Project-AgML/AgML

AgML is a centralized framework for agricultural machine learning. AgML provides access to public agricultural datasets for common agricultural deep learning tasks, with standard benchmarks and pretrained models, as well the ability to generate synthetic data and annotations.

What it solves

AgML addresses the difficulty of accessing and standardizing diverse agricultural datasets for machine learning. It provides a centralized library to discover, download, and load public agricultural data, reducing the time researchers and developers spend on data collection and preprocessing.

How it works

AgML acts as a data layer that integrates with the Hugging Face Hub. It uses a HuggingFaceDataLoader to fetch datasets and provides tools to split data into training, validation, and test sets with optional stratification. The library standardizes common agricultural tasks into specific formats: labels for classification, COCO-style bounding boxes for object detection, and single-channel masks for semantic segmentation. It is designed to be compatible with PyTorch and TensorFlow, and integrates directly with the Hugging Face datasets and transformers libraries for training and inference.

Who it’s for

Agricultural AI researchers, data scientists, and developers building deep learning models for crop monitoring, pest detection, and other agricultural tasks.

Highlights

  • Extensive Dataset Access: Provides a wide variety of public agricultural datasets from across the globe, including the massive iNatAg collection with over 4 million images.
  • Standardized Formats: Unifies different data sources into consistent formats for classification, detection, and segmentation.
  • Seamless Integration: Works out-of-the-box with Hugging Face's ecosystem, allowing users to use standard Trainer and Pipeline classes.
  • Framework Agnostic: Supports both PyTorch and TensorFlow.

Related

  • Project
    rllm-team/rllm

    A PyTorch library for Relational Table Learning (RTL) that standardizes and combines GNNs, LLMs, and TNNs to analyze data across linked tables.

  • Project
    PINTO0309/PINTO_model_zoo

    A comprehensive collection of AI models pre-converted and quantized across various frameworks like TensorFlow, PyTorch, ONNX, and CoreML for easier deployment on edge devices.

  • Project
    polyaxon/traceml

    An engine for ML and data tracking, visualization, and explainability that integrates with major deep learning frameworks to log metrics, inputs, and artifacts.

  • Project
    ZHZisZZ/dllm

    A library that unifies the training, inference, and evaluation of diffusion language models, providing scalable pipelines and recipes for converting AR models into diffusion models.

  • Project
    pyg-team/pytorch-frame

    A modular deep learning extension for PyTorch designed to build neural network models on heterogeneous tabular data containing mixed column types like text, images, and numerical values.