Completed
11:44 – Guided Decoding with Speculative Inference
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Efficient Inference with Command R+ - Optimizing Speed and Cost for Enterprise AI
Automatically move to the next video in the Classroom when playback concludes
- 1 0:00 – Introduction to Command R+ Inference Optimization
- 2 0:55 – Sparse Attention Architecture & Sliding Window
- 3 2:21 – Speculative Decoding Overview
- 4 4:32 – Using Medusa for Parallel Token Prediction
- 5 6:29 – Evaluation and Training with W&B
- 6 7:54 – Synthetic vs. Original Data in Speculative Training
- 7 9:00 – Final Gains and Performance Tradeoffs
- 8 11:44 – Guided Decoding with Speculative Inference
- 9 14:29 – Dynamic Guided Decoding and FSM Integration
- 10 19:03 – Combining Guided Decoding with Speculative Tokens
