Add a --write_distogram flag to save the distogram in .npz (AF3 mimic) - #699
Open
Jahaciel12 wants to merge 1 commit into
Open
Add a --write_distogram flag to save the distogram in .npz (AF3 mimic)#699Jahaciel12 wants to merge 1 commit into
Jahaciel12 wants to merge 1 commit into
Conversation
The distogram head is run on every prediction but its output stays inside the forward pass, so the predicted distance distribution is not reachable from the CLI. Carry the logits through predict_step and dump them next to the models, as predictions/<id>/distogram_<id>.npz, together with the annotation of their token axes. The distogram is a trunk output, shared by the diffusion samples of a prediction, so one file is written per input rather than one per sample. Boltz-2 predicts a stack of distograms where Boltz-1 predicts a single one, and the stacking axis is dropped when it holds one distogram.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The distogram head is run on every prediction but its output stays inside the forward pass, so the predicted distance distribution is not reachable. Carry the logits through predict_step and dump them next to the models, as predictions//distogram_.npz, together with the annotation of their token axes.
The distogram is a trunk output, shared by the diffusion samples of a prediction, so one file is written per input rather than one per sample. Boltz-2 predicts a stack of distograms where Boltz-1 predicts a single one, and the stacking axis is dropped when it holds one distogram.