https://github.com/huggingface/transformers/blob/546dc24e0883e5e9f5eb06ec8060e3e6ccc5f6d7/src/transformers/models/gpt2/modeling_gpt2.py#L698

Assertions can't be relied upon for control flow because they can be disabled, as per the following:

$ python --help
usage: python [option] ... [-c cmd | -m mod | file | -] [arg] ...
...
-O     : remove assert and __debug__-dependent statem

Reproduction

1. What command or script did you run?
   I use ATSS to train on cityscapes dataset, and all the configurations are exactly the same as the official faster-rcnn. AP75 is always -1 at the time of evaluation.
2. Did you make any modifications on the code or config? Did you understand what you have modified?
3. What dataset did you use?
   cityscapes
   Environment
   sys.platform:

Proposed refactoring or deprecation

Conditionally configure syncbatchnorm in distributed plugins

Motivation

This issue is closely related to PyTorchLightning/pytorch-lightning#6977
Carrying forward discussion from PyTorchLightning/pytorch-lightning#9096
Related issue in PyTorch: pytorch/pytorch#48988

S

Change tensor.data to tensor.detach() due to
pytorch/pytorch#6990 (comment)
tensor.detach() is more robust than tensor.data.

New Operator

Describe the operator

Why is this operator necessary? What does it accomplish?

This is a frequently used operator in tensorflow/keras

Can this operator be constructed using existing onnx operators?

If so, why not add it as a function?

I don't know.

Is this operator used by any model currently? Which one?

Are you willing to contribute it?

Is your feature request related to a problem? Please describe.
I typically used compressed datasets (e.g. gzipped) to save disk space. This works fine with AllenNLP during training because I can write my dataset reader to load the compressed data. However, the predict command opens the file and reads lines for the Predictor. This fails when it tries to load data from my compressed files.