Gradsflow + Ray Tune integration

• migrate to ray_tune 🌟. Read more here. PR #36 by @aniketmaurya.
• render jupyter notebooks in documentation. PR #38 by @aniketmaurya.

$ pip install -U gradsflow==0.0.3a2

📝 Read more on documentation

0.0.3a1

0.0.3a1

• add optuna visualizations 🎨 . PR #27 by @aniketmaurya.
• add max_steps for HPO. PR #25 by @aniketmaurya.
• 📝 update docs & license. PR #23 by @aniketmaurya.
• fetch best trial model. PR #21 by @aniketmaurya.

0.0.2

• fetch best trial model. PR #21 by @aniketmaurya.

First stable Release 🎉🚀✨

Release Notes

0.0.1

• This changed: 📝 update example and documentation. Done by aniketmaurya. Check the Pull Request 20 with the changes and stuff. now back to code. 🤓\n
• This changed: 🎉✨ First Release - v0.0.1 - Refactor API & tested Python 3.7+. Done by aniketmaurya. Check the Pull Request 18 with the changes and stuff. now back to code. 🤓
• This changed: Adding example notebook for AutoSummarization. Done by the GitHub user gagan3012. Check the Pull Request 19 with the changes and stuff. now back to code. 🤓
• This changed: Adding text summarisation. Done by the GitHub user gagan3012. Check the Pull Request 14 with the changes and stuff. now back to code. 🤓
• This changed: add codecov CI. Done by the GitHub user aniketmaurya. Check the Pull Request 15 with the changes and stuff. now back to code. 🤓
• This changed: 📚 update documentation - added citation, acknowledgments, docstrings automation. Done by the GitHub user aniketmaurya. Check the Pull Request 13 with the changes and stuff. now back to code. 🤓
• Refactor API Design, CI & Docs PR #10 by @aniketmaurya.
• auto docstring. PR #7 by @aniketmaurya.
• Add AutoImageClassifier. PR #1 by @aniketmaurya.

What is Gradsflow?

Gradsflow is based on Optuna and PyTorch Lightning ⚡️. It leverages PyTorch Lightning Flash so that you don't have to write any PyTorch or Optuna code for model building or hyperparameter tuning 🚀

Although you might want to train a custom model and search hyperparameters, You can easily integrate any PyTorch/Lightning Flash Model with Gradsflow AutoModel ✨

gradsflow.core: Core defines the building blocks of AutoML tasks.

gradsflow.taskauto: AutoTasks defines different ML/DL tasks which is provided by Gradsflow AutoML API.

Image classification example

    from flash.core.data.utils import download_data
    from flash.image import ImageClassificationData

    from gradsflow import AutoImageClassifier

    # 1. Create the DataModule
    download_data("https://pl-flash-data.s3.amazonaws.com/hymenoptera_data.zip", "./data")

    datamodule = ImageClassificationData.from_folders(
        train_folder="data/hymenoptera_data/train/",
        val_folder="data/hymenoptera_data/val/",
    )

    suggested_conf = dict(
        optimizers=["adam", "sgd"],
        lr=(5e-4, 1e-3),
    )
    model = AutoImageClassifier(datamodule,
                                suggested_conf=suggested_conf,
                                max_epochs=10,
                                optimization_metric="val_accuracy",
                                timeout=300)
    model.hp_tune()

An AutoML Library made with Optuna and PyTorch Lightning

Refactoring and bug fixes

🎉 AutoClassification

• Added AutoImageClassification
• Added AutoTextClassification

Example

model = AutoImageClassifier(datamodule,
                            suggested_backbones=['ssl_resnet18'],
                            suggested_conf=suggested_conf,
                            max_epochs=1,
                            optimization_metric="val_accuracy",
                            timeout=30)

print("AutoImageClassifier initialised!")
model.hp_tune()