Skip to content

Pet dos #714

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Draft
wants to merge 38 commits into
base: main
Choose a base branch
from
Draft

Pet dos #714

wants to merge 38 commits into from

Conversation

jwa7
Copy link
Member

@jwa7 jwa7 commented Aug 12, 2025 •
edited by github-actions bot
Loading

Please go the the Preview tab and select the appropriate PR template:

📚 Documentation preview 📚: https://metatrain--714.org.readthedocs.build/en/714/

HowWeiBin and others added 29 commits May 13, 2025 22:01
@HowWeiBin
Updated for latest release of metatrain
7c61bf5
@HowWeiBin
Fixed gradient not taking into account excess target
4423ce0
@HowWeiBin
Fixed eval function
4ac0877
@HowWeiBin
Fixed sampler bug
ea968a7
@SanggyuChong
num_subtarget > 1 for LLPR
56ed4d8
@SanggyuChong
remove debug prints, update to metatomic
5aba703
@HowWeiBin
Optimized gradient calculation and now reports mean loss instead of t…
c2aac9b 
…otal loss
@HowWeiBin
Fixed Training loss not using mean losses but total losses
c98f669
@HowWeiBin
Fixed device mismatch
f2279e4
@HowWeiBin
Trying to fixc issue with evaluation script
bed409a
@HowWeiBin
Fixed validation criteria
7eedbbc
@HowWeiBin
Fixed restart issue
d6d523d
@HowWeiBin
Modified trainer to save validation set
5a76632
@HowWeiBin
Fixed validation set savepath
5978c42
@HowWeiBin
Removed debug message
039e9b1
@HowWeiBin
Detach predictions for memory purposes
ddab09f
@HowWeiBin
Removed saving of vaLidation predictions
b4b8256
@SanggyuChong
account for model device, batch shift calculation in LLPR
26557d4
@SanggyuChong
Merge branch 'pet-dos' of github.com:metatensor/metatrain into pet-dos
3d33024
@HowWeiBin
Allowed the option to not have a permanent subset
d3002d1
@HowWeiBin
Removed reading target during eval.py since evaluation is disabled an…
498118f 
…yway
@SanggyuChong
fix shift bug in LLPR for DOS
edabf67
@SanggyuChong
Merge branch 'pet-dos' of github.com:metatensor/metatrain into pet-dos
330d7ba
@SanggyuChong
fix bug with device and dtype
6bf6899
@HowWeiBin
bugfix for multigpu
46416e4
@HowWeiBin
Fix code to not make model output everything
d1a300d
@HowWeiBin
Further fixes to make the model output only necessary stuff
d74f190
@SanggyuChong
very rough prototyping to make LLPR ensemble work for DOS
e71d368
@SanggyuChong
revamp ensemble to be part of model
b520388
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@frostedoyster frostedoyster Awaiting requested review from frostedoyster frostedoyster will be requested when the pull request is marked ready for review frostedoyster is a code owner

@SanggyuChong SanggyuChong Awaiting requested review from SanggyuChong SanggyuChong will be requested when the pull request is marked ready for review SanggyuChong is a code owner

@DavideTisi DavideTisi Awaiting requested review from DavideTisi DavideTisi will be requested when the pull request is marked ready for review DavideTisi is a code owner

@abmazitov abmazitov Awaiting requested review from abmazitov abmazitov will be requested when the pull request is marked ready for review abmazitov is a code owner

At least 1 approving review is required to merge this pull request.

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@jwa7 @HowWeiBin @SanggyuChong