Many models have identical implementations of prune_heads it would be nice to store that implementation as a method on PretrainedModel and reduce the redundancy.

Hi, is there any plan to provide a tutorial of showing an example of employing the Transformer as an alternative of RNN for seq2seq task such as machine translation?

For some reason, when I open the web document, real_a and fake_b are matching, but the real_b is from another image; however in the images folder the images are correct. Does someone know why does this happen?

Example scripts contains some dependencies not listed for Horovod, and in some cases require datasets without explaining how to obtain them. We should provide a README file along with a set of packages (requirements.txt) for successfully running the examples.

I tried selecting hyper parameters of my model following "Tutorial 8: Model Tuning" below:
https://github.com/flairNLP/flair/blob/master/resources/docs/TUTORIAL_8_MODEL_OPTIMIZATION.md

Although I got the "param_selection.txt" file in the result directory, I am not sure how to interpret the file, i.e. which parameter combination to use. At the bottom of the "param_selection.txt" file, I found "

It would be nice to separately log model computed loss from regularization loss in tensorboard. Involves minor changes to the Trainer.

❓ Questions and Help

I followed the fine-tuning example described in here: https://github.com/pytorch/fairseq/blob/master/examples/mbart/README.md
However I didn't manage to reproduce the results described in the paper for EN-RO translation.

How to reproduce fine tuning with mbart?

1. Can you clarify where did you get the data and how did you preprocess it for training in more de

The documentation about edge orientation is inconsistent. In the Creating Message Passing Networks tutorial, the main expression says that e𝑖,𝑗 denotes (optional) edge features from node 𝑖 to node 𝑗., the attached expression also suggests it. However, in documentation to MessagePassing.message(), the documentation says Constructs messages from node 𝑗 to node 𝑖 (this is actually true).

I

Describe the bug
I try to run tensorboardX/examples/demo_graph.py for jupyter notebook (launched by anaconda navigator) and I get the error seen at Additional context.

I just copy paste the code to notebook from Github.

Minimal runnable code to reproduce the behavior
class SimpleModel(nn.Module):
def init(self):
super(SimpleModel, self).init()

Let's enable loading weights from a URL directly

Option 1:

Automate it with our current API

Trainer.load_from_checkpoint('http://')

Option 2:

Have a separate method

Trainer.load_from_checkpoint_at_url('http://')

Resources

We can use this under the hood:
(https://pytorch.org/docs/stable/hub.html#torch.hub.load_state_dict_from_url)

Any tho

Describe the bug
The test_torch_tanh_approx test fails intermittently during automated PR testing.

To Reproduce
Run the test (or full suite) until it fails.

Screenshots

2020-04-24T13:00:44.9923763Z method = 'sigmoid', prec_frac = 3, tolerance = 0.1
2020-04-24T13:00:44.9925054Z workers = {'alice': <VirtualWorker id:alice #objects:112>, 'bob': <VirtualWorker id:bob #

Can someone explain how dimensions of the anchor boxes are calculated from anchor ANCHOR_SCALES and ANCHOR_RATIOS? How do they relate to generating 1:1, 1:2 or 2:1 aspect ratio anchor boxes with box areas 128^2, 256^2 as mentioned in the Faster RCNN paper?

Sorry to bother you.

Platform (like ubuntu 16.04/win10): Windows 10
Python version: 3.7.4, mmdnn==0.2.5

Running scripts: mmconvert -f caffe -df keras -om test

I know that this command is not supposed to run without passing an input file, but the error message is incorrect and should be improved:

mmconvert: error: argument --srcFramework/-f: invalid choice: 'None' (choose from 'caffe', 'caffe2', 'cn

when I change the size of input image, may I need to change the value of "g_conv_dim" and "d_conv_dim"

According to scipy, scipy.misc.toimage()

toimage is deprecated! toimage is deprecated in SciPy 1.0.0, and will be removed in 1.2.0. Use Pillow’s Image.fromarray directly instead.

which is used on line 46 of utils/visualizer.py is now a deprecated function under the newest scipy version. As a result this co

Hi! Great package!

Both NLTK and Spacy offer the option to install models from local files, as with:

pip install /Users/you/en_core_web_sm-2.2.0.tar.gz

Do you have any thoughts on adding this to stanza? This makes it easier to deploy in an environment where the resources for the cod

When positional encoding is disabled, the embedding scaling is also disabled even though the operations are independent:

https://github.com/OpenNMT/OpenNMT-py/blob/1.0.0/onmt/modules/embeddings.py#L48

In consequence, Transformer models with relative position representations do not follow the reference implementation which scales the embedding [by default](https://github.com/tensorflow/tensor