ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
3,105 commits 391 branches 58 tags 304 MiB
Commit graph

3,105 commits

This branch
This branch
All branches
Author SHA1 Message Date
electroglyph
6cc710e6e3 Merge branch 'unslothai:main' into FST 2025-12-19 00:05:08 -08:00
Daniel Han
4f7651c11c FunctionGemma December-2025 2025-12-18 09:27:46 -08:00
DoubleMathew
ac0835d49f Fix Deepseek OCR Lora Model Load (#3738) 
* fix deepseek ocr lora_model load: trust_remote_code option

check for import error in autoconfig/peftconfig from_pretrained error

handle import

* Apply suggestion from @gemini-code-assist[bot]

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

* Apply suggestion from @gemini-code-assist[bot]

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

* Apply suggestion from @gemini-code-assist[bot]

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

* Apply suggestion from @gemini-code-assist[bot]

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

---------

Co-authored-by: Daniel Han <danielhanchen@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
 2025-12-18 04:09:17 -08:00
pre-commit-ci[bot]
f8c451e22a [pre-commit.ci] auto fixes from pre-commit.com hooks 
for more information, see https://pre-commit.ci
 2025-12-17 11:39:08 +00:00
electroglyph
8927715168 add super hacky mpnet patch from hell 2025-12-17 03:38:55 -08:00
Daniel Han
0f5bf8833c Nightly (#3737) 
* Update _utils.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* [FIX] [Transformers] VLM input embeds fix for gradients (#3715)

* Fix get_input_embeds call for VLMs

* patch input_require_grads instead

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* cleanup old patch

* cleanup old patch

* cleanup

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Apply suggestion from @danielhanchen

* use logger instead of prints

* Move unsloth present set

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>

* Update rope_embedding.py

* Fixes

* Update _utils.py

* Update import_fixes.py

* Update rl_replacements.py

* fix_openenv_no_vllm

* Fix

* Update __init__.py

* Update __init__.py

* Update __init__.py

* Update import_fixes.py

* Update import_fixes.py

* Update import_fixes.py

* logger

* Update __init__.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Update __init__.py

* Update import_fixes.py

* Update __init__.py

* Update import_fixes.py

* Update import_fixes.py

* Update import_fixes.py

* Update import_fixes.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Update import_fixes.py

* Update unsloth/import_fixes.py

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

* Update save.py

* [fbgemm] Silence tma fbgemm (#3735)

* Silence fbgemm TMA print

Also safer .push_to_hub

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

* Update loader.py

* Update save.py

* Update save.py

* Update _utils.py

* Update _utils.py

* Diffusers warnings

* Update pyproject.toml

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Datta Nimmaturi <venkatadattasainimmaturi@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
 2025-12-17 03:31:48 -08:00
pre-commit-ci[bot]
8cb8ccc80f [pre-commit.ci] auto fixes from pre-commit.com hooks 
for more information, see https://pre-commit.ci
 2025-12-17 06:53:37 +00:00
electroglyph
d51dbded19 propagate trust_remote_code properly 2025-12-16 22:52:51 -08:00
Daniel Han
935b133606 Update FUNDING.yml (#3736) 2025-12-16 21:36:25 -08:00
Daniel Han
96b49b91b5 Bug fixes (#3734) 
* Update _utils.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* [FIX] [Transformers] VLM input embeds fix for gradients (#3715)

* Fix get_input_embeds call for VLMs

* patch input_require_grads instead

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* cleanup old patch

* cleanup old patch

* cleanup

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Apply suggestion from @danielhanchen

* use logger instead of prints

* Move unsloth present set

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>

* Update rope_embedding.py

* Fixes

* Update _utils.py

* Update import_fixes.py

* Update rl_replacements.py

* fix_openenv_no_vllm

* Fix

* Update __init__.py

* Update __init__.py

* Update __init__.py

* Update import_fixes.py

* Update import_fixes.py

* Update import_fixes.py

* logger

* Update __init__.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Update __init__.py

* Update import_fixes.py

* Update __init__.py

* Update import_fixes.py

* Update import_fixes.py

* Update import_fixes.py

* Update import_fixes.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Update import_fixes.py

* Update unsloth/import_fixes.py

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Datta Nimmaturi <venkatadattasainimmaturi@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
 2025-12-16 20:52:57 -08:00
Daniel Han
28f742ee19 Update import_fixes.py 2025-12-16 16:58:03 -08:00
Daniel Han
3600b5341d Update import_fixes.py 2025-12-16 16:57:38 -08:00
Daniel Han
002f2a2157 Update import_fixes.py 2025-12-16 15:46:27 -08:00
electroglyph
4aec1a4891 revert save_pretrained_merged change 2025-12-16 14:07:47 -08:00
electroglyph
3b39be0150 comment 2025-12-16 13:38:31 -08:00
electroglyph
cd02c09916 rename read_pooling_mode, load modules instead of hard-coding em 2025-12-16 13:24:26 -08:00
electroglyph
da42e16505 fix _save_pretrained_merged 2025-12-16 04:04:39 -08:00
electroglyph
3acea66cc5 add save_pretrained_merged method which gets modules and config 2025-12-16 03:05:05 -08:00
electroglyph
a589ec34d1 refactor pooling detection, add missing pooling types 2025-12-16 02:50:37 -08:00
electroglyph
8481327d12 force disable FP8 loading 2025-12-16 02:32:58 -08:00
electroglyph
ca77709db7 add some comments 2025-12-16 02:10:02 -08:00
electroglyph
a099456372 refactor do_lower_case 2025-12-16 02:01:31 -08:00
Daniel Han
4e04e56b82 Update rl.py 2025-12-15 22:43:16 -08:00
pre-commit-ci[bot]
6aff56a61a [pre-commit.ci] pre-commit autoupdate (#3731) 
updates:
- [github.com/astral-sh/ruff-pre-commit: v0.14.8 → v0.14.9](https://github.com/astral-sh/ruff-pre-commit/compare/v0.14.8...v0.14.9)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
 2025-12-15 17:54:15 -08:00
pre-commit-ci[bot]
da9438b257 [pre-commit.ci] auto fixes from pre-commit.com hooks 
for more information, see https://pre-commit.ci
 2025-12-15 05:42:38 +00:00
electroglyph
15ead93619 unsloth-zoo patch only fixed usage for XLMRobertaForMaskedLM, this is a fix for XLMRobertaModel 2025-12-14 21:42:25 -08:00
electroglyph
c8be8fff88 Merge branch 'unslothai:main' into FST 2025-12-14 20:41:01 -08:00
Michael Han
7201ba3919 Update README.md 2025-12-13 16:44:44 -08:00
Daniel Han
2fc67f463a Nightly (#3720) 
* Update _utils.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* [FIX] [Transformers] VLM input embeds fix for gradients (#3715)

* Fix get_input_embeds call for VLMs

* patch input_require_grads instead

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* cleanup old patch

* cleanup old patch

* cleanup

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Apply suggestion from @danielhanchen

* use logger instead of prints

* Move unsloth present set

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>

* Update rope_embedding.py

* Fixes

* Update _utils.py

* Update import_fixes.py

* Update rl_replacements.py

* fix_openenv_no_vllm

* Fix

* Update __init__.py

* Update __init__.py

* Update __init__.py

* Update import_fixes.py

* Update import_fixes.py

* Update import_fixes.py

* logger

* Update __init__.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Update __init__.py

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Datta Nimmaturi <venkatadattasainimmaturi@gmail.com>
 2025-12-12 05:53:08 -08:00
Scott Roy
6451e5cae5 Update torchao save (#3679) 
* Update torchao save

* up

* up

* up

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Apply suggestion from @danielhanchen

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
 2025-12-12 04:07:02 -08:00
pre-commit-ci[bot]
7c28830980 [pre-commit.ci] auto fixes from pre-commit.com hooks 
for more information, see https://pre-commit.ci
 2025-12-12 11:45:24 +00:00
electroglyph
855f719e76 Gemini code review suggestions 2025-12-12 03:45:13 -08:00
pre-commit-ci[bot]
3c97fbcdfa [pre-commit.ci] auto fixes from pre-commit.com hooks 
for more information, see https://pre-commit.ci
 2025-12-12 11:27:02 +00:00
electroglyph
c483c3591c add FastSentenceTransformer 2025-12-12 03:22:54 -08:00
Dan Saunders
e21640e905 Mistral packing, train on completions only, simplifications (#3709) 
* pipe kwargs through mistral model

* simplify / bugfix

* bugfix for train_on_completions_only

* wire up is_unsupported_model

* nits, edge cases
 2025-12-10 23:15:59 -08:00
Lei Zhenyuan
21d34d8228 [intel] skip xpu fbgemm fp8 (#3625) 
* skip xpu fbgemm fp8

* Apply suggestion from @gemini-code-assist[bot]

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Apply suggestion from @danielhanchen

---------

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
 2025-12-10 21:13:29 -08:00
Michael Han
2f51d6b457 Padding free packing update 2025-12-10 21:12:13 -08:00
Michael Han
2412f78bd2 Adding new padding free packing support 2025-12-10 21:10:19 -08:00
Dan Saunders
85852fd902 update TRL filter (#3707) 
* update TRL filter

* both filters

* Apply suggestion from @danielhanchen

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: Daniel Han <danielhanchen@gmail.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
 2025-12-10 07:49:52 -08:00
Daniel Han
427656d891 Gemma issue 2025-12-10 06:25:49 -08:00
Daniel Han
2ecf368b53 Update _utils.py 2025-12-10 06:11:14 -08:00
Daniel Han
1e08afbca1 Update trainer.py 2025-12-10 06:11:03 -08:00
Daniel Han
9a1772b570 Nightly (#3706) 
* Update _utils.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
 2025-12-10 04:13:13 -08:00
Daniel Han
9ac14e4a33 Update import_fixes.py 2025-12-10 04:05:14 -08:00
Daniel Han
b586930726 Update _utils.py 2025-12-10 03:48:01 -08:00
Datta Nimmaturi
b0b154fbca [FIX] fbgemm version check (#3704) 
* fbgemm version check

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* safer version check

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Add check for torchvision-torch compatibility

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* refactor package check logic

* Remove logs and enforce torch

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
 2025-12-10 03:46:30 -08:00
Dan Saunders
89b042f23b Auto-enable padding-free SFT (#3672) 
* implement (sdpa, xformers, fa2) sample packing

* attention dispatching

* ddp working OOTB with CLI

* packed SWA and softcap support

* enable batch flattening

* LGPL license headers

* mask packed sequence boundaries

* auto-enable sample packing

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Add explicit toggle for sample packing

* Add explicit toggle for sample packing

* Update __init__.py

* Update unsloth/kernels/rope_embedding.py

* Update unsloth/kernels/rope_embedding.py

* remove grad output clones; restore deleted FastLanguageModel arg

* fix

* restore rope embedding clones

* xformers mask cache

* implement (sdpa, xformers, fa2) sample packing

* attention dispatching

* ddp working OOTB with CLI

* packed SWA and softcap support

* enable batch flattening

* LGPL license headers

* mask packed sequence boundaries

* auto-enable sample packing

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Add explicit toggle for sample packing

* Add explicit toggle for sample packing

* Update __init__.py

* Update unsloth/kernels/rope_embedding.py

* Update unsloth/kernels/rope_embedding.py

* remove grad output clones; restore deleted FastLanguageModel arg

* fix

* restore rope embedding clones

* xformers mask cache

* add back accidental deletion

* Update unsloth/kernels/rope_embedding.py

Co-authored-by: Daniel Han <danielhanchen@gmail.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* fix merge conflicts

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Add **kwargs

* add back clobbered

* Update rope_embedding.py

* Update rope_embedding.py

* simplify trl warnings filter

* docstring

* nit

* bugfix

* add padding-free seqlen metadata

* auto-enable padding free

* gemma2 disable

* Apply suggestion from @danielhanchen

* Update trainer.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Update trainer.py

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
 2025-12-10 03:07:29 -08:00
Dan Saunders
538d2c0719 SFT sample packing (#3566) 
* implement (sdpa, xformers, fa2) sample packing

* attention dispatching

* ddp working OOTB with CLI

* packed SWA and softcap support

* enable batch flattening

* LGPL license headers

* mask packed sequence boundaries

* auto-enable sample packing

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Add explicit toggle for sample packing

* Add explicit toggle for sample packing

* Update __init__.py

* Update unsloth/kernels/rope_embedding.py

* Update unsloth/kernels/rope_embedding.py

* remove grad output clones; restore deleted FastLanguageModel arg

* fix

* restore rope embedding clones

* xformers mask cache

* implement (sdpa, xformers, fa2) sample packing

* attention dispatching

* ddp working OOTB with CLI

* packed SWA and softcap support

* enable batch flattening

* LGPL license headers

* mask packed sequence boundaries

* auto-enable sample packing

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Add explicit toggle for sample packing

* Add explicit toggle for sample packing

* Update __init__.py

* Update unsloth/kernels/rope_embedding.py

* Update unsloth/kernels/rope_embedding.py

* remove grad output clones; restore deleted FastLanguageModel arg

* fix

* restore rope embedding clones

* xformers mask cache

* add back accidental deletion

* Update unsloth/kernels/rope_embedding.py

Co-authored-by: Daniel Han <danielhanchen@gmail.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* fix merge conflicts

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Add **kwargs

* add back clobbered

* Update rope_embedding.py

* Update rope_embedding.py

* simplify trl warnings filter

* docstring

* nit

* bugfix

* Apply suggestion from @danielhanchen

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Update unsloth/trainer.py

* Update unsloth/trainer.py

* Update unsloth/trainer.py

* Update unsloth/trainer.py

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
 2025-12-09 17:36:45 -08:00
Daniel Han
1b70a5f15e Update _utils.py (#3698) 2025-12-09 03:31:20 -08:00
Datta Nimmaturi
c8c8f168bb [Fix] [TRL] load_lora for multi line llm.chat/generate (#3696) 
* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Remove reload_weights rpc call from grpo trainer

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Use regex instead of static string

* patch openenv reload_weights call

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Better handle sleep and wakeup

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Reset indentation

* Handle multi line self.llm.chat better

* Use logger

* re-indent

* Stricter regex to replace wildcard

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Daniel Han <danielhanchen@gmail.com>
 2025-12-09 03:30:23 -08:00