Commit graph

1,556 commits

Author SHA1 Message Date
Daniel Han
bb0d3f900f Update llama.py 2025-02-20 07:36:23 -08:00
Daniel Han
ebf8187d70 Update llama.py 2025-02-20 07:25:31 -08:00
Daniel Han
5d4182c774 Update llama.py 2025-02-20 07:01:14 -08:00
Daniel Han
d30bbfd3cf Merge branch 'main' into nightly 2025-02-20 05:14:57 -08:00
Daniel Han
02c19d97f9 bug fix 2025-02-20 04:47:12 -08:00
Daniel Han
7b6cd79c2f Memory Efficient GRPO (#1773)
* Update __init__.py

* Update loader.py

* Update rl.py

* Update rl.py

* Update _utils.py

* Update tokenizer_utils.py

* Update tokenizer_utils.py

* Better TRL handling

* Update rl.py

* Update tokenizer_utils.py

* Auto patching

* Update tokenizer_utils.py

* Update tokenizer_utils.py

* Update tokenizer_utils.py

* Update rl.py

* Update tokenizer_utils.py

* Update rl.py

* Update tokenizer_utils.py

* Update tokenizer_utils.py

* Update tokenizer_utils.py

* Update tokenizer_utils.py

* Update tokenizer_utils.py

* Update tokenizer_utils.py

* Update tokenizer_utils.py

* Update tokenizer_utils.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update tokenizer_utils.py

* Update rl.py

* Update rl.py

* Update rl.py

* max seq length

* Update rl.py

* Update rl.py

* Patching

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* NEFTune

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Extra replacements

* Update rl_replacements.py

* Update rl.py

* extra RL replacements

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update llama.py

* Update rl_replacements.py

* Update _utils.py

* Update loader_utils.py

* Update rl.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* autocast

* Update rl_replacements.py

* Update llama.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update llama.py

* Update rl_replacements.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update rl_replacements.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update pyproject.toml

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update llama.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update llama.py

* Update _utils.py

* Update llama.py

* Update _utils.py

* Update rl_replacements.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update rl_replacements.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* GRPO optimized

* Update rl.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Selective Log softmax

* Fix GRPO bsz

* Update rl.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Fix TRL

* Metrics GRPO

* Update rl_replacements.py

* Update rl_replacements.py

* No compile

* Update rl.py

* Remove docs

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl_replacements.py

* Update rl.py

* Update rl.py

* Update rl_replacements.py

* Update rl_replacements.py

* llama-quantize on WINDOWS WSL error fix - edit save.py (gguf saving breaks) (#1649)

* edit save.py to fix gguf saving breaks.

* add check for .exe or not exe file extension for linux and windows

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update llama.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update llama.py

* Update llama.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl.py

* Update rl.py

* Update rl_replacements.py

* Update rl.py

* Update rl.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* unsloth_num_chunks

* Update rl.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl_replacements.py

* Update rl_replacements.py

* Update rl_replacements.py (#1754)

Fix typo in comment: know -> now.

This was printed when running the Llama3.1_(8B)-GRPO.ipynb example notebook, so I'd expect others to run into it as well.

* Optional logits

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl_replacements.py

* Update rl.py

* Update rl.py

* Update rl.py

* Update rl.py

* fix an import error (#1767)

* fix an import error

* Delete .gitignore

* Update loader.py

* Update save.py

---------

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

* SamplingParams

* Convert mask to float (#1762)

* [Windows Support] Add latest `xformers` wheels to pyproject.toml (#1753)

* Add latest xformers

* Add a couple of lines to docs

* vLLMSamplingParams

* Update __init__.py

* default num_chunks == -1

* Versioning

---------

Co-authored-by: Gennadii Manzhos <105049664+everythingisc00l@users.noreply.github.com>
Co-authored-by: Seth Weidman <seth@sethweidman.com>
Co-authored-by: Nino Risteski <95188570+NinoRisteski@users.noreply.github.com>
Co-authored-by: Edd <68678137+Erland366@users.noreply.github.com>
Co-authored-by: Ben <6579034+versipellis@users.noreply.github.com>
2025-02-20 04:23:28 -08:00
Daniel Han
c0bdacd505 Versioning 2025-02-20 04:22:17 -08:00
Daniel Han
2ff2cdad63 default num_chunks == -1 2025-02-19 23:51:06 -08:00
Daniel Han
af709e1dd6 Update __init__.py 2025-02-19 23:45:07 -08:00
Daniel Han
f89ae2b255 vLLMSamplingParams 2025-02-19 23:43:52 -08:00
Daniel Han
0c6ee39481 Merge branch 'nightly' of https://github.com/unslothai/unsloth into nightly 2025-02-19 23:40:50 -08:00
Ben
5b816e735c [Windows Support] Add latest xformers wheels to pyproject.toml (#1753)
* Add latest xformers

* Add a couple of lines to docs
2025-02-19 23:40:07 -08:00
Edd
8e5aeacbee Convert mask to float (#1762) 2025-02-19 23:38:48 -08:00
Daniel Han
f9ed548b54 SamplingParams 2025-02-19 23:37:52 -08:00
Nino Risteski
d95f5782d0 fix an import error (#1767)
* fix an import error

* Delete .gitignore

* Update loader.py

* Update save.py

---------

Co-authored-by: Daniel Han <danielhanchen@gmail.com>
2025-02-19 23:32:25 -08:00
Daniel Han
abeb373878 Update rl.py 2025-02-19 23:27:16 -08:00
Daniel Han
309fd6cd90 Merge branch 'main' into nightly 2025-02-19 23:24:21 -08:00
Daniel Han
25988683f8 Update README.md (#1768) 2025-02-19 23:24:05 -08:00
Daniel Han
e83d31632c Update rl.py 2025-02-19 23:23:39 -08:00
Daniel Han
b15b98c697 Update rl.py 2025-02-19 23:15:13 -08:00
Daniel Han
381b025779 Update rl.py 2025-02-19 23:06:18 -08:00
Daniel Han
18dfabd8e5 Update rl_replacements.py 2025-02-19 22:03:47 -08:00
Daniel Han
078d623b5d Update rl.py 2025-02-19 21:41:17 -08:00
Daniel Han
8368734e48 Update rl.py 2025-02-19 17:58:25 -08:00
Daniel Han
cc7188e7d2 Update rl.py 2025-02-19 16:40:37 -08:00
Daniel Han
8a703abe10 Update rl.py 2025-02-19 16:38:44 -08:00
Daniel Han
94264f7017 Update rl.py 2025-02-19 16:37:12 -08:00
Daniel Han
d9b31ee764 Update rl.py 2025-02-19 12:51:22 -08:00
Daniel Han
5ee0990347 Update rl.py 2025-02-19 12:47:15 -08:00
Daniel Han
17245595fa Update rl.py 2025-02-19 03:41:51 -08:00
Daniel Han
10b236d634 Optional logits 2025-02-19 02:23:00 -08:00
Seth Weidman
9f354c1447 Update rl_replacements.py (#1754)
Fix typo in comment: know -> now.

This was printed when running the Llama3.1_(8B)-GRPO.ipynb example notebook, so I'd expect others to run into it as well.
2025-02-19 02:12:07 -08:00
Daniel Han
550cddefb5 Update rl_replacements.py 2025-02-18 01:57:18 -08:00
Daniel Han
606fc099f3 Update rl_replacements.py 2025-02-18 00:17:07 -08:00
Daniel Han
f444c7517d Update rl.py 2025-02-18 00:13:26 -08:00
Daniel Han
9455ef40c9 Update rl.py 2025-02-18 00:09:11 -08:00
Daniel Han
fa831423b3 Update rl.py 2025-02-18 00:05:40 -08:00
Daniel Han
afa5f08c45 Update rl.py 2025-02-18 00:01:09 -08:00
Daniel Han
3b5c0d3c74 Update rl.py 2025-02-17 23:57:57 -08:00
Daniel Han
72999b3885 Update rl_replacements.py 2025-02-17 23:47:52 -08:00
Daniel Han
cf0f907321 Update rl_replacements.py 2025-02-17 22:30:20 -08:00
Daniel Han
6bac0780bc Update rl_replacements.py 2025-02-17 22:30:13 -08:00
Daniel Han
c355a7457d Update rl.py 2025-02-17 22:24:57 -08:00
Daniel Han
e554eb6f28 unsloth_num_chunks 2025-02-17 22:17:11 -08:00
Daniel Han
e77d21688d Update rl_replacements.py 2025-02-17 21:17:58 -08:00
Daniel Han
26aca36db8 Update rl_replacements.py 2025-02-17 21:10:26 -08:00
Daniel Han
46f470bea6 Update rl_replacements.py 2025-02-17 21:08:32 -08:00
Daniel Han
68d7c9941d Update rl_replacements.py 2025-02-17 21:03:44 -08:00
Daniel Han
2bba4e02d3 Update rl_replacements.py 2025-02-17 20:38:18 -08:00
Daniel Han
673325dea7 Update rl_replacements.py 2025-02-17 20:21:20 -08:00