Wallaroo 2024.3 Release

Updated the tutorials for the latest release of Wallaroo, version 2024.3

Wallaroo 2024.2 Release

This release is based on the updates for Wallaroo 2024.2. For full details, see the Wallaroo 2024.2 Release Notes.

These also include the new LLM (Large Language Model) deployment and monitoring tutorials. For access to these sample models and a demonstration on using LLMs with Wallaroo:

• Contact your Wallaroo Support Representative OR
• Schedule Your Wallaroo.AI Demo Today

Wallaroo 2024.1 Tutorial Release

This tutorial release is geared for Wallaroo 2024.1, with the following updates:

• Enhanced XGBoost performance
• Model Drift Detection via Assays updates
• New Python step process mainly used for pre and post processing
• Edge Observability enhancements - perform model drift from edge and multicloud deployments via assays, log uploads to the Wallaroo model ops center and on the edge deployments
• Model Acceleration and Model Architecture support for deployments and model edge deployments
• Enhanced parallel infer to work with DataFrames
• Assay improvements to the Wallaroo SDK

1.24-2023.4.1

Updated the tutorials for the 2023.4.1 release. This allows preprocessing and postprocessing model steps to function.

1.33 - 2023.4.0 Release

This release is tied to the new Wallaroo 2023.4.0 which includes:

• Wallaroo Community for GitHub Codespaces
• The Wallaroo Inference Server available on Azure Marketplace
• Updates on ONNX models. Default inputs for ONNX models are tied to the model, and are overridden with the model.config("tensor_fields") setting.
• Edge Observability: Wallaroo pipelines published to an OCI registry are deployable to a Docker or Kubernetes service as Wallaroo Server. Edge Observability enables those Wallaroo Server deployments to communicate with the Wallaroo Ops instance to transmit inference results from the Wallaroo Server that are added to the Wallaroo Pipeline.

See https://docs.wallaroo.ai for more details.

1.32 - 2023.3.0 Release

This release has the new tutorials for Edge and Arm based deployments for Wallaroo pipelines.

1.31 - 2023.2.1 Release

Updated with Wallaroo Frameworks for model uploads.

1.30 - 2023.2.1 Release

2023.2.1 release includes:

• Updated House Price Life Cycle: wallaroo-testing-tutorials/houseprice-saga
• Model Uploads and Registrations: model_uploads/
• GPU pipeline deployment feature: wallaroo-features/gpu-deployment
• Forecast model life cycle including parallel inference, data connections, and ML Workload Orchestrations: wallaroo-features/pipeline_multiple_replicas_forecast_tutorial
• Parallel Inference with the Aloha Model: wallaroo-features/parallel-inference-aloha-tutorial

See list of .zip files below to download specific tutorials.

1.29 2023.2.0

Updated the tutorials to the new release of Wallaroo 2023.2. This includes:

• All tutorials set to Arrow Enabled by default.
• New tutorials for ML Workload Orchestrations through the SDK and API.
• New tutorials for pipeline log retrieval through the SDK and API.
• Assays updated for multi-input and output models.

1.28 2023.1

This major release provides support for version 2023.1 and Arrow support. This release also provides the tutorials as separate .zip files for easy downloads.