_initializer is set as "glorot_normal" for Conv2d; However, in QuantConv2D, the kernel_initializer is set as "glorot_normal". Kernel_initializer ="glorot_normal" means the kernel weights are all set...

puhan123

Which backend is chosen for training, CPU or GPU ?

4

I am running [this](https://github.com/plumerai/rethinking-bnn-optimization) experiment on Google Collaboratory. GPU is chosen as runtime. But, Training is taking longer to complete. Is CPU chosen explicitly in the code or Is this...

Sreerag-ibtl

how to transfer imagenet raw datasets into corresponding rfrecords?

3

Hi, when I ran this programe, it said 'cannot found imagenet2012-train.tfrecord-00823-of-01000' error, how can I transfer imagenet_train.tar and imagenet_val.tar into corresponding tfrecord?

HamLaertes

Running Nets with ImageNet2012

1

Hello everyone, I've been trying to run the Nets using the ImageNet dataset but the issue is that it cannot be downloaded from the default URL. I have the tars...

CuauSuarez

BinaryNet setting in section 5.2

1

Hi Koen, Thanks for the great work! I noticed to reproduce BinaryNet on CIFAR-10 experiment of section 5.2, you use the setting as followed: `class bop_sec52(default):` ` epochs = 500...

Sandbox3aster

PyTorch Implementation

2

Thanks for really good paper. Do you plan to implement it as a PyTorch version?

jack-kjlee