fondant 0.4.0

Sweet data-centric foundation model fine-tuning
Explore the docs »

Fondant helps you create high quality datasets to train or fine-tune foundation models such as:

• 🎨 Stable Diffusion
• 📄 GPT-like Large Language Models (LLMs)
• 🔎 CLIP
• ✂️ Segment Anything (SAM)
• ➕ And many more

🪤 Why Fondant?

Foundation models simplify inference by solving multiple tasks across modalities with a simple prompt-based interface. But what they've gained in the front, they've lost in the back. These models require enormous amounts of data, moving complexity towards data preparation, and leaving few parties able to train their own models.

We believe that innovation is a group effort, requiring collaboration. While the community has been building and sharing models, everyone is still building their data preparation from scratch. Fondant is the platform where we meet to build and share data preparation workflows.

Fondant offers a framework to build composable data preparation pipelines, with reusable components, optimized to handle massive datasets. Stop building from scratch, and start reusing components to:

• Extend your data with public datasets
• Generate new modalities using captioning, segmentation, translation, image generation, ...
• Distill knowledge from existing foundation models
• Filter out low quality data
• Deduplicate data

And create high quality datasets to fine-tune your own foundation models.

(back to top)

💨 Getting Started

Anxious to get started? Here's is a step by step guide to get your first pipeline up and running.

🪄 Example pipelines

Curious to see what Fondant can do? Have a look at our example pipelines:

Fine-tuning ControlNet

Our example pipeline to generate data for ControlNet fine-tuning allows you to create models that you can control using inpainting, segmentation, and regeneration. All you need to get started is a set of prompts describing the type of images to generate.

For instance, using our ControlNet model fine-tuned on interior design images, allows you to generate the room of your dreams:

Input image	Output image

Want to try out the resulting model yourself, head over to our Hugging Face space!

Fine-tuning Stable Diffusion

Using our example pipeline to fine-tune Stable Diffusion allows you to create models that generate better images within a specific domain. All you need to get started is a small seed dataset of example images.

Eg. generating logos:

Stable Diffusion 1.5	Fine-tuned Stable Diffusion 1.5

Training Starcoder

Using our example pipeline to train StarCoder provides a starting point to create datasets for training code assistants.

(back to top)

🧩 Reusable components

Fondant comes with a library of reusable components, which can jumpstart your pipeline.

(back to top)

⚒️ Installation

Fondant can be installed using pip:

pip install fondant

For the latest development version, you might want to install from source instead:

pip install git+https://github.com/ml6team/fondant.git

🧱 Deploying Fondant

There are 2 ways of using fondant:

• Leveraging Kubeflow pipelines on any Kubernetes cluster. All Fondant needs is an url pointing to the Kubeflow pipeline host and an Object Storage provider (S3, GCS, etc) to store data produced in the pipeline between steps. We have compiled some references and created some scripts to get you started with setting up the required infrastructure.
• Or locally by using docker compose. This way is mainly aimed at helping you develop fondant pipelines and components faster by making it easier to run things on a smaller scale.

The same pipeline can be used in both variants allowing you to quickly develop and iterate using the local Docker Compose implementation and then using the power of Kubeflow pipelines to run a large scale pipeline.

(back to top)

👨‍💻 Usage

Pipeline

Fondant allows you to easily define data pipelines comprised of both reusable and custom components. The following pipeline for instance uses the reusable load_from_hf_hub component to load a dataset from the Hugging Face Hub and process it using a custom component:

from fondant.pipeline import ComponentOp, Pipeline, Client


def build_pipeline():
    pipeline = Pipeline(pipeline_name="example pipeline", base_path="fs://bucket")

    load_from_hub_op = ComponentOp.from_registry(
        name="load_from_hf_hub",
        arguments={"dataset_name": "lambdalabs/pokemon-blip-captions"},
    )
    pipeline.add_op(load_from_hub_op)

    custom_op = ComponentOp(
        component_dir="components/custom_component",
        arguments={
            "min_width": 600,
            "min_height": 600,
        },
    )
    pipeline.add_op(custom_op, dependencies=load_from_hub_op)

    return pipeline
    

if __name__ == "__main__":
    client = Client(host="https://kfp-host.com/")
    pipeline = build_pipeline()
    client.compile_and_run(pipeline=pipeline)

Component

To create a custom component, you first need to describe its contract as a yaml specification. It defines the data consumed and produced by the component and any arguments it takes.

name: Custom component
description: This is a custom component
image: custom_component:latest

consumes:
  images:
    fields:
      data:
        type: binary

produces:
  captions:
    fields:
      data:
        type: utf8

args:
  argument1:
    description: An argument passed to the component at runtime
    type: str
  argument2:
    description: Another argument passed to the component at runtime
    type: str

Once you have your component specification, all you need to do is implement a constructor and a single .transform method and Fondant will do the rest. You will get the data defined in your specification partition by partition as a Pandas dataframe.

import pandas as pd
from fondant.component import PandasTransformComponent
from fondant.executor import PandasTransformExecutor


class ExampleComponent(PandasTransformComponent):

    def __init__(self, *args, argument1, argument2) -> None:
        """
        Args:
            argumentX: An argument passed to the component
        """
        # Initialize your component here based on the arguments

    def transform(self, dataframe: pd.DataFrame) -> pd.DataFrame:
        """Implement your custom logic in this single method
        Args:
            dataframe: A Pandas dataframe containing the data
        Returns:
            A pandas dataframe containing the transformed data
        """

For more advanced use cases, you can use the DaskTransformComponent instead.

Running your pipeline

Once you have a pipeline you can easily run (and compile) it by using the built-in CLI:

fondant run pipeline.py --local

To see all available arguments you can check the fondant CLI help pages

fondant --help

Or for a subcommand:

fondant <subcommand> --help

(back to top)

🚧 Current state and roadmap

Fondant is currently in the alpha stage, offering a minimal viable interface. While you should expect to run into rough edges, the foundations are ready and Fondant should already be able to speed up your data preparation work.

The following topics are on our roadmap

• Local pipeline execution
• Non-linear pipeline DAGs
• LLM-focused example pipelines and reusable components
• Static validation, caching, and partial execution of pipelines
• Data lineage and experiment tracking
• Distributed execution, both on and off cluster
• Support other dataframe libraries such as HF Datasets, Polars, Spark
• Move reusable components into a decentralized component registry
• Create datasets of copy-right free data for fine-tuning
• Create reusable components for bias detection and mitigation

The roadmap and priority are defined based on community feedback. To provide input, you can join our discord or submit an idea in our Github Discussions.

For a detailed view on the roadmap and day to day development, you can check our github project board.

(back to top)

👭 Contributing

We welcome contributions of different kinds:

Environment setup

We use poetry and pre-commit to enable a smooth developer flow. Run the following commands to set up your development environment:

pip install poetry
poetry install
pre-commit install

(back to top)