* make sure that logging_first_step evaluates
* fix bug with incorrect loss on logging_first_step
* fix style
* logging_first_step only logs, not evals
* first attempt to add AzureML callbacks
* func arg fix
* var name fix, but still won't fix error...
* fixing as in https://discuss.huggingface.co/t/how-to-integrate-an-azuremlcallback-for-logging-in-azure/1713/2
* Avoid lint check of azureml import
* black compliance
* Make isort happy
* Fix point typo in docs
* Add AzureML to Callbacks docs
* Attempt to make sphinx happy
* Format callback docs
* Make documentation style happy
* Make docs compliant to style
Co-authored-by: Davide Fiocco <davide.fiocco@frontiersin.net>
* Fix comet_ml import and add ensure availability
* Make isort happy
* Make flake8 happy
* Don't show comet_ml warn if COMET_MODE=DISABLED
* Make isort happy
* Important files
* Styling them all
* Revert "Styling them all"
This reverts commit 7d029395fdae8513b8281cbc2a6c239f8093503e.
* Syling them for realsies
* Fix syntax error
* Fix benchmark_utils
* More fixes
* Fix modeling auto and script
* Remove new line
* Fixes
* More fixes
* Fix more files
* Style
* Add FSMT
* More fixes
* More fixes
* More fixes
* More fixes
* Fixes
* More fixes
* More fixes
* Last fixes
* Make sphinx happy
* Add MLflow integration class
Add integration code for MLflow in integrations.py along with the code
that checks that MLflow is installed.
* Add MLflowCallback import
Add import of MLflowCallback in trainer.py
* Handle model argument
Allow the callback to handle model argument and store model config items as hyperparameters.
* Log parameters to MLflow in batches
MLflow cannot log more than a hundred parameters at once.
Code added to split the parameters into batches of 100 items and log the batches one by one.
* Fix style
* Add docs on MLflow callback
* Fix issue with unfinished runs
The "fluent" api used in MLflow integration allows only one run to be active at any given moment. If the Trainer is disposed off and a new one is created, but the training is not finished, it will refuse to log the results when the next trainer is created.
* Add MLflow integration class
Add integration code for MLflow in integrations.py along with the code
that checks that MLflow is installed.
* Add MLflowCallback import
Add import of MLflowCallback in trainer.py
* Handle model argument
Allow the callback to handle model argument and store model config items as hyperparameters.
* Log parameters to MLflow in batches
MLflow cannot log more than a hundred parameters at once.
Code added to split the parameters into batches of 100 items and log the batches one by one.
* Fix style
* Add docs on MLflow callback
* Fix issue with unfinished runs
The "fluent" api used in MLflow integration allows only one run to be active at any given moment. If the Trainer is disposed off and a new one is created, but the training is not finished, it will refuse to log the results when the next trainer is created.
* Start simplification
* More progress
* Finished script
* Address comments and update tests instructions
* Wrong test
* Accept files as inputs and fix test
* Update src/transformers/trainer_utils.py
Co-authored-by: Julien Chaumond <chaumond@gmail.com>
* Fix labels and add combined score
* Add special labels
* Update TPU command
* Revert to old label strategy
* Use model labels
* Fix for STT-B
* Styling
* Apply suggestions from code review
Co-authored-by: Thomas Wolf <thomwolf@users.noreply.github.com>
* Code styling
* Fix review comments
Co-authored-by: Julien Chaumond <chaumond@gmail.com>
Co-authored-by: Thomas Wolf <thomwolf@users.noreply.github.com>
I'm using transformers 3.3.1 and run a training script with `--save_total_limit 3`. I hit the exception below, and after debugging the code found that it wrongly tries to index into the `best_model_checkpoint`'s *str* rather than the `sorted_checkpoints` array. When running without the fix I got this exception:
```
Traceback (most recent call last):
File "/<HOME>/.conda/envs/transformers/lib/python3.7/site-packages/transformers/trainer.py", line 921, in _save_training
self._rotate_checkpoints(use_mtime=True)
File "/<HOME>/.conda/envs/transformers/lib/python3.7/site-packages/transformers/trainer.py", line 1283, in _rotate_checkpoints
checkpoints_sorted = self._sorted_checkpoints(use_mtime=use_mtime)
File "/<HOME>/.conda/envs/transformers/lib/python3.7/site-packages/transformers/trainer.py", line 1274, in _sorted_checkpoints
checkpoints_sorted[best_model_index],
TypeError: 'str' object does not support item assignment
```
* fix 5990
* accomodate iterable dataset without predefined length
* set it as 1 use case: provide max_steps, and NO num_epochs
* Is a merge of master and PR 5995
* fix trainer test under TF
* fix only for torch
* TF trainer untouched
* trainer tests are skipped when no torch
* address comments
* fix quality checks
* remove torch.dataset from test_trainer
* unnecessary inheritance
* RegressionDataset implements all needed methods __len__ and __getitem__
* fix quality checks
* restore RegressionDataset
* was wrongly under is_torch_available()
* splitting fast and slow tokenizers [WIP]
* [WIP] splitting sentencepiece and tokenizers dependencies
* update dummy objects
* add name_or_path to models and tokenizers
* prefix added to file names
* prefix
* styling + quality
* spliting all the tokenizer files - sorting sentencepiece based ones
* update tokenizer version up to 0.9.0
* remove hard dependency on sentencepiece 🎉
* and removed hard dependency on tokenizers 🎉
* update conversion script
* update missing models
* fixing tests
* move test_tokenization_fast to main tokenization tests - fix bugs
* bump up tokenizers
* fix bert_generation
* update ad fix several tokenizers
* keep sentencepiece in deps for now
* fix funnel and deberta tests
* fix fsmt
* fix marian tests
* fix layoutlm
* fix squeezebert and gpt2
* fix T5 tokenization
* fix xlnet tests
* style
* fix mbart
* bump up tokenizers to 0.9.2
* fix model tests
* fix tf models
* fix seq2seq examples
* fix tests without sentencepiece
* fix slow => fast conversion without sentencepiece
* update auto and bert generation tests
* fix mbart tests
* fix auto and common test without tokenizers
* fix tests without tokenizers
* clean up tests lighten up when tokenizers + sentencepiece are both off
* style quality and tests fixing
* add sentencepiece to doc/examples reqs
* leave sentencepiece on for now
* style quality split hebert and fix pegasus
* WIP Herbert fast
* add sample_text_no_unicode and fix hebert tokenization
* skip FSMT example test for now
* fix style
* fix fsmt in example tests
* update following Lysandre and Sylvain's comments
* Update src/transformers/testing_utils.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/testing_utils.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Update src/transformers/tokenization_utils_base.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* use DDP no_sync when possible
* fix is_nlp_available addition mistake
* reformat trainer.py
* reformat trainer.py
* drop support for pytorch < 1.2
* return support for pytorch < 1.2
* Initial callback proposal
* Finish various callbacks
* Post-rebase conflicts
* Fix tests
* Don't use something that's not set
* Documentation
* Remove unwanted print.
* Document all models can work
* Add tests + small fixes
* Update docs/source/internal/trainer_utils.rst
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Address review comments
* Fix TF tests
* Real fix this time
* This one should work
* Fix typo
* Really fix typo
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
* Trainer should not modify its TrainingArguments
* Trainer should not modify its TrainingArguments
* Trainer should not modify its TrainingArguments
* Add test of resumed training
* Fixes
* Non multiGPU test
* Clean Trainer state
* Add more to the state
* Documentation
* One last test
* Make resume training test more complete
* Unwanted changes
* GPT2 gradient checkpointing
* find_unused_parameters removed if checkpointing
* find_unused_parameters removed if checkpointing
* Update src/transformers/configuration_gpt2.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* Added a test for generation with checkpointing
* Update src/transformers/configuration_gpt2.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
* Add dataloader_num_workers to TrainingArguments
This argument is meant to be used to set the
number of workers for the PyTorch DataLoader.
* Pass num_workers argument on DataLoader init
* fix ZeroDivisionError and epoch counting
* Add test for num_train_epochs calculation in trainer.py
* Remove @require_non_multigpu for test_num_train_epochs_in_training
* Should check if `torch` is available
* fixed samples_count error, distributed_concat arguments
* style
* Import torch at beginning of file
Co-authored-by: TevenLeScao <teven.lescao@gmail.com>
* fixed trainer tr_loss memory leak
* detached returned training loss from computation graph in the Trainer class' training_step() method
* Revert "fixed trainer tr_loss memory leak"
This reverts commit 47226e4e
* Introduce HPO checkpointing for PBT
* Moved checkpoint saving
* Fixed checkpoint subdir pass
* Fixed style
* Enable/disable checkpointing, check conditions for various tune schedulers incl. PBT
* Adjust number of GPUs to number of jobs
* Avoid mode pickling in ray
* Move hp search to integrations
