Version 2.9 of 🤗 Transformers introduced a new Trainer class for PyTorch, and its equivalent TFTrainer for TF 2. Running the examples requires PyTorch 1.3.1+ or TensorFlow 2.2+.

Here is the list of all our examples:

• grouped by task (all official examples work for multiple models)
• with information on whether they are built on top of Trainer/TFTrainer (if not, they still work, they might just lack some features),
• whether or not they leverage the 🤗 Datasets library.
• links to Colab notebooks to walk through the scripts and run them easily,
• links to Cloud deployments to be able to deploy large-scale trainings in the Cloud with little to no setup.

Important

To make sure you can successfully run the latest versions of the example scripts, you have to install the library from source and install some example-specific requirements. Execute the following steps in a new virtual environment:

git clone https://github.com/huggingface/transformers
cd transformers
pip install .
pip install -r ./examples/requirements.txt

Alternatively, you can run the version of the examples as they were for your current version of Transformers via (for instance with v3.4.0):

git checkout tags/v3.4.0

Task	Example datasets	Trainer support	TFTrainer support	🤗 Datasets	Colab
language-modeling	Raw text	✅	-	✅
text-classification	GLUE, XNLI	✅	✅	✅
token-classification	CoNLL NER	✅	✅	✅	-
multiple-choice	SWAG, RACE, ARC	✅	✅	-
question-answering	SQuAD	✅	✅	✅	-
text-generation	-	n/a	n/a	-
distillation	All	-	-	-	-
summarization	CNN/Daily Mail	✅	-	-	-
translation	WMT	✅	-	-	-
bertology	-	-	-	-	-
adversarial	HANS	✅	-	-	-

Coming soon!

When using Tensorflow, TPUs are supported out of the box as a tf.distribute.Strategy.

When using PyTorch, we support TPUs thanks to pytorch/xla. For more context and information on how to setup your TPU environment refer to Google's documentation and to the very detailed pytorch/xla README.

In this repo, we provide a very simple launcher script named xla_spawn.py that lets you run our example scripts on multiple TPU cores without any boilerplate. Just pass a --num_cores flag to this script, then your regular training script with its arguments (this is similar to the torch.distributed.launch helper for torch.distributed). Note that this approach does not work for examples that use pytorch-lightning.

For example for run_glue:

python examples/xla_spawn.py --num_cores 8 \
	examples/text-classification/run_glue.py \
	--model_name_or_path bert-base-cased \
	--task_name mnli \
	--data_dir ./data/glue_data/MNLI \
	--output_dir ./models/tpu \
	--overwrite_output_dir \
	--do_train \
	--do_eval \
	--num_train_epochs 1 \
	--save_steps 20000

Feedback and more use cases and benchmarks involving TPUs are welcome, please share with the community.

You can easily log and monitor your runs code. The following are currently supported:

• TensorBoard
• Weights & Biases
• Comet ML

To use Weights & Biases, install the wandb package with:

pip install wandb

Then log in the command line:

wandb login

If you are in Jupyter or Colab, you should login with:

import wandb
wandb.login()

Whenever you use Trainer or TFTrainer classes, your losses, evaluation metrics, model topology and gradients (for Trainer only) will automatically be logged.

When using 🤗 Transformers with PyTorch Lightning, runs can be tracked through WandbLogger. Refer to related documentation & examples.

To use comet_ml, install the Python package with:

pip install comet_ml

or if in a Conda environment:

conda install -c comet_ml -c anaconda -c conda-forge comet_ml