Commit Graph

12 Commits

Author SHA1 Message Date
salman
143dea4753 FSDPConfig (#3170) 2025-10-10 14:44:25 +01:00
Wing Lian
af92151a7b FSDP2 fix validation and add tests (#2910)
* fix validation and add tests

* remove debugging and add more tests

* remove migrate_fsdp
2025-07-14 09:25:44 -04:00
salman
d6e4a611e5 FSDP1 -> FSDP2 (#2760)
* FSDP2 args migration implementation

This commit implements the migration to FSDP2 arguments including:
- FSDP2 support with LoRA training
- DPO integration with FSDP2
- Model loading fixes and refactoring
- CPU offloading and PEFT handling
- Test updates and CI improvements
- Bug fixes for dtype errors and various edge cases
2025-07-12 15:18:01 +01:00
Wing Lian
40f4ea23ab replace references to random 68m model w 135m smollm2 (#2570) [skip ci]
* replace references to random 68m model w 135m smollm2

* use AutoTokenizer for smollm2
2025-04-28 10:08:07 -04:00
Dan Saunders
c907ac173e adding pre-commit auto-update GH action and bumping plugin versions (#2428)
* adding pre-commit auto-update GH action and bumping plugin versions

* running updated pre-commit plugins

* sorry to revert, but pylint complained

* Update .pre-commit-config.yaml

Co-authored-by: Wing Lian <wing.lian@gmail.com>

---------

Co-authored-by: Dan Saunders <dan@axolotl.ai>
Co-authored-by: Wing Lian <wing.lian@gmail.com>
2025-03-21 11:02:43 -04:00
Wing Lian
fd3b80716a remove fastchat and sharegpt (#2021)
* remove fastchat and sharegpt

* remove imports

* remove more fastchat imports

* chore: remove unused functions

* feat: remove sharegpt and deprecate from docs

* chore: remove unused sharegpt checks

* fix: remove sharegpt type from tests

* feat: add sharegpt deprecation error

* feat: update readme

---------

Co-authored-by: NanoCode012 <nano@axolotl.ai>
2024-11-08 13:45:49 -05:00
Wing Lian
0f985e12fe more fixes 20240228 (#1342) [skip ci]
* add missing evals_per_epoch setting

* more pydantic fixes

* more fixes

* move test from normalization to validation

* increase eval size for sample packing tests
2024-02-28 12:57:45 -05:00
Wing Lian
782b6a4216 set fp16 to false if bf16, update bf16: auto in example YAMLs (#1122) [skip ci]
* set fp16 to false if bf16, update bf16: auto in example YAMLs

* unset fp16 so that it fallsback properly if bf16 isn't available

* Update README.md [skip-ci]

Co-authored-by: NanoCode012 <kevinvong@rocketmail.com>

* test that bf16 disables fp16

---------

Co-authored-by: NanoCode012 <kevinvong@rocketmail.com>
2024-01-22 18:44:01 -05:00
Simon Hällqvist
086561326f Enable or disable bf16 support based on availability (#1116) 2024-01-14 12:06:56 -05:00
Wing Lian
0ce1a6594e update sharegpt conversations when chatml chat template is set (#1075) [skip ci]
* update sharegpt conversations when chatml chat template is set

* add info log when updating sharegpt/chatml conversation
2024-01-10 00:49:07 -05:00
Wing Lian
2d8def68dc simplify by removing duplicate base_model_config (#772) 2023-10-23 01:42:38 -04:00
Wing Lian
ca84cca2c0 convert exponential notation lr to floats (#771) 2023-10-22 15:37:03 -04:00