GitLab

* Fix the attention_mask in DPRReaderTokenizer

* Add an integration test for DPRReader inference

* Run make style

* Add head_mask/decoder_head_mask for BART

This branch implement head_mask and decoder_head_mask
for BART-based models. Full list below:
- BART
- MBart
- Blenderbot
- BlenderbotSmall
- Marian
- Pegasus

Everything is accompanied with updated testing.

* Fix test_headmasking for BART models

* Fix text_headmasking for BART-like models
which has only 2 layers in each modules.
The condition
```
self.assertNotEqual(attentions[1][..., 0, :, :].flatten().sum().item(), 0.0)
```
is, therefore, invalid for encoder-decoder models considering
the `head_mask`
```
head_mask = torch.ones(
    self.model_tester.num_hidden_layers,
    self.model_tester.num_attention_heads,
    device=torch_device,
)
head_mask[0, 0] = 0
head_mask[-1, :-1] = 0
```
specified in the `test_headmasking` test/function.

* Adjust test_modeling_common.py to reflect T5 input args

* Update tests/test_modeling_common.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* make style

* make fix-copies
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Ignore lm_head decoder bias warning

* Revert "Ignore lm_head decoder bias warning"

This reverts commit f25177a9da6ca898e351f46c8b1515971de5c670.

* predictions -> lm_head

* Add warning

* Remove unused import

* Fix missing call

* Fix missing call

* Completely remove token_type_ids

* Apply style

* Remove unused import

* Update src/transformers/models/mpnet/modeling_tf_mpnet.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* fix tf led

* remove loop file

This reverts commit 3f40070c.

* [doc] install + 1-gpu deployment

* Apply suggestions from code review
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* improvements
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Upstream (and rename) sortish sampler

* Use proper sampler

* Update src/transformers/trainer_pt_utils.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* gradient accumulation for tftrainer

* label naming
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* label naming
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* conda build -> conda-build

* Syntax error

* conda build -> conda-build + 4.2.0

* Prepare to merge in `master`

* note on how to get to deps from shell

* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* fix text
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* Compliancy with tf-nightly

* Add more version + restore min version check

* Switch metrics in run_ner to datasets

* Add flag to return all metrics

* Upstream (and rename) sortish_sampler

* Revert "Upstream (and rename) sortish_sampler"

This reverts commit e07d0dcf650c2bae36da011dd76c77a8bb4feb0d.

* Fix Trainer with a parallel model

* More clean up

* Fix conversational pipeline test

* LayoutLM

* ProphetNet

* BART

* Blenderbot & small

* Marian

* mBART

* Pegasus

* Tapas tokenizer

* BERT2BERT test

* Style

* Example requirements

* TF BERT2BERT test

* Fix data parallelism in Trainer

* Update src/transformers/training_args.py
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Update run_glue for do_predict with local test data (#9442)

* Update run_glue (#9442): fix comments ('files' to 'a file')

* Update run_glue (#9442): reflect the code review

* Update run_glue (#9442): auto format

* Update run_glue (#9442): reflect the code review

* make TopKLogitsWarper faster

* make TopPLogitsWarper faster

Co-authored-by: Pavel Tarashkevich <Pavel.Tarashkievich@orange.com>

* Update pretrained_models.rst

To clarify things cf. this tweet for instance https://twitter.com/RTomMcCoy/status/1349094111505211395

* format

* add model_input_names

* fix test

* deepspeed integration

* style

* add test

* ds wants to do its own backward

* fp16 assert

* Update src/transformers/training_args.py
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* style

* for clarity extract what args are being passed to deepspeed

* introduce the concept of self.wrapped_model

* s/self.wrapped_model/self.model_wrapped/

* complete transition to self.wrapped_model / self.model

* fix

* doc

* give ds its own init

* add custom overrides, handle bs correctly

* fix test

* clean up model_init logic, fix small bug

* complete fix

* collapse --deepspeed_config into --deepspeed

* style

* start adding doc notes

* style

* implement hf2ds optimizer and scheduler configuration remapping

* oops

* call get_num_training_steps absolutely when needed

* workaround broken auto-formatter

* deepspeed_config arg is no longer needed - fixed in deepspeed master

* use hf's fp16 args in config

* clean

* start on the docs

* rebase cleanup

* finish up --fp16

* clarify the supported stages

* big refactor thanks to discovering deepspeed.init_distributed

* cleanup

* revert fp16 part

* add checkpoint-support

* more init ds into integrations

* extend docs

* cleanup

* unfix docs

* clean up old code

* imports

* move docs

* fix logic

* make it clear which file it's referring to

* document nodes/gpus

* style

* wrong format

* style

* deepspeed handles gradient clipping

* easier to read

* major doc rewrite

* Apply suggestions from code review
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

* docs

* switch to AdamW optimizer

* style

* Apply suggestions from code review
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* clarify doc
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Co-authored-by: Lysandre Debut <lysandre@huggingface.co>

* Use the right version of tokenizers

* Try another way

* Try another way

* Deps are installed from there...

* Deps are installed from there...

* Revert last

* remove needless comment

* Add target contextmanager and rework prepare_seq2seq_batch

* Fix tests, treat BART and Barthez

* Add last tokenizers

* Fix test

* Set src token before calling the superclass

* Remove special behavior for T5

* Remove needless imports

* Remove needless asserts