Gustavo de Rosa

Using ```tf.nn.bidirectional_dynamic_rnn``` and concatenating its outputs ```outputs = tf.concat([outputs[0], outputs[1]], 2)``` should perform better. It will create the graph on-the-fly and allow for variable batch sizes feeding.

Looks like that the process gets killed due to torch.distributed.launch/run timeout of 30 minutes? (https://pytorch.org/docs/stable/distributed.html#torch.distributed.init_process_group) I had the same problem, where my job would be stopped when using DDP due...

That's amazing @dvlshah! Thank you so much for doing it!

These are some messy modifications but it should hold for your BERT case: ```import torch import torch.nn as nn from torch.autograd import Variable from collections import OrderedDict import numpy as...

The problem is that on VGG16 the last layer prior to the top classification ones is a Pooling Layer, therefore you are basically getting the output of this layer, which...

Just use the whole file instead of only the example: https://github.com/fchollet/deep-learning-models/blob/master/resnet50.py And I can confirm it is working with python3 as I'm using it and had no problems so far.

It is fixed on main repository: https://github.com/fchollet/keras/blob/master/keras/applications/imagenet_utils.py And yes, you were right!

I'm also having the same problem as you! No matter which network I use, I always get the same predictions for different images. This only happens when I try to...

It's based on this format: {"0": ["n01440764", "tench"], "1": ["n01443537", "goldfish"], "2": ["n01484850", "great_white_shark"], "3": ["n01491361", "tiger_shark"], ..., }, where it stands for "class id": ["imagenet_reference_number", "real_class_name"]. To solve your...

Just use this straight link: https://github.com/fchollet/deep-learning-models/releases/download/v0.1/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5