Sign up
Log in
Sign up
Log in
On-demand webinar
Learn from our technical deep dive into using function calling to develop AI agents.
Watch now
Back to demos & webinars
Solutions

OctoAI SaaS, OctoStack

Format

9 minute walkthrough

Key Learnings

1. Selecting the right model, 2. Using checkpoints or LoRAs, 3. Evaluating model success, 4. Self-hosting options, 5. Getting the most from your hardware

Date Published

Jun 12, 2024

Publishers

Alyss Noland, Thierry Moreau

Back to demos & webinars

Selecting the right GenAI model for production

Solutions

OctoAI SaaS, OctoStack

Format

9 minute walkthrough

Key Learnings

1. Selecting the right model, 2. Using checkpoints or LoRAs, 3. Evaluating model success, 4. Self-hosting options, 5. Getting the most from your hardware

Date Published

Jun 12, 2024

Publishers

Alyss Noland, Thierry Moreau

Watch our on-demand webinar as our engineers review all steps of model evaluation, testing, when to use checkpoints vs LoRAs, and how to get the best results for your needs.

There are over 90k text gen models in the market. Picking the right LLM for your job is a critical step in productizing your GenAI project, but the sheer number of options can trip up even the savviest of teams.

You can’t afford to slow down your AI initiatives or bank on the wrong model. This on-demand shortened webinar walkthrough will help you streamline the process of model selection and onboarding, and provide a framework for evaluating LLMs in your GenAI stack.

This on-demand session will provide practical advice for engineering leads and technical decision makers tasked with delivering innovation with generative models.

You’ll learn:
  • How to narrow the universe of models to a short list fit for your task

  • Why would you choose to use a checkpoint vs. LoRA vs. a Community tune

  • How to evaluate model success across your prompts, data, training, and tuning

  • Options for self-hosting open source models in privacy/security sensitive use cases

  • How to achieve highly performant inference with optimal unit economics and hardware utilization

Watch now