Keynote: Anyscale Product Demo
summary
Demo builds a movie recommender end-to-end: color-palette KNN search with FAISS and a BERT-based plot model fine-tuned via Ray Tune, then deployed with Ray Serve on AnyScale across GCP and AWS. Code runs locally, scales to hundreds of cores (demo hits ~180 cores) and GPU nodes (shown as Tesla T4s), uses auto sync to keep code in sync, exposes a retrainer actor that yields a new logistic ranker at ~87% accuracy and rolls it out with zero downtime.
AI-written summary · about
chapters (9)
- 0:00Start
- 5:55Extract Color Palettes
- 8:10Progress Bar Utility
- 9:25Auto Sync
- 10:11Ray Dashboard
- 12:15Recap
- 12:50Deep Learning Model
- 14:44Ecosystem Integration
- 24:25Production Monitoring
related talks
27m[Opening Keynote] Anyscale Demo: Machine Learning Application from Dev to Prod
16mKeynote: An Open-Source Stack for AI Compute: Kubernetes + Ray + PyTorch + vLLM
30mScaling Machine Learning at Discord: Our Journey with Ray and Anyscale
21mHow Ray and Anyscale Make it Easy to do Massive-scale ML on Aerial Imagery
53mKeynote Interview
23mKeynote
31mHosting AI models with Ray and GKE
05mInterview