Anaconda builds and maintains over 17,000 open source Python packages alongside a catalog of more than 70 open source AI models. In this session, Seth Clark, VP of Product, breaks down the trends Anaconda has seen across serving, quantization, fine-tuning, evals, and guardrails projects, and how those insights shaped Anaconda’s own curated model catalog.

Key topics include:

• Trends across serving, quantization, fine-tuning, evals, and guardrails
• The nuts and bolts of running, quantizing, and benchmarking models
• How to run red teaming on large open source models, and common pitfalls
• What matters most in a model catalog at the enterprise level