LLM Fine-Tuning for Modern AI Teams: How One E-Commerce Unicorn Cut Inference Cost by 90%

Speaker: Emmanuel Turlay, CEO/Founder, Airtrain AI
While commercial LLMs such as GPT-4 and Claude 3 Opus offer amazing generative quality, small open-source fine-tuned models such as Mistral 7B and Phi-2/3 can offer similar performance on specific tasks, for a fraction of the cost, and with much more control. However, this has been proven to be true only when the tuning dataset is of high quality.
In this session, we will discuss the core concepts behind choosing a base model, how to prepare your dataset for success, how to run the fine-tuning job, and how to evaluate the tuned model.
In this talk, we will show how to use tools such as Airtrain AI to embark on your fine-tuning journey. How to pick a base model, how to prepare your dataset, how to run the fine-tuning job, and how to evaluate the tuned model.