Skip to content

chore: Merge set of changes for release v3.1.0 #630

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
Merged
dushyantbehl merged 30 commits into from
Nov 10, 2025
Merged

chore: Merge set of changes for release v3.1.0 #630

dushyantbehl merged 30 commits into from
Nov 10, 2025

Conversation

@dushyantbehl
Copy link
Collaborator

@dushyantbehl dushyantbehl commented Nov 10, 2025

Description of the change

Related issue number

How to verify the PR

Was the PR tested

  • I have added >=1 unit test(s) for every new method I have added.
  • I have ensured all unit tests pass

dushyantbehl and others added 29 commits July 24, 2025 21:28
@dushyantbehl
Update offline-data-preprocessing.md (#589)
048a0e8 
Signed-off-by: Dushyant Behl <[email protected]>
@dushyantbehl
use default dataprocessor if one is not provided (#590)
7258505 
Signed-off-by: Dushyant Behl <[email protected]>
@dushyantbehl
fix data processing job exiting with failure even though successful (#…
5d49c76 
…593)

Signed-off-by: Dushyant Behl <[email protected]>
@romitjain @dushyantbehl
fix: Disable caching in transformers via use_cache flag to avoid un…
ef11bd1 
…ncessary memory overhead (#592)

* Added use_cache in the model args

Signed-off-by: romit <[email protected]>

* Removed model args use_cache

Signed-off-by: r0 <[email protected]>
Signed-off-by: romit <[email protected]>

* Update sft_trainer.py, default use_cache

Signed-off-by: r0 <[email protected]>
Signed-off-by: romit <[email protected]>

* Updated cache usage in vision models

Signed-off-by: romit <[email protected]>

---------

Signed-off-by: romit <[email protected]>
Signed-off-by: r0 <[email protected]>
Co-authored-by: Dushyant Behl <[email protected]>
@seshapad
fix: Rank is set to zero by default (#594)
5951c9a 
Signed-off-by: Padmanabha V Seshadri <[email protected]>
@dushyantbehl
upgrade trl (#601)
9c7858e 
Signed-off-by: Dushyant Behl <[email protected]>
@YashasviChaurasia
Add changes to support granite 4 models (#599)
607984c 
Signed-off-by: yashasvi <[email protected]>
@dushyantbehl
feat: Support gpt-oss class of models with flash attention 3 support (#…
7e261d2 
…603)

Signed-off-by: Dushyant Behl <[email protected]>
@dushyantbehl @praveenj83
feat: Restructure README (#598)
bc39f95 
* restructure README

Signed-off-by: Dushyant Behl <[email protected]>

* split readme

Signed-off-by: Dushyant Behl <[email protected]>

* Update README.md

Co-authored-by: Praveen Jayachandran <[email protected]>
Signed-off-by: Dushyant Behl <[email protected]>

* update readme

Signed-off-by: Dushyant Behl <[email protected]>

* Update README.md

Co-authored-by: Praveen Jayachandran <[email protected]>
Signed-off-by: Dushyant Behl <[email protected]>
Signed-off-by: Dushyant Behl <[email protected]>

---------

Signed-off-by: Dushyant Behl <[email protected]>
Signed-off-by: Dushyant Behl <[email protected]>
Co-authored-by: Praveen Jayachandran <[email protected]>
@YashasviChaurasia
fix: add default optim arg in training arg (#607)
f41eb2c 
Signed-off-by: yashasvi <[email protected]>
@dushyantbehl
fix typo in dockerfile
8232e63 
Signed-off-by: Dushyant Behl <[email protected]>
@dushyantbehl
fix typo which ignored qlora config
47e80ce 
Signed-off-by: Dushyant Behl <[email protected]>
@dushyantbehl
fix typos in data handler names
b6aa877 
Signed-off-by: Dushyant Behl <[email protected]>
@dushyantbehl
Merge pull request #610 from dushyantbehl/main
2949a3a 
fix: typos
@romitjain @dushyantbehl
fix: subclass Lora config from upstream peft.LoraConfig (#609)
abeb12c 
* Updated LoraConfig to subclass from peft.LoraConfig
Signed-off-by: romit <[email protected]>

* Added some fields under custom dataclass to let is pass through HfArgumentParser
Signed-off-by: romit <[email protected]>

* Lint and fmt fixes
Signed-off-by: romit <[email protected]>

* Updated config utils

Signed-off-by: romit <[email protected]>

* Update comment in LoraConfig

Signed-off-by: r0 <[email protected]>

* Lint changes

Signed-off-by: romit <[email protected]>

* Updated comment

Signed-off-by: romit <[email protected]>

---------

Signed-off-by: romit <[email protected]>
Signed-off-by: r0 <[email protected]>
Co-authored-by: Dushyant Behl <[email protected]>
@YashasviChaurasia
fix: update fms-accel to main (#608)
384d424 
Signed-off-by: yashasvi <[email protected]>
@dushyantbehl
Update advanced-data-preprocessing.md (#613)
d6dc4c9 
Fix image in README.

Signed-off-by: Dushyant Behl <[email protected]>
@YashasviChaurasia
feat: add ckpt conversion script fp32-bf16 (#614)
d9ee35f 
* feat: add ckpt conversion script fp32-bf16

Signed-off-by: yashasvi <[email protected]>

* feat: add remove optim func to checkpoint_util.py

Signed-off-by: yashasvi <[email protected]>

* feat: add inplace file deletion capability

Signed-off-by: yashasvi <[email protected]>

---------

Signed-off-by: yashasvi <[email protected]>
@YashasviChaurasia
feat: Allow chat template to be specified via a path in data config. (#…
f337875 
…615)

Signed-off-by: yashasvi <[email protected]>
@kmehant
feat: add online data mixing plugin (#612)
4ec1340 
feat: add odm plugin

Signed-off-by: Mehant Kammakomati <[email protected]>

---------

Signed-off-by: Mehant Kammakomati <[email protected]>
@kmehant
feat: Adopt resumption feature of online data mixing (#617)
452c13b 
* feat: resume functionality

Signed-off-by: Mehant Kammakomati <[email protected]>

* feat: resume functionality

Signed-off-by: Mehant Kammakomati <[email protected]>

* feat: resume functionality

Signed-off-by: Mehant Kammakomati <[email protected]>

* feat: resume functionality

Signed-off-by: Mehant Kammakomati <[email protected]>

* feat: resume functionality

Signed-off-by: Mehant Kammakomati <[email protected]>

* fix: refactor code

Signed-off-by: Mehant Kammakomati <[email protected]>

* fix: refactor code

Signed-off-by: Mehant Kammakomati <[email protected]>

* fix: refactor code

Signed-off-by: Mehant Kammakomati <[email protected]>

* fix: refactor sampling weight

Signed-off-by: Mehant Kammakomati <[email protected]>

---------

Signed-off-by: Mehant Kammakomati <[email protected]>
@YashasviChaurasia
feat: Alora migration to PEFT upstream. (#618)
5187516 
feat: Alora migration to PEFT upstream.

Signed-off-by: yashasvi <[email protected]>

---------

Signed-off-by: yashasvi <[email protected]>
@YashasviChaurasia
feat: ensure fms-acceleration callbacks run before TrainerController (#…
ebf5743 
…620)

Signed-off-by: yashasvi <[email protected]>
@YashasviChaurasia
fix: add on_init_end before adding tc callback (#621)
6f2134d 
Signed-off-by: yashasvi <[email protected]>
@YashasviChaurasia
fix: add hf compatible path update (#622)
7ff1ce9 
Signed-off-by: yashasvi <[email protected]>
@YashasviChaurasia
fix: avoid updating path kwarg (#624)
3344193 
Signed-off-by: yashasvi <[email protected]>
@YashasviChaurasia
fix: directly save final ckpt in save_model_dir (#626)
1e844d3 
Signed-off-by: yashasvi <[email protected]>
@dushyantbehl
Add free up disk space to gh runners (#628)
faea400 
Signed-off-by: Dushyant Behl <[email protected]>
@dushyantbehl
fix: image build failure due to flash attn (#629)
2706020 
* Add free up disk space to gh runners

Signed-off-by: Dushyant Behl <[email protected]>

* Build flash attn without cache to avoid problems

Signed-off-by: Dushyant Behl <[email protected]>

---------

Signed-off-by: Dushyant Behl <[email protected]>
@github-actions
Copy link

github-actions bot commented Nov 10, 2025

Thanks for making a pull request! 😃
One of the maintainers will review and advise on the next steps.

@github-actions github-actions bot added the chore label Nov 10, 2025
@dushyantbehl dushyantbehl merged commit 8659901 into release Nov 10, 2025
10 of 11 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@ashokponkumar ashokponkumar ashokponkumar approved these changes

@anhuong anhuong Awaiting requested review from anhuong anhuong is a code owner

@aluu317 aluu317 Awaiting requested review from aluu317 aluu317 is a code owner

@fabianlim fabianlim Awaiting requested review from fabianlim fabianlim is a code owner

@kmehant kmehant Awaiting requested review from kmehant kmehant is a code owner

Assignees

No one assigned

Labels

chore

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

7 participants

@dushyantbehl @ashokponkumar @romitjain @seshapad @YashasviChaurasia @kmehant