Flexible interface for high performance research using SOTA Transformers leveraging Pytorch Lightning, Transformers, and Hydra.
What is Lightning Transfomers • Using Lightning Transformers • Docs • Community • License
Installation
Option 1: from PyPI
pip install lightning-transformers
# instead of: `python train.py ...`, run with:
pl-transformers-train ...
Option 2: from source
git clone https://github.com/PyTorchLightning/lightning-transformers.git
cd lightning-transformers
pip install .
python train.py ...
# the `pl-transformers-train` endpoint is also available!
What is Lightning-Transformers
Lightning Transformers offers a flexible interface for training and fine-tuning SOTA Transformer models using the PyTorch Lightning Trainer.
- Train using HuggingFace Transformers models and datasets with Lightning custom Callbacks, Loggers, Accelerators and high performance scaling.
- Seamless Memory and Speed Optimizations such as DeepSpeed ZeRO or FairScale Sharded Training with no code changes.
- Powerful config composition backed by Hydra - Easily swap out models, optimizers, schedulers and many more configurations without touching the code.
- Transformer Task Abstraction for Rapid Research & Experimentation - Built from the ground up to be task agnostic, the library supports creating transformer tasks across all modalities with little friction.
Lightning Transformers tasks allow you to train models using HuggingFace Transformer models and datasets, use Hydra to hotswap models, optimizers or schedulers and leverage all the advances features that Lightning has to offer, including custom Callbacks, Loggers, Accelerators and high performance scaling with minimal changes.
Using Lightning-Transformers
Grid is our platform for training models at scale on the cloud! Sign up here.
| Task | Quick Commands | Run |
|---|---|---|
| Language Modeling | python train.py task=nlp/language_modeling dataset=nlp/language_modeling/wikitext trainer.gpus=1 training.batch_size=8 |
 |
| Multiple Choice | python train.py task=nlp/multiple_choice dataset=nlp/multiple_choice/race trainer.gpus=1 |
|
| Question Answering | python train.py task=nlp/question_answering dataset=nlp/question_answering/squad trainer.gpus=1 |
|
| Summarization | python train.py task=nlp/summarization dataset=nlp/summarization/xsum trainer.gpus=1 |
|
| Text Classification | python train.py task=nlp/text_classification dataset=nlp/text_classification/emotion trainer.gpus=1 |
|
| Token Classification | python train.py task=nlp/token_classification dataset=nlp/token_classification/conll trainer.gpus=1 |
|
| Translation | python train.py task=nlp/translation dataset=nlp/translation/wmt16 trainer.gpus=1 |
|
Quick recipes
Train bert-base-cased on the CARER emotion dataset using the Text Classification task.
python train.py \
task=nlp/text_classification \
dataset=nlp/text_classification/emotion
See the composed Hydra config used under-the-hood
optimizer:
_target_: torch.optim.AdamW
lr: ${training.lr}
weight_decay: 0.001
scheduler:
_target_: transformers.get_linear_schedule_with_warmup
num_training_steps: -1
num_warmup_steps: 0.1
training:
run_test_after_fit: true
lr: 5.0e-05
output_dir: .
batch_size: 16
num_workers: 16
trainer:
_target_: pytorch_lightning.Trainer
logger: true
checkpoint_callback: true
callbacks: null
default_root_dir: null
gradient_clip_val: 0.0
process_position: 0
num_nodes: 1
num_processes: 1
gpus: null
auto_select_gpus: false
tpu_cores: null
log_gpu_memory: null
progress_bar_refresh_rate: 1
overfit_batches: 0.0
track_grad_norm: -1
check_val_every_n_epoch: 1
fast_dev_run: false
accumulate_grad_batches: 1
max_epochs: 1
min_epochs: 1
max_steps: null
min_steps: null
limit_train_batches: 1.0
limit_val_batches: 1.0
limit_test_batches: 1.0
val_check_interval: 1.0
flush_logs_every_n_steps: 100
log_every_n_steps: 50
accelerator: null
sync_batchnorm: false
precision: 32
weights_summary: top
weights_save_path: null
num_sanity_val_steps: 2
truncated_bptt_steps: null
resume_from_checkpoint: null
profiler: null
benchmark: false
deterministic: false
reload_dataloaders_every_epoch: false
auto_lr_find: false
replace_sampler_ddp: true
terminate_on_nan: false
auto_scale_batch_size: false
prepare_data_per_node: true
plugins: null
amp_backend: native
amp_level: O2
move_metrics_to_cpu: false
task:
_recursive_: false
backbone: ${backbone}
optimizer: ${optimizer}
scheduler: ${scheduler}
_target_: lightning_transformers.task.nlp..text_classification.TextClassificationTransformer
downstream_model_type: transformers.AutoModelForSequenceClassification
dataset:
cfg:
batch_size: ${training.batch_size}
num_workers: ${training.num_workers}
dataset_name: emotion
dataset_config_name: null
train_file: null
validation_file: null
test_file: null
train_val_split: null
max_samples: null
cache_dir: null
padding: max_length
truncation: only_first
preprocessing_num_workers: 1
load_from_cache_file: true
max_length: 128
limit_train_samples: null
limit_val_samples: null
limit_test_samples: null
_target_: lightning_transformers.task.nlp.text_classification.TextClassificationDataModule
experiment_name: ${now:%Y-%m-%d}_${now:%H-%M-%S}
log: false
ignore_warnings: true
tokenizer:
_target_: transformers.AutoTokenizer.from_pretrained
pretrained_model_name_or_path: ${backbone.pretrained_model_name_or_path}
use_fast: true
backbone:
pretrained_model_name_or_path: bert-base-cased
Swap the backbone to RoBERTa and the optimizer to RMSprop:
python train.py \
task=nlp/text_classification \
dataset=nlp/text_classification/emotion
backbone.pretrained_model_name_or_path=roberta-base
optimizer=rmsprop
See the changed Hydra config under-the-hood
optimizer:
- _target_: torch.optim.AdamW
+ _target_: torch.optim.RMSprop
lr: ${training.lr}
- weight_decay: 0.001
scheduler:
_target_: transformers.get_linear_schedule_with_warmup
num_training_steps: -1
....
tokenizer:
pretrained_model_name_or_path: ${backbone.pretrained_model_name_or_path}
use_fast: true
backbone:
- pretrained_model_name_or_path: bert-base-cased
+ pretrained_model_name_or_path: roberta-base
Enable Sharded Training.
python train.py \
task=nlp/text_classification \
dataset=nlp/text_classification/emotion \
trainer=sharded
See the changed Hydra config under-the-hood
Without the need to modify any code, the config updated automatically for sharded training:optimizer:
_target_: torch.optim.AdamW
lr: ${training.lr}
trainer:
process_position: 0
num_nodes: 1
num_processes: 1
- gpus: null
+ gpus: 1
auto_select_gpus: false
tpu_cores: null
log_gpu_memory: null
...
log_every_n_steps: 50
- accelerator: null
+ accelerator: ddp
sync_batchnorm: false
- precision: 32
+ precision: 16
weights_summary: top
....
terminate_on_nan: false
auto_scale_batch_size: false
prepare_data_per_node: true
- plugins: null
+ plugins:
+ _target_: pytorch_lightning.plugins.DDPShardedPlugin
amp_backend: native
amp_level: O2
move_metrics_to_cpu: false
tokenizer:
pretrained_model_name_or_path: ${backbone.pretrained_model_name_or_path}
use_fast: true
backbone:
pretrained_model_name_or_path: bert-base-cased
Enable DeepSpeed ZeRO Training.
python train.py \
task=nlp/text_classification \
dataset=nlp/text_classification/emotion \
trainer=deepspeed
See the changed Hydra config under-the-hood
Without the need to modify any code, the config updated automatically for DeepSpeed:optimizer:
_target_: torch.optim.AdamW
lr: ${training.lr}
trainer:
process_position: 0
num_nodes: 1
num_processes: 1
- gpus: null
+ gpus: 1
auto_select_gpus: false
tpu_cores: null
log_gpu_memory: null
...
val_check_interval: 1.0
flush_logs_every_n_steps: 100
log_every_n_steps: 50
- accelerator: null
+ accelerator: ddp
sync_batchnorm: false
- precision: 32
+ precision: 16
...
- plugins: null
+ plugins:
+ _target_: pytorch_lightning.plugins.DeepSpeedPlugin
+ stage: 2
+ cpu_offload: true
amp_backend: native
amp_level: O2
move_metrics_to_cpu: false
...
Train with a pre-trained t5-base backbone, on the XSUM dataset using the Summarization task.
python train.py \
task=nlp/summarization \
dataset=nlp/summarization/xsum \
backbone.pretrained_model_name_or_path=t5-base
Train with a pre-trained mt5-base backbone, on the WMT16 dataset using the Translation task with 2 GPUs.
python train.py \
task=nlp/translation \
dataset=nlp/translation/wmt16 \
backbone.pretrained_model_name_or_path=google/mt5-base \
trainer.gpus=2
Custom Files & Datasets
You can train, validate and test Lightning transformers tasks on your own data files, and you can extend datasets for custom processing and your own tasks.
How to train, validate and test on custom files
How to extend datasets
Custom Tasks
Extending the Language Modeling Task
Contribute
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.
Please make sure to update tests as appropriate.
Community
For help or questions, join our huge community on Slack!
License
Please observe the Apache 2.0 license that is listed in this repository. In addition, the Lightning framework is Patent Pending.
2.3k Jan 05, 2023
4 Feb 23, 2022
1.5k Jan 02, 2023
37 Dec 14, 2022
613 Jan 07, 2023
76 Nov 12, 2022
70 Dec 11, 2022
1 Dec 22, 2021
71 Jan 06, 2023
375 Dec 28, 2022
213 Jan 04, 2023
592 Dec 18, 2022
847 Dec 19, 2022
94 Dec 27, 2022
1k Jan 02, 2023
2 Nov 13, 2021
120 Dec 27, 2022
9 Nov 09, 2021
19 Sep 07, 2022
2 Nov 11, 2022