119 Commits

Author SHA1 Message Date
hiyouga
a66ff6052b add unittest
Former-commit-id: 8a1f0c5f922989e08a19c65de0b2c4afd2a5771f
2024-07-19 01:06:27 +08:00
hiyouga
6b16da0cd1 fix metrics #4786
Former-commit-id: 7d0c4bd394fc3cba197db1719f1164b9dd66ac21
2024-07-17 00:47:00 +08:00
hiyouga
746e9b352e support batch_eval_metrics, fix #4826
Former-commit-id: 3fe1df17188825f8a32fbe6a1294b4b532ce0c85
2024-07-17 00:33:00 +08:00
hiyouga
473b77f6a0 fix #4820
Former-commit-id: 8c0f8357e1eebee32010fe715554f1136b68b4ba
2024-07-15 22:32:07 +08:00
hiyouga
8ce43766c6 fix up
Former-commit-id: 43a56cb331fae899ca35b0c312730d4ab79d0c42
2024-07-15 01:04:56 +08:00
hoshi-hiyouga
07ea5796e5 Merge pull request #4691 from codemayq/feature-suppot-eval-dataset
add eval dataset support

Former-commit-id: 51eb379b44fad0336fc96c329ec98dc4528b9c2c
2024-07-15 01:00:34 +08:00
hzhaoy
7043219f00 tiny fix
Former-commit-id: 48be67c41eb394d276b41ca22b28e1ef10af4920
2024-07-12 00:28:44 +08:00
hoshi-hiyouga
a4e3b21b9e Update callbacks.py
Former-commit-id: 526376967deaad73b7ca11063a2e3f0c9a0add98
2024-07-10 13:32:20 +08:00
-.-
acde60b6d8 fix src/llamafactory/train/callbacks.py
Former-commit-id: c79a21aeaa5462770790887a6826d335e1ded5a2
2024-07-10 12:05:51 +08:00
hiyouga
88cf5c3cc2 fix #4731
Former-commit-id: 99e016ee552a551b52b6fcf3616cb57a5b927715
2024-07-10 11:32:36 +08:00
hiyouga
b778f3f949 fix ppo trainer
Former-commit-id: a03b2e5ef0d5d6b1b27753438745385d290cb211
2024-07-10 11:05:45 +08:00
hiyouga
970031b25c fix #4742
Former-commit-id: ae9cf84347878fcc462f35db941c14e1df104276
2024-07-09 23:24:24 +08:00
codingma
82e941ff61 1. add custom eval dataset support
2. merge load dataset and split dataset function


Former-commit-id: 963d97ba07e7efa3a4544c4d077283d9e112b3ad
2024-07-05 15:52:10 +08:00
hiyouga
a0df8be4e8 fix packing for eager/sdpa attn
Former-commit-id: 735a033ceb7f2da6da71d138ea091d8a665411a9
2024-07-04 01:52:43 +08:00
hiyouga
767aae4b72 fix #4609
unwrap_model_for_generation(reward_model) is necessary for zero3 training


Former-commit-id: c8d5b21700577cae8d6ca03359bcf1762c8b7cb8
2024-07-03 19:45:51 +08:00
hiyouga
e8a1dc2785 tiny fix
Former-commit-id: d944020257f363f38e62de6279b337e399b7c65e
2024-07-03 02:31:50 +08:00
hiyouga
3f2b9e9326 tiny fix
Former-commit-id: 98c4a0af6b3e27ae393d2847f48a01d23d9c8780
2024-07-02 23:06:13 +08:00
hiyouga
1aaee45a94 remove rlhf support for chatglm2&3
Former-commit-id: bcbb5b71961b89719bffb0d202c431c82e6067cc
2024-07-02 23:03:17 +08:00
hiyouga
6b755749b9 upcast logits
Former-commit-id: df61660351c8af30591471807a20869a45bb055a
2024-07-02 22:32:05 +08:00
hiyouga
ca106d1f1b improve rlhf
Former-commit-id: e441780e3db256ca09a442ea9254e7ce16898a07
2024-07-02 22:23:08 +08:00
hiyouga
8c3b285da2 fix ppo callbacks
Former-commit-id: 54f1c67c2a802b1d8368a6d1837d4c9a729f2695
2024-07-02 17:34:56 +08:00
hiyouga
14b37e1e03 tiny fix
Former-commit-id: 5dd2e5c3323f56420b5845a5ed28bcd9d4da5e41
2024-07-01 05:43:17 +08:00
hiyouga
8baf04d772 add eval acc
Former-commit-id: 7ffde76fbfb6192e3aac31ccc098f31ce89181ae
2024-07-01 03:51:20 +08:00
hiyouga
f7a4f3d9c0 loose gemma2 attention
Former-commit-id: a0b645017a2de3d58b6cbc71bd91ec96fc7a818b
2024-06-29 01:42:14 +08:00
hiyouga
81094dc09a add Gemma2 models
Former-commit-id: 8fc5a248ecfd6861cb90dac6c14fe89cdeaf8921
2024-06-28 01:26:50 +08:00
hiyouga
884a4a33ee refactor pissa, improve llamaboard
Former-commit-id: 619556e46c19718f702c97df5d570a2a4c5fb13a
2024-06-28 01:04:24 +08:00
hzhaoy
3491525d96 fix #4579
Former-commit-id: 0fa298ff6a4febea36ea9f11c7594277a77e6e9b
2024-06-27 13:49:57 +08:00
hiyouga
e823b200cf fix #4458
Former-commit-id: aab14b15268dbe74ded22549dbd3677474868cbb
2024-06-26 19:52:35 +08:00
hiyouga
8a6405f39e fix #4379
Former-commit-id: 96bedb4b6445a04ff8b97fb2aadace50b2f882df
2024-06-25 02:31:44 +08:00
hiyouga
4d2c279083 tiny fix about badam
Former-commit-id: 03f49267c7406e36aee35639f86e6e0383897090
2024-06-25 01:54:53 +08:00
hoshi-hiyouga
0105689cf4 Merge pull request #4352 from Ledzy/main
[Enhancement] Support ZeRO-3 when using BAdam

Former-commit-id: 0dc75275efa7d7540b472783a52ea6aeaa503c0b
2024-06-25 01:49:13 +08:00
hiyouga
83ee461b9a update readme
Former-commit-id: a1477208471039d3578980f929f1ca8c2a07aa96
2024-06-24 18:22:12 +08:00
hiyouga
da3b0aab6d fix templates
Former-commit-id: 6f357d59b73309c5955683008632e7f320e7dcb1
2024-06-19 17:44:05 +08:00
Jonery
a22e932b4f Cleaner integration.
Former-commit-id: 26d4b05d424bd71f570195dd433258caf6465d92
2024-06-19 12:29:40 +08:00
Jonery
b567216702 Support distributed BAdam.
Former-commit-id: bdcb986e37975911c190a74d3e60bb77aa2033bd
2024-06-18 12:27:47 +08:00
Jonery
d1edcfb135 Merge remote-tracking branch 'upstream/main'
Former-commit-id: 37834a7e79473ccf50ad7f67745b97c274c326d9
2024-06-17 18:44:51 +08:00
Jonery
34e98b7fc6 adapt for badam with ds zero3
Former-commit-id: fff2a020ec8713022bd8145f4a7168168ea07ca4
2024-06-17 18:18:10 +08:00
hiyouga
d5a0cc93a2 fix tol
Former-commit-id: bdb54bcb477126687db789bd89f2df84e424a2a3
2024-06-16 01:38:44 +08:00
hiyouga
0b571f84b4 support pissa
Former-commit-id: ef8e45f2eaf466c54e9a671512a2974575677b08
2024-06-16 01:08:12 +08:00
hiyouga
640372cb66 tiny fix
Former-commit-id: f7f440986b0ae3b38ea9f2da80789629d4f79ea1
2024-06-16 01:06:41 +08:00
hiyouga
acfae2e677 add license
Former-commit-id: 69cfc98d7c81756a5ab6bf962240e393e449fef0
2024-06-15 17:54:33 +08:00
hiyouga
43cfbde0d1 fix #4295
Former-commit-id: 08f657868f9d605b837c5d8c2946a25cc05c8735
2024-06-15 04:34:55 +08:00
hiyouga
bbeb3b10aa add test cases
Former-commit-id: 731176ff34cdf0cbf6b41c40c69f4ceb54c2daf6
2024-06-15 04:05:54 +08:00
hiyouga
e8885443a9 fix #4221
Former-commit-id: 05a3be4853b941909e7d193c31e8d62c8c5f879b
2024-06-13 02:48:21 +08:00
hiyouga
045cef901e fix #4209
DeepSpeed ZeRO3 has inflight param error when calling model.eval()


Former-commit-id: 4be013f18ea6a35b5a11db98db5f0670ffb41619
2024-06-13 02:25:50 +08:00
hiyouga
344d1192ac clean code
Former-commit-id: f54cafd5c7f0383370d1a2f357834a61a97397ce
2024-06-13 01:58:16 +08:00
hiyouga
95f95bef60 fix #4198
Former-commit-id: 945d2c6cc73542adf9272ebd9aa332ea2c1c7361
2024-06-11 15:38:38 +08:00
hiyouga
4785c723f8 tiny fix
Former-commit-id: 0621bcad1dfbe8ce2464f741d4256c5df2a8d1b6
2024-06-07 05:19:21 +08:00
hiyouga
90f706c05b fix ppo trainer save zero3 model
accelerator.get_state_dict(ds_model) should be called at all ranks


Former-commit-id: 3a0f60f0aa072531e4ae5819ec00c8fa42aa0913
2024-06-07 05:14:19 +08:00
hiyouga
7b7dc2f423 fix ppo in trl 0.8.6
Former-commit-id: 5e0d66a0d80b4bd4a8506e2317209d8fb9d25ff6
2024-06-07 04:48:29 +08:00