GitLab

* per_device instead of per_gpu/error thrown when argument unknown

* [docs] Restore examples.md symlink

* Correct absolute links so that symlink to the doc works correctly

* Update src/transformers/hf_argparser.py
Co-authored-by: Julien Chaumond <chaumond@gmail.com>

* Warning + reorder

* Docs

* Style

* not for squad
Co-authored-by: Julien Chaumond <chaumond@gmail.com>

As discussed w/ @lysandrejik

packaging is maintained by PyPA (the Python Packaging Authority), and should be lightweight and stable

* TPU hangs when saving optimizer/scheduler

* Style

* ParallelLoader is not a DataLoader

* Style

* Addressing @julien-c's comments

* Map optimizer to correct device after loading from checkpoint.

* Make style test pass
Co-authored-by: Julien Chaumond <chaumond@gmail.com>

* Distributed eval: SequentialDistributedSampler + gather all results

* For consistency only write to disk from world_master

Close https://github.com/huggingface/transformers/issues/4272

* Working distributed eval

* Hook into scripts

* Fix #3721 again

* TPU.mesh_reduce: stay in tensor space

Thanks @jysohn23

* Just a small comment

* whitespace

* torch.hub: pip install packaging

* Add test scenarii

* makes fetching last learning late in trainer backward compatible

* split comment to multiple lines

* fixes black styling issue

* uses version to create a more explicit logic

* fix loss calculation in evaluation

* fix evaluation on TPU when prediction_loss_only is True

* Improvements to the wandb integration

* small reorg + no global necessary

* feat(trainer): log epoch and final metrics

* Simplify logging a bit

* Fixup

* Fix crash when just running eval
Co-authored-by: Chris Van Pelt <vanpelt@gmail.com>
Co-authored-by: Boris Dayma <boris.dayma@gmail.com>

* [TPU] Doc, fix xla_spawn.py, only preprocess dataset once

* Update examples/README.md

* [xla_spawn] Add `_mp_fn` to other Trainer scripts

* [TPU] Fix: eval dataloader was None

* Created using Colaboratory

* [examples] reorganize files

* remove run_tpu_glue.py as superseded by TPU support in Trainer

* Bugfix: int, not tuple

* move files around

cc @patrickvonplaten @thomwolf

* wip

* wip

* a last wip

* Better logging when using TPUs

* Correct argument name

* Tests

* fix

* Metrics in evaluation

* Update src/transformers/training_args.py

* [tpu] Use launcher script instead

* [tpu] lots of tweaks

* Fix formatting
Co-authored-by: Julien Chaumond <chaumond@gmail.com>

* First commit to add a TF version of the trainer.

* Make the TF trainer closer to what looks the PT trainer

* Refactoring common code between the PT and TF trainer into an util file.

* Some bugfix + better similarity with the PT trainer

* Add missing class in transformers init

* Bugfix over prediction + use classification report instead of simple metrics

* Fix name error

* Fix optimization tests + style

* Apply style

* Several bugfix for multi-gpu training

* Apply style

* Apply style

* Add glue example for the TF trainer

* Several bugix + address the reviews

* Fix on the TF training args file

* Add a debug mode

* Bugfix in utils_ner.py when segment_ids is None

* Apply style

* Apply style

* Add TPU strategy

* Fix selection strategy

See https://github.com/huggingface/transformers/pull/3916

* feat: add logging through Weights & Biases

* feat(wandb): make logging compatible with all scripts

* style(trainer.py): fix formatting

* [Trainer] Tweak wandb integration
Co-authored-by: Julien Chaumond <chaumond@gmail.com>

* Continue training args

* Continue training args

* added explaination

* added explaination

* added explaination

* Fixed tqdm auto

* Update src/transformers/training_args.py
Co-Authored-By: Julien Chaumond <chaumond@gmail.com>

* Update src/transformers/training_args.py

* Update src/transformers/training_args.py
Co-authored-by: Julien Chaumond <chaumond@gmail.com>

Close #3920

* doc

* [tests] Add sample files for a regression task

* [HUGE] Trainer

* Feedback from @sshleifer

* Feedback from @thomwolf + logging tweak

* [file_utils] when downloading concurrently, get_from_cache will use the cached file for subsequent processes

* [glue] Use default max_seq_length of 128 like before

* [glue] move DataTrainingArguments around

* [ner] Change interface of InputExample, and align run_{tf,pl}

* Re-align the pl scripts a little bit

* ner

* [ner] Add integration test

* Fix language_modeling with API tweak

* [ci] Tweak loss target

* Don't break console output

* amp.initialize: model must be on right device before

* [multiple-choice] update for Trainer

* Re-align to 827d6d6e