GitLab

Custom schedulers are currently initiated by wrapping Pytorch's LambdaLR
class and passing a method of the wrapping class to the __init__
function of LambdaLR. This approach is not appropriate for several
reasons:

1. one does not need to define a class when it only defines a
__init__() method;
2. instantiating the parent class by passing a method of the child class
creates a cyclical reference which leads to memory leaks. See issues #1742 and #1134.

In this commit we replace the wrapper classes with functions that
instantiate `LambdaLR` with a custom learning rate function. We use a
closure to specify the parameter of the latter. We also do a bit of
renaming within the function to explicit the behaviour and removed
docstrings that were subsequently not necessary.

converted from https://github.com/openai/gpt-2-output-dataset/tree/master/detector

Co-Authored-By: Lysandre Debut <lysandre.debut@reseau.eseo.fr>
Co-Authored-By: Jong Wook Kim <jongwook@nyu.edu>
Co-Authored-By: Jeff Wu <wuthefwasthat@gmail.com>

cf. https://docs.pytest.org/en/latest/example/simple.html

model forwards can take an inputs_embeds param

add progress bar to convert_examples_to_features

Add speed log to examples/run_squad.py

Fixing mode in evaluate during training

It takes considerate amount of time (~10 min) to parse the examples to features, it is good to have a progress-bar to track this

Fix #1623

Co-Authored-By: Santosh-Gupta <San.Gupta.ML@gmail.com>

Add common getter and setter for input_embeddings & output_embeddings

Fix token order in xlnet preprocessing for SQuAD

Fix https://github.com/huggingface/transformers/issues/1673