1. Accelerating Production LLMs with Combined Token/Embedding Speculators
- Author
-
Wertheimer, Davis, Rosenkranz, Joshua, Parnell, Thomas, Suneja, Sahil, Ranganathan, Pavithra, Ganti, Raghu, and Srivatsa, Mudhakar
- Subjects
Computer Science - Computation and Language - Abstract
This technical report describes the design and training of novel speculative decoding draft models, for accelerating the inference speeds of large language models in a production environment. By conditioning draft predictions on both context vectors and sampled tokens, we can train our speculators to efficiently predict high-quality n-grams, which the base model then accepts or rejects. This allows us to effectively predict multiple tokens per inference forward pass, accelerating wall-clock inference speeds of highly optimized base model implementations by a factor of 2-3x. We explore these initial results and describe next steps for further improvements., Comment: Original upload 4/29/24, updated 6/6/24 with additional references to concurrent work
- Published
- 2024