Comment by salter2
6 days ago
Perhaps something similar to speculative decoding.
Speculating Experts Accelerates Inference for Mixture-of-Experts: https://arxiv.org/abs/2603.19289
6 days ago
Perhaps something similar to speculative decoding.
Speculating Experts Accelerates Inference for Mixture-of-Experts: https://arxiv.org/abs/2603.19289
No comments yet
Contribute on Hacker News ↗