Hi authors,
I'm currently trying to run the experiments and noticed that the data/ directory is meant to contain self_distilled_data for training the drafter model. I've looked through the repository and the paper, but I couldn't find specific instructions or a script.
Could you please provide some guidance or a script on how to generate the self_distilled_data? Any help would be greatly appreciated.
Thanks again for your brilliant work!
Hi authors,
I'm currently trying to run the experiments and noticed that the
data/directory is meant to containself_distilled_datafor training the drafter model. I've looked through the repository and the paper, but I couldn't find specific instructions or a script.Could you please provide some guidance or a script on how to generate the
self_distilled_data? Any help would be greatly appreciated.Thanks again for your brilliant work!