Skip to content
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.

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

release: merge set of changes for v1.2.0 #299

Merged
merged 79 commits into from
Aug 14, 2024
Merged

release: merge set of changes for v1.2.0 #299

merged 79 commits into from
Aug 14, 2024

Conversation

willmj
Copy link
Collaborator

@willmj willmj commented Aug 14, 2024

Description of the change

v1.2.0 release

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

willmj and others added 30 commits July 29, 2024 09:52
* initial code for deleting lm_head

Signed-off-by: Anh-Uong <[email protected]>

* fix logic for copying checkpoint

Signed-off-by: Anh-Uong <[email protected]>

* fix check that embed_tokens and lm_head weights are the same

Signed-off-by: Anh-Uong <[email protected]>

* fix warning assertion

Signed-off-by: Anh-Uong <[email protected]>

* fix lm_head check, remove test

Signed-off-by: Anh-Uong <[email protected]>

* small fixes from code review

Signed-off-by: Anh-Uong <[email protected]>

* fmt

Signed-off-by: Anh-Uong <[email protected]>

---------

Signed-off-by: Anh-Uong <[email protected]>
Co-authored-by: Anh-Uong <[email protected]>
Signed-off-by: Abhishek <[email protected]>
Signed-off-by: Angel Luu <[email protected]>
Signed-off-by: Angel Luu <[email protected]>
* feat: On save event added to callback

Signed-off-by: Padmanabha V Seshadri <[email protected]>

* fix: Removed additional bracket

Signed-off-by: Padmanabha V Seshadri <[email protected]>

* fix: Removed additional bracket

Signed-off-by: Padmanabha V Seshadri <[email protected]>

* fix: Format issues resolved

Signed-off-by: Padmanabha V Seshadri <[email protected]>

* fix: rebase with upstream and add new line

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

---------

Signed-off-by: Padmanabha V Seshadri <[email protected]>
Signed-off-by: Mehant Kammakomati <[email protected]>
Co-authored-by: Mehant Kammakomati <[email protected]>
* feat: All metric handling changes

Signed-off-by: Padmanabha V Seshadri <[email protected]>

* fix: Format issues

Signed-off-by: Padmanabha V Seshadri <[email protected]>

---------

Signed-off-by: Padmanabha V Seshadri <[email protected]>
* feat: Added the triggered login in the operation

Signed-off-by: Padmanabha V Seshadri <[email protected]>

* fix: Formatting issues

Signed-off-by: Padmanabha V Seshadri <[email protected]>

* fix: Added default config

Signed-off-by: Padmanabha V Seshadri <[email protected]>

* fix: Moved the variable to right scope

Signed-off-by: Padmanabha V Seshadri <[email protected]>

* fix: Checked added to validate config log level

Signed-off-by: Padmanabha V Seshadri <[email protected]>

* fix: Removed some unwanted log file

Signed-off-by: Padmanabha V Seshadri <[email protected]>

---------

Signed-off-by: Padmanabha V Seshadri <[email protected]>
* refactor code to preprocess datasets

Co-authored-by: Alex-Brooks <[email protected]>
Signed-off-by: Sukriti-Sharma4 <[email protected]>

* fix formatting

Co-authored-by: Alex-Brooks <[email protected]>
Signed-off-by: Sukriti-Sharma4 <[email protected]>

* allow input/output in validate args

Co-authored-by: Alex-Brooks <[email protected]>
Signed-off-by: Sukriti-Sharma4 <[email protected]>

* format input/output JSON and mask

Co-authored-by: Alex-Brooks <[email protected]>
Signed-off-by: Sukriti-Sharma4 <[email protected]>

* function to return suitable collator

Co-authored-by: Alex-Brooks <[email protected]>
Signed-off-by: Sukriti-Sharma4 <[email protected]>

* add tests for SFT Trainer input/output format

Co-authored-by: Alex-Brooks <[email protected]>
Signed-off-by: Sukriti-Sharma4 <[email protected]>

* remove unused functions

Co-authored-by: Alex-Brooks <[email protected]>
Signed-off-by: Sukriti-Sharma4 <[email protected]>

* add eos token to input/output format

Signed-off-by: Sukriti-Sharma4 <[email protected]>

* fix tests

Signed-off-by: Sukriti-Sharma4 <[email protected]>

* improve docstrings

Signed-off-by: Sukriti-Sharma4 <[email protected]>

* keeping JSON keys constant

Signed-off-by: Sukriti-Sharma4 <[email protected]>

* support for input/output format

Signed-off-by: Sukriti-Sharma4 <[email protected]>

* formatting fixes

Signed-off-by: Sukriti-Sharma4 <[email protected]>

* update rEADME formats

Signed-off-by: Sukriti-Sharma4 <[email protected]>

* formatting README

Signed-off-by: Sukriti-Sharma4 <[email protected]>

---------

Signed-off-by: Sukriti-Sharma4 <[email protected]>
Co-authored-by: Alex-Brooks <[email protected]>
Signed-off-by: Harikrishnan Balagopal <[email protected]>
* Add a dockerfile argument at the end of final layer to enable aimstack.
Currenlty guarded by a dockerfile argument.

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

* Set the default value of ENABLE_AIM to false

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

---------

Signed-off-by: Dushyant Behl <[email protected]>
willmj and others added 21 commits August 6, 2024 17:47
* Add unit test to verify target_modules defaults correctly

Signed-off-by: Will Johnson <[email protected]>

* Add sft_trainer.main test to ensure target modules properly default for LoRA when set to None from CLI

Signed-off-by: Will Johnson <[email protected]>

* fmt

Signed-off-by: Will Johnson <[email protected]>

* Use model_args instead of importing, fix nits

Signed-off-by: Will Johnson <[email protected]>

* Add test to ensure target_modules defaults to None in job config

Signed-off-by: Will Johnson <[email protected]>

* Add additional check, fix nits

Signed-off-by: Will Johnson <[email protected]>

---------

Signed-off-by: Will Johnson <[email protected]>
Signed-off-by: Abhishek <[email protected]>
Signed-off-by: Padmanabha V Seshadri <[email protected]>
Signed-off-by: Abhishek <[email protected]>
Signed-off-by: Abhishek <[email protected]>
Signed-off-by: Abhishek <[email protected]>
Fix: Removal of transformers logger and addition of python native logger
…p, on_substep_end (#293)

Signed-off-by: Padmanabha V Seshadri <[email protected]>
Signed-off-by: James Busche <[email protected]>
Co-authored-by: Anh Uong <[email protected]>
* add save_model_dir flag for final checkpoint

Signed-off-by: Anh-Uong <[email protected]>

* remove output_dir logic, add save method

Signed-off-by: Anh-Uong <[email protected]>

* update accelerate_launch, remove save tokenizer

Signed-off-by: Anh-Uong <[email protected]>

* fix: put back creation of .complete file

Signed-off-by: Anh-Uong <[email protected]>

* fix failing tests and add new ones

Signed-off-by: Anh-Uong <[email protected]>

* tests: add sft_trainer test to train and save

- small refactor of tests

Signed-off-by: Anh-Uong <[email protected]>

* add docs on saving checkpoints and fix help msg

Signed-off-by: Anh-Uong <[email protected]>

* update example and note best checkpoint

Signed-off-by: Anh-Uong <[email protected]>

* changes based on PR review

Signed-off-by: Anh-Uong <[email protected]>

* add logging to save, fix error out properly

Signed-off-by: Anh-Uong <[email protected]>

---------

Signed-off-by: Anh-Uong <[email protected]>
@anhuong
Copy link
Collaborator

anhuong commented Aug 14, 2024

@willmj I looked through all of the commits and they look good except for the most recent commit from you. This does not look to be signed,

Merge tag 'v1.2.0-rc.1' into release

Some of the PRs that were done via merge commit instead of squash commit also have lots of commits spread out which makes it harder to read. So we should do squash commits going forward.

@anhuong anhuong changed the title v1.2.0 rc1 release: merge set of changes for v1.2.0 Aug 14, 2024
@anhuong anhuong merged commit 49305ad into release Aug 14, 2024
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

10 participants