Exploring and Practicing Large Models at Ctrip

Wei Peng – Algorithm Expert at Ctrip

Personal introduction: Master’s graduate of East China Normal University with nearly ten years of machine‑learning application experience, currently working at Ctrip, focusing on intelligent decision‑making and large‑model applications.

Talk Title: Exploration and Practice of Large Models at Ctrip

Talk introduction: Large language models are evolving rapidly, with core abilities that can be roughly divided into generation, summarization, extraction, classification, retrieval, and rewriting. We explore how these capabilities perform in real business scenarios. This talk presents large‑model deployments in the tourism industry, including AIGC for content marketing, multimodal models for interactive pages, and open‑source models that improve customer‑service efficiency. Topics cover open‑source model comparisons, fine‑tuning practices, and performance and effect comparisons of different inference frameworks.

Talk Outline:

1. Related business background introduction

2. Explore core capabilities of large models in specific scenarios

· Comparison of open‑source large‑model applications

· Large‑model fine‑tuning practice

· Large‑model inference framework performance and effect

3. Summary of related practice experience

Audience Benefits:

1. Understand large‑model capabilities and their applications across various scenarios

2. Learn large‑model fine‑tuning experience

3. Learn large‑model inference performance‑optimization techniques and related framework applications

Scan QR code to inquire and purchase tickets