Khue Le

> Ran into the same issue and looks like apart from removing the `jekyll-feed` gem, monkey patching is the only thing that works. Can confirm neither commenting out `- jekyll-feed`...

Hi @datumbox, I am updating my personal vision library to take into account the new API and I find that it is a bit inconvenient to work with, mostly because...

@datumbox Thanks a lot for the detailed reply! > Examples: > > ```python > # These all work: > model = models.__dict__['resnet50'](weights='DEFAULT') # Use ResNet50_Weights.DEFAULT > model = models.__dict__['resnet50'](weights='IMAGENET1K_V1') #...

@datumbox Is there any plan to add the weights pre-trained (with self-supervision) on ImageNet-21K for `vision_transformer`? (cc @sallysyw @YosuaMichael) Thanks.

@thwjoy Any updates on this, please? I got the same errors as @sadeepj :( > out of memoryCuda kernel failed. Error: out of memoryinvalid argumentinvalid argumentCuda kernel failed. Error: invalid...

@zhanghang1989 That's great! I am looking forward to the updates. Thanks for the prompt reply!

In addition, I got the following error when training on multiple GPUs: > RuntimeError: Expected to have finished reduction in the prior iteration before starting a new one. This error...

Any updates on this, @JulienSiems and @MiguelMonteiro ? I have tried putting Miguel's implementation on top a CNN but it did not seem to do anything. What I did is...

Hi Miguel, Let me answer your questions: 1) No I did not train anything. What I did is loading the pretrained FCN weights into the models, i.e. only the FCN...

For debugging I tried to print the energy at each iteration, but failed :( I tried this code: ``` energy = tf.reduce_sum(tf.multiply(tf.subtract(0.5*pairwise, unaries), x)) print_energy_keras = keras.backend.print_tensor(energy, message="energy: ") print_energy_tf...