Weights & Biases on Google Cloud

Accelerate AI development by running Weights & Biases on Google Cloud. Evaluate, monitor, and iterate on cutting-edge GenAI applications powered by the leading foundation models hosted on Google Cloud or train and fine-tune AI models. Combine Google Cloud’s robust infrastructure with the AI developer platform from Weights & Biases to deliver GenAI applications with confidence.

Trusted by the world's leading companies

By building W&B Weave and Models on Google Cloud and natively integrating with Vertex AI, Weights & Biases delivers an end-to-end platform to accelerate development of LLM-powered applications, for example Bayers’s drug discovery, Devsisters’s game development, and Cohere’s business-critical model training

Stephen Orban

VP of Migrations, ISVs and Marketplace at Google Cloud

Weights & Biases Wins Google Cloud Technology Partner of the Year Award for Generative AI

As Google Cloud’s Technology Partner of the Year for Generative AI, Weights & Biases integrates seamlessly with Google Cloud. Developers leveraging both the scalable infrastructure provided by Google Cloud Storage, Google Compute Engine, and Google Kubernetes Engine and ML development platforms such as VertexAI can take advantage of the Weights & Biases platform easily. In addition, enterprises using Gemini for Google Cloud can easily leverage and evaluate the models using Weights & Biases.

Why Weights & Biases on Google Cloud

Cutting-edge GenAI

Evaluate, monitor, and iterate on GenAI applications with W&B Weave on Vertex AI, or fine-tune state-of-the-art and open-source LLMs like Gemini Pro or Llama 3 using W&B Models and Vertex AI.

Simplified MLOps workflows

W&B Models with Vertex AI to provide ML engineers with easy access to accelerated compute like TPUs required to develop AI models. W&B Models also integrates with Vertex AI Private Endpoints to deploy models faster.

Unified system of record

Deploy Weights & Biases on Google Cloud and gain auditable history, consistency, reliability, and transparency across your entire workflow. Weights & Biases lets you keep track of all experiments and results, and access full model and data lineage.

AI security and compliance

Implement robust security measures using Google Cloud security services and granular access controls. Weights & Biases natively supports IAM, KMS, VPC and implements security and governance best practices from Google Cloud.

How it works

Features

Deep Google Cloud integration

Deploy W&B Models or Weave on Google Cloud to benefit from data isolation, private connectivity, and native integrations with Google Cloud AI tools and frameworks (Vertex AI, Tensorflow, Jax) and Google Cloud infrastructure.

LLM evaluations

W&B Weave integrates with Vertex AI to accelerate the development of LLM-powered applications. Leverage W&B Weave with Vertex AI, including how to evaluate the different foundation models available on Vertex AI.

Google Cloud-optimized performance

Leverage Google Cloud’s purpose-built ML infrastructure and compute services, like Google Kubernetes Engine (GKE), GPUs and TPUs to run large-scale ML experiments and scale workloads as needed.

Centralized model and dataset management

Version and manage your models and datasets seamlessly with W&B Registry and Google Cloud Storage. Ensure reproducibility and traceability throughout the model lifecycle while ensuring efficient data management and versioning.

Automate model deployment

Automatically deploy custom models and new model versions from the model registry to Vertex AI Endpoints. Ensure the best model versions get deployed to production as they iteratively improve over time, while capturing full pipelines for reproducibility.

Comprehensive experiment tracking

Automatically log hyperparameters, metrics, and artifacts for all your Vertex AI runs in W&B Models. Visualize and compare experiments to make data-driven decisions, all within your Google Cloud environment.

Collaborative workspaces

Seamlessly track and visualize ML experiments running on Google Cloud. Share results, leave comments, and collaborate on model optimization with Weights & Biases, fostering better communication throughout the ML development process.

Advanced LLM fine-tuning

Fine-tune foundation models like Gemini Pro and Gemini Ultra using W&B Models features, including logging, hyperparameter tuning, and visualization tools, fully integrated with Vertex AI’s powerful capabilities.

Governance framework

Implement robust governance with fine-grained access controls, audit trails, and seamless integration with Google Cloud security services and identity management services such as Identity and Access Management (IAM), Cloud Key Management Service (KMS), and VPC.

Bayer Pharmaceuticals uses AI to revolutionize drug discovery, from identifying potential drug targets to optimizing candidate molecules

Cohere trains business-critical LLMs by quickly understanding which candidate model outperforms the others, unlocking insights about the particular architectures and data recipes that work best.

Devsisters revolutionizes game AI development, enabling rapid experimentation and performance tracking for their machine learning models.

Get Started with Weights & Biases on Google Cloud

Learn more about Weights & Biases

Blogs and tutorials

Podcasts and webinars