428 Commits

Author	SHA1	Message	Date
hiyouga	9ea9380145	support GPTQ tuning #729 #1481 #1545 , fix chatglm template #1453 #1480 #1569	2023-11-20 22:52:11 +08:00
hiyouga	5021062493	update ppo trainer	2023-11-20 21:39:15 +08:00
hoshi-hiyouga	48211e3799	Merge pull request #1553 from hannlp/hans ... Change the default argument settings for PPO training	2023-11-20 20:32:55 +08:00
hiyouga	2a36fd5064	fix value head model resuming	2023-11-20 19:01:37 +08:00
hiyouga	99a3f06377	fix #1567	2023-11-20 18:46:36 +08:00
hiyouga	00baaa990e	better data streaming	2023-11-19 23:32:47 +08:00
hiyouga	211b2db5a8	fix model card network issue	2023-11-19 23:03:19 +08:00
hiyouga	bfb9433165	fix Mistral template ... https://github.com/lm-sys/FastChat/pull/2547	2023-11-19 16:29:30 +08:00
hiyouga	065bfaeed4	fix #1263	2023-11-19 16:05:18 +08:00
hiyouga	1740131d63	fix #1558	2023-11-19 14:15:47 +08:00
hiyouga	ff6056405d	fix evaluator and cached_file in 4.31.0	2023-11-18 19:39:23 +08:00
hiyouga	ccb0f58e22	fix quantization	2023-11-17 22:21:29 +08:00
hiyouga	1bbc1be95e	fix #1550	2023-11-17 17:23:13 +08:00
Yuchen Han	eeb5249d0b	Update workflow.py	2023-11-17 00:16:27 -08:00
Yuchen Han	b24635d22b	Update finetuning_args.py	2023-11-17 00:15:51 -08:00
hiyouga	999bc0ed93	fix packages	2023-11-17 16:11:48 +08:00
Shaowen Wang	397e948984	Fix: Change rouge-chinese package name to rouge_chinese ... To reproduce: python: importlib.util.find_spec('rouge-chinese') -> None importlib.util.find_spec('rouge_chinese') -> ModuleSpec(name='rouge_chinese'...) from rouge_chinese import Rouge print(Rouge.__module__) -> rouge_chinese	2023-11-16 20:12:35 -06:00
hiyouga	ed9f7705ef	fix chatglm template	2023-11-16 22:54:15 +08:00
hiyouga	10ce87e088	fix web ui demo	2023-11-16 18:41:55 +08:00
hiyouga	1c80e9a09e	fix web ui demo	2023-11-16 17:12:23 +08:00
hiyouga	c4facc03af	release v0.3.0	2023-11-16 16:00:11 +08:00
hiyouga	08f3c11429	fix css	2023-11-16 15:45:38 +08:00
hiyouga	6efa38be46	fix bug in web ui	2023-11-16 15:21:24 +08:00
hiyouga	7537dd434f	update ppo and demo in webui	2023-11-16 14:55:26 +08:00
hiyouga	ff52b1779c	fix bug in freeze tuning	2023-11-16 14:25:11 +08:00
hiyouga	83cee2a604	tiny fix	2023-11-16 03:27:19 +08:00
hiyouga	1817ffc86f	fix rlhf callback	2023-11-16 03:26:19 +08:00
hiyouga	856522a3df	fix bug in PPO training	2023-11-16 02:32:54 +08:00
hiyouga	35b91ea34c	fix import bug	2023-11-16 02:27:03 +08:00
hiyouga	ce78303600	support full-parameter PPO	2023-11-16 02:08:04 +08:00
hiyouga	8350bcf85d	add demo mode for web UI	2023-11-15 23:51:26 +08:00
hiyouga	1e19cf242a	update readme and constants	2023-11-15 18:04:37 +08:00
hiyouga	4907452d95	support multiple modules in freeze training #1514	2023-11-15 17:08:18 +08:00
hiyouga	bbbce1f516	fix imports	2023-11-15 16:47:45 +08:00
hiyouga	4736344eb1	disentangle model from tuner and rename modules	2023-11-15 16:29:09 +08:00
hiyouga	2f02f688e1	fix #1507	2023-11-15 16:22:32 +08:00
hiyouga	42c8fc4fb9	add cal_lr.py	2023-11-14 20:58:37 +08:00
hiyouga	d125ef5535	fix #1494	2023-11-14 18:07:20 +08:00
hiyouga	3743b7420b	fix #1489	2023-11-14 15:27:05 +08:00
hiyouga	2d42be32c1	support eval remote dataset	2023-11-14 02:42:30 +08:00
hiyouga	35cc1e28f6	release v0.2.2, fix #1478 #1466	2023-11-13 23:09:05 +08:00
hiyouga	87390ae3b7	fix #424	2023-11-13 22:42:23 +08:00
hiyouga	442aefb925	refactor evaluation, upgrade trl to 074	2023-11-13 22:20:35 +08:00
hiyouga	4bd8e3906d	fix flashattn warning	2023-11-10 18:34:54 +08:00
hiyouga	a0c31c68c4	add todo	2023-11-10 14:38:18 +08:00
hiyouga	3697a3dc9a	refactor constants	2023-11-10 14:16:10 +08:00
hiyouga	415bca900e	tiny fix	2023-11-09 17:20:49 +08:00
Yanqing	3684dffa14	Update finetuning_args.py ... 更新 chatglm/falcon/bloom 的 lora_target 的名称	2023-11-09 17:04:40 +08:00
hiyouga	0e86527d7f	fix #1452	2023-11-09 16:41:32 +08:00
hiyouga	1db59832fd	release v0.2.1	2023-11-09 15:54:16 +08:00