Resume checkpoint does not include loading embedding?

[para-zhou]

trafficstars

Hi, may I ask the implementation of resuming training from a checkpoint? To the best of my understanding, when specify "args.resume_checkpoint", there's no operation on loading the saved embedding. Is this a bug or something? Also I found there's no training on the embedding. Did I miss points or it is intended? Thank you very much.

[summmeer]

Hi, We didn't use the saved embedding. The word embedding params are built into the model, so the resume operation could load it.

[para-zhou]

Thank you for your reply. Btw, could you please release the checkpoints trained on other tasks such as open-domain QA? Thanks!

2023年1月13日 上午11:47，Sansa G @.***> 写道：

Hi, We didn't use the saved embedding. The word embedding params are built into the model, so the resume operation could load it.

— Reply to this email directly, view it on GitHubhttps://github.com/Shark-NLP/DiffuSeq/issues/19#issuecomment-1381280128, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AN2CS5ZUSIZB73T4MIYE3SDWSDF53ANCNFSM6AAAAAATVDHP6Q. You are receiving this because you authored the thread.Message ID: @.***>