Commit Graph
12 Commits
Author SHA1 Message Date
8f63f565c6 [utils] add merge_lora utility function (#227)
* add merge_lora utility function

* forward contrib credits from original script

* some changes

* make style

* fix tets

* finally fix tests

* Update tests/test_peft_model.py

* adapt from suggestions

* adapt

* Update src/peft/tuners/lora.py

Co-authored-by: Sourab Mangrulkar <13534540+pacman100@users.noreply.github.com>

* fix 8bit

* Update src/peft/tuners/lora.py

Co-authored-by: Sourab Mangrulkar <13534540+pacman100@users.noreply.github.com>

---------

Co-authored-by: edbeeching <edbeeching@users.noreply.github.com>
Co-authored-by: Sourab Mangrulkar <13534540+pacman100@users.noreply.github.com>
2023-03-30 13:45:37 +02:00
Younes Belkada df71b84341 [CI] Add more ci tests (#223)
* add more tests

* fix

* add generate tests

* make style

* fix test

* add -n

* skip llama
2023-03-29 15:28:38 +02:00
Younes Belkada 2632e7eba7 [CI] Add ci tests (#203)
* add ci tests

* fix some tests

* fix tests

* rename

* fix

* update tests

* try

* temp hotfix

* refactor tests

* Update .github/workflows/tests.yml

* fix test
2023-03-23 12:38:40 +01:00
Sourab Mangrulkar b9451ab458 fixing issues and quality ✨ 2023-03-07 14:04:19 +05:30
younesbelkada 36c7e3b441 apply suggestions 2023-02-14 11:50:55 +00:00
younesbelkada 0e80648010 add prepare_model_for_training 2023-02-14 11:12:37 +00:00
Sourab Mangrulkar 06e49c0a87 fixes 2023-01-30 13:31:01 +05:30
Sourab Mangrulkar 6cf2cf5dae fix hf hub util tests 2023-01-30 12:51:30 +05:30
younesbelkada 6c9534e660 adapt for other models 2023-01-29 11:18:31 +00:00
younesbelkada 22295c4278 adapt from code review
- remove `README`
- inherit from `dataclass`
- add new test
2023-01-29 10:49:31 +00:00
younesbelkada 634f3692d8 working v1
- push to hub method works
- add tests
- add config super class
- add Lora support for `from_pretrained`
2023-01-26 11:17:24 +00:00
younesbelkada 2cc7f2cbac add config tests 2023-01-26 10:12:51 +00:00