polyaxon/traceml

Engine for AI/ML/Data tracking, visualization, explainability, drift detection, and dashboards for Polyaxon.

What it solves

TraceML provides a centralized engine for tracking machine learning and data experiments. It addresses the need for consistent logging of hyperparameters, metrics, data references, and model artifacts across different frameworks, while also providing tools for data summary and visualization.

How it works

TraceML integrates into ML workflows via a Python API that allows users to initialize tracking sessions and log various types of data. It offers specialized callbacks for popular deep learning frameworks like Keras, PyTorch, TensorFlow, Fastai, PyTorch Lightning, and HuggingFace to automate the collection of metrics and models. Additionally, it includes a DataFrameSummary tool to extend pandas' describe() function for deeper data profiling.

Who it’s for

Data scientists and ML engineers who need to track their experiments, monitor model performance, and analyze datasets across multiple different ML libraries.

Highlights

  • Multi-framework support: Native callbacks for Keras, PyTorch, TensorFlow, Fastai, PyTorch Lightning, and HuggingFace.
  • Rich artifact logging: Support for logging images (Matplotlib) and interactive charts (Bokeh, Altair, Plotly).
  • Data profiling: Enhanced pandas DataFrame summaries including column statistics, types, and detailed numerical analysis.
  • Flexible deployment: Supports both online tracking via Polyaxon and an offline mode for local logging.

Related

  • Project
    clearml/clearml

    An open-source MLOps and LLMOps suite that provides experiment tracking, orchestration, and data management to streamline the deep learning development lifecycle.

  • Project
    open-mmlab/mmengine

    A foundational PyTorch library that serves as a training engine for deep learning models, providing standardized components for training loops, configuration, and monitoring.

  • Project
    Project-AgML/AgML

    AgML is a comprehensive library for agricultural machine learning that provides standardized access to a wealth of public agricultural datasets for common deep learning tasks.

  • Project
    traceloop/openllmetry

    An open-source observability framework built on OpenTelemetry that provides complete tracing and monitoring for LLM applications, providers, and vector databases.

  • Project
    apple/coremltools

    A Python package for converting machine learning models from frameworks like PyTorch and TensorFlow into the Core ML format for on-device deployment on Apple devices.