PyTorch

Feature request

Is the addition of the 'OPTforSequenceClassification' class scheduled?
Is someone handling it?
When adding these functions, I wonder if it is possible to PR one by one, or if I have to PR all classes supported by other models.

Motivation

Added function of OPT class, which is being actively discussed recently

Your contribution

I personally use the forSequenceCla

Reporters do not apply sort_by_metric if metric is passed through tune.run.

I tried to pass metric argument to CLIReporter, however, I obtained this error:

raise ValueError(
                "You passed a `metric` or `mode` argument to `tune.run()`, but "
                "the reporter you are using was already instantiated with their "
                "own `metric` and `mode

https://github.com/open-mmlab/mmdetection/blob/7a9bc498d5cc972171ec4f7332afcd70bb50e60e/tools/analysis_tools/coco_error_analysis.py#L43
This I believe is for coco format, but I couldn't find any files for plotting precision or precision vs recall chart for pascal voc format.

🚀 Feature

The message here is printed per rank on purpose https://github.com/PyTorchLightning/pytorch-lightning/blob/bfa8b7be2d99b980afa62f5cb0433326bcfd2ef0/pytorch_lightning/utilities/seed.py#L69-L71 but this confuses users who are not aware of the difference

Pitch

Prefix the message with the rank just as we do for EarlyStopping https://github.com/PyTorchLightning/pytorch-light

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

🚀 Feature

Motivation

paper "LEARNING TO REPRESENT PROGRAMS WITH GRAPHS" which encode computer programs as graphs, with rich semantic information, however, most code implementation on this dataset VarMisuse is based on TensorFlow, like [tf-gnn-samples](https://github.com/microsof

Although the results look nice and ideal in all TensorFlow plots and are consistent across all frameworks, there is a small difference (more of a consistency issue). The result training loss/accuracy plots look like they are sampling on a lesser number of points. It looks more straight and smooth and less wiggly as compared to PyTorch or MXNet.

It can be clearly seen in chapter 6([CNN Lenet](ht

As mentioned in huggingface/datasets#2552 it would be nice to improve the error message when a dataset fails to build because there are duplicate example keys.

The current one is

datasets.keyhash.DuplicatedKeysError: FAILURE TO GENERATE DATASET !
Found duplicate Key: 48
Keys should be unique and deterministic in nature

and we could have something

We should sort imports with isort to keep the import section clean

Bug with GPU Model

Currently, while using pruning methods like TaylorFOWeight Pruner, If I use a model on GPU for getting the metrics (as calculated for getting masks), it fails on line while creating masks. The reason why it fails i