Please provide a barebones "pick up and go" GPT-2 colab notebook for text generation, just like gpt-2-simple does

Environment:

1. Framework: PyTorch
2. Framework version: 1.3.1
3. Horovod version: 0.19.0
4. MPI version: 4.0.2
5. CUDA version: N/A
6. NCCL version: N/A
7. Python version: 3.7.5
8. OS and version: Mac OS 10.15.2
9. GCC version: 9.2.0

Checklist:

1. Did you search issues to find if somebody asked this question before? Yes
2. If your question is about hang, did you read [this d

While wanting to try out the tutorials, I noticed they were in markdown format, ".md".
As I am not aware of an easy tool to convert markdown to python, and I was unwilling to copy paste code blocks, I created a quick and dirty script to convert the .md to .py files.

Maybe it helps others as well.
md_to_py.py.zip

Describe the bug

Calling Predictor.get_gradients() returns an empty dictionary

To Reproduce
I am replicating the binary sentiment classification tasked described in the paper 'Attention is not Explanation ' (Jain and Wallace 2019 - https://arxiv.org/pdf/1902.10186.pdf).

My first experiment is on the Stanford Sentiment TreeBank Dataset. I need to measure the correlation between th

Several parts of the op sec like the main op description, attributes, input and output descriptions become part of the binary that consumes ONNX e.g. onnxruntime causing an increase in its size due to strings that take no part in the execution of the model or its verification.

Setting __ONNX_NO_DOC_STRINGS doesn't really help here since (1) it's not used in the SetDoc(string) overload (s

Hi, Is there any pretrained BART model for Japanese? If not, could you please explain the procedure to train new BART model for Japanese data from scratch?

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

I noticed when hovering over a graph which was created via SummaryWriter.add_scalar that each data point has Name: . . It would be nice feature to give each data point a description which would be displayed as Name.

This would be very helpful, e.g., when measuring the accuracy outliers could be easily identified with the corresponding input (e.g., name of image) for further study.

I

Similar to the tutorial on custom losses in SVI, we should have a tutorial on implementing custom MCMC kernels using the new MCMC API. Something simple like SGLD seems like a good starting point.

is it Grid Search can solve CASH problems with NNI , it seems that it is usually used for hyper-parameters optimization, have you guys have finished some revision for Grid Search for solving CASH problems.

about Cash problems can refer to ：microsoft/nni#1178

Describe the bug
This whole file seems to be largely un-tested by our unit testing suite.
https://github.com/OpenMined/PySyft/blob/master/syft/frameworks/torch/nn/rnn.py

Expected behavior
This file should have 100% test coverage.

Make sure to uncomment the codecov omit flag for this file OpenMined/PySyft#2896 (review)