4 ms·
There is no reason that it couldn’t be beneficial for training though.
by Zacharias030 1y ago
There is no reason that it couldn’t be beneficial for training though.
- cubefox 1y agoExcept that speculative decoding is de facto only an inference time optimization. But the H-Net architecture from the previous reference, which doesn't require tokens or speculative decoding, does something similar both for inference and training.
- Zacharias030 1y agoYes, but the discussion is about Multi-Token Prediction (Gloeckle et al. 2024) which is only incidentally useful for speculative decoding.