https://github.com/Lightning-Universe/lightning-transformers

Flexible components pairing 🤗 Transformers with :zap: Pytorch Lightning
https://github.com/Lightning-Universe/lightning-transformers

hydra pytorch pytorch-lightning transformers

Last synced: about 2 months ago
JSON representation

Flexible components pairing 🤗 Transformers with :zap: Pytorch Lightning

• Host: GitHub
• URL: https://github.com/Lightning-Universe/lightning-transformers
• Owner: Lightning-Universe
• License: apache-2.0
• Archived: true
• Created: 2020-12-14T10:45:57.000Z (almost 4 years ago)
• Default Branch: master
• Last Pushed: 2022-11-21T16:35:36.000Z (almost 2 years ago)
• Last Synced: 2024-05-14T00:05:57.609Z (4 months ago)
• Topics: hydra, pytorch, pytorch-lightning, transformers
• Language: Python
• Homepage: https://lightning-transformers.readthedocs.io
• Size: 825 KB
• Stars: 607
• Watchers: 23
• Forks: 77
• Open Issues: 8
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md
  • License: LICENSE
  • Codeowners: .github/CODEOWNERS

Awesome Lists containing this project

README

        
# Deprecation notice 🔒

**This repository has been archived (read-only) on Nov 21, 2022**. Thanks to everyone who contributed to `lightning-transformers`, we feel it's time to move on.

:hugs: Transformers can **already be easily trained using the Lightning :zap: Trainer**. Here's a recent example from the community: . Note that there are **no limitations or workarounds**, things just work out of the box.

The `lightning-transformers` repo explored the possibility to provide task-specific modules and pre-baked defaults, at the cost of introducing extra abstractions. In the spirit of keeping ourselves focused, these abstractions are not something we wish to continue supporting.

If you liked `lightning-transformers` and want to continue developing it in the future, feel free to fork the repo and choose another name for the project.

______________________________________________________________________





**Flexible components pairing :hugs: Transformers with [Pytorch Lightning](https://github.com/PyTorchLightning/pytorch-lightning) :zap:**

______________________________________________________________________



  Docs •

  Community



______________________________________________________________________



## Installation

```bash

pip install lightning-transformers

```

From Source

```bash

git clone https://github.com/PyTorchLightning/lightning-transformers.git

cd lightning-transformers

pip install .

```

______________________________________________________________________

## What is Lightning-Transformers

Lightning Transformers provides `LightningModules`, `LightningDataModules` and `Strategies` to use :hugs: Transformers with the [PyTorch Lightning Trainer](https://pytorch-lightning.readthedocs.io/en/stable/common/trainer.html).

## Quick Recipes

#### Train [bert-base-cased](https://huggingface.co/bert-base-cased) on the [CARER](https://huggingface.co/datasets/emotion) emotion dataset using the Text Classification task.

```python

import pytorch_lightning as pl

from transformers import AutoTokenizer

from lightning_transformers.task.nlp.text_classification import (

    TextClassificationDataModule,

    TextClassificationTransformer,

)

tokenizer = AutoTokenizer.from_pretrained(

    pretrained_model_name_or_path="bert-base-cased"

)

dm = TextClassificationDataModule(

    batch_size=1,

    dataset_name="emotion",

    max_length=512,

    tokenizer=tokenizer,

)

model = TextClassificationTransformer(

    pretrained_model_name_or_path="bert-base-cased", num_labels=dm.num_classes

)

trainer = pl.Trainer(accelerator="auto", devices="auto", max_epochs=1)

trainer.fit(model, dm)

```

#### Train a pre-trained [mt5-base](https://huggingface.co/google/mt5-base) backbone on the [WMT16](https://huggingface.co/datasets/wmt16) dataset using the Translation task.

```python

import pytorch_lightning as pl

from transformers import AutoTokenizer

from lightning_transformers.task.nlp.translation import (

    TranslationTransformer,

    WMT16TranslationDataModule,

)

tokenizer = AutoTokenizer.from_pretrained(

    pretrained_model_name_or_path="google/mt5-base"

)

model = TranslationTransformer(

    pretrained_model_name_or_path="google/mt5-base",

    n_gram=4,

    smooth=False,

    val_target_max_length=142,

    num_beams=None,

    compute_generate_metrics=True,

)

dm = WMT16TranslationDataModule(

    # WMT translation datasets: ['cs-en', 'de-en', 'fi-en', 'ro-en', 'ru-en', 'tr-en']

    dataset_config_name="ro-en",

    source_language="en",

    target_language="ro",

    max_source_length=128,

    max_target_length=128,

    padding="max_length",

    tokenizer=tokenizer,

)

trainer = pl.Trainer(accelerator="auto", devices="auto", max_epochs=1)

trainer.fit(model, dm)

```

Lightning Transformers supports a bunch of :hugs: tasks and datasets. See the [documentation](https://lightning-transformers.readthedocs.io/en/latest/).

## Billion Parameter Model Support

### Big Model Inference

It's really easy to enable large model support for the pre-built LightningModule :hugs: tasks.

Below is an example to enable automatic model partitioning (across CPU/GPU and even leveraging disk space) to run text generation using a 6B parameter model.

```python

import torch

from accelerate import init_empty_weights

from transformers import AutoTokenizer

from lightning_transformers.task.nlp.language_modeling import (

    LanguageModelingTransformer,

)

with init_empty_weights():

    model = LanguageModelingTransformer(

        pretrained_model_name_or_path="EleutherAI/gpt-j-6B",

        tokenizer=AutoTokenizer.from_pretrained("EleutherAI/gpt-j-6B"),

        low_cpu_mem_usage=True,

        device_map="auto",  # automatically partitions the model based on the available hardware.

    )

output = model.generate("Hello, my name is", device=torch.device("cuda"))

print(model.tokenizer.decode(output[0].tolist()))

```

For more information see [Big Transformers Model Inference](https://lightning-transformers.readthedocs.io/en/latest/features/large_model.html).

### Big Model Training with DeepSpeed

Below is an example of how you can train a 6B parameter transformer model using Lightning Transformers and DeepSpeed.

```python

import pytorch_lightning as pl

from transformers import AutoTokenizer

from lightning_transformers.task.nlp.language_modeling import (

    LanguageModelingDataModule,

    LanguageModelingTransformer,

)

tokenizer = AutoTokenizer.from_pretrained(pretrained_model_name_or_path="gpt2")

model = LanguageModelingTransformer(

    pretrained_model_name_or_path="EleutherAI/gpt-j-6B",

    tokenizer=AutoTokenizer.from_pretrained("EleutherAI/gpt-j-6B"),

    deepspeed_sharding=True,  # defer initialization of the model to shard/load pre-train weights

)

dm = LanguageModelingDataModule(

    batch_size=1,

    dataset_name="wikitext",

    dataset_config_name="wikitext-2-raw-v1",

    tokenizer=tokenizer,

)

trainer = pl.Trainer(

    accelerator="gpu",

    devices="auto",

    strategy="deepspeed_stage_3",

    precision=16,

    max_epochs=1,

)

trainer.fit(model, dm)

```

For more information see [DeepSpeed Training with Big Transformers Models](https://lightning-transformers.readthedocs.io/en/latest/features/large_model_training.html) or the [Model Parallelism](https://pytorch-lightning.readthedocs.io/en/latest/advanced/model_parallel.html#fully-sharded-training) documentation.

## 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](https://www.pytorchlightning.ai/community)!