165 Commits

Author	SHA1	Message	Date
hiyouga	d5b2c57a35	fix modelscope data hub	2023-12-12 18:33:06 +08:00
hoshi-hiyouga	6382efec52	Merge branch 'main' into feat/support_ms	2023-12-12 17:55:32 +08:00
xingjun.wang	09533e95ed	update args for MsDataset.load	2023-12-12 13:02:54 +08:00
hiyouga	8cace77808	update readme	2023-12-12 11:44:30 +08:00
hiyouga	28d5de7e78	fix #1784	2023-12-09 20:53:18 +08:00
yuze.zyz	9c2247d700	support ms dataset	2023-12-08 18:00:57 +08:00
hiyouga	7df4f3ab20	implement rm server #1543	2023-12-03 20:52:54 +08:00
hiyouga	bd42c229b0	patch modelscope	2023-12-01 22:53:15 +08:00
hoshi-hiyouga	00f5c9ee16	Merge branch 'main' into feat/support_ms	2023-12-01 20:23:46 +08:00
hiyouga	a5a248d569	fix err hint	2023-12-01 17:13:22 +08:00
yuze.zyz	d38a2e7341	support ms	2023-11-29 20:36:55 +08:00
hiyouga	475a3fa0f4	fix #1659	2023-11-28 20:52:28 +08:00
hiyouga	859a6ea942	support export size setting	2023-11-26 18:34:09 +08:00
hiyouga	5021062493	update ppo trainer	2023-11-20 21:39:15 +08:00
Yuchen Han	b24635d22b	Update finetuning_args.py	2023-11-17 00:15:51 -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	ce78303600	support full-parameter PPO	2023-11-16 02:08:04 +08:00
hiyouga	4907452d95	support multiple modules in freeze training #1514	2023-11-15 17:08:18 +08:00
hiyouga	d125ef5535	fix #1494	2023-11-14 18:07:20 +08:00
hiyouga	442aefb925	refactor evaluation, upgrade trl to 074	2023-11-13 22:20:35 +08:00
hiyouga	a0c31c68c4	add todo	2023-11-10 14:38:18 +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	01260d9754	fix ppo train and dpo eval	2023-11-07 22:48:51 +08:00
hiyouga	479d0af2dc	delete file	2023-11-07 16:20:12 +08:00
hiyouga	b2a60905f3	upgrade peft, fix #1088 #1411	2023-11-07 16:13:36 +08:00
hiyouga	cc8ffa10d8	update data readme (zh)	2023-11-02 23:42:49 +08:00
hiyouga	a837172413	support sharegpt format, add datasets	2023-11-02 23:10:04 +08:00
hiyouga	dff128c7e3	fix #1356	2023-11-02 16:51:52 +08:00
hiyouga	083787dbfe	fix #1325	2023-11-01 23:38:49 +08:00
hiyouga	3fe7df628d	support dataset cache	2023-10-26 21:48:45 +08:00
hiyouga	7b4acf7265	reimplement neftune	2023-10-22 16:15:08 +08:00
anvie	57fb40aa04	add NEFTune optimization	2023-10-21 13:24:10 +07:00
hiyouga	b665e9e133	fix #1232	2023-10-20 23:28:52 +08:00
hiyouga	7a11a42dfd	fix #1218	2023-10-19 16:17:41 +08:00
hiyouga	ea82f8a82a	refactor export, fix #1190	2023-10-15 16:01:48 +08:00
hiyouga	af18b0dce7	fix #1184	2023-10-14 19:20:11 +08:00
hiyouga	11bd271364	fix ppo args	2023-10-11 23:40:50 +08:00
hiyouga	2818af0b09	refactor model_dtype, fix PPO trainer	2023-10-11 23:16:01 +08:00
hiyouga	84b7486885	fix layer norm dtype	2023-09-28 00:25:55 +08:00
hiyouga	620efe1d8d	refactor finetuning Args	2023-09-27 22:28:06 +08:00
hiyouga	90375f600d	support LongLoRA	2023-09-27 21:55:50 +08:00
hiyouga	338b8664ed	fix #944	2023-09-21 19:51:02 +08:00
hiyouga	d8aa1404be	support FlashAttention2	2023-09-10 20:43:56 +08:00
hiyouga	8ea32e4046	change to right-padding, update reward score #803	2023-09-08 20:04:31 +08:00
hiyouga	a9d1fb72f7	refactor dataset_attr, add eos in pt, fix #757	2023-09-01 19:00:45 +08:00
codemayq	ba94c8729d	add stage in DatasetAttr	2023-08-23 20:54:53 +08:00
hiyouga	4318347d3f	update template	2023-08-22 19:46:09 +08:00
hiyouga	53e33418d0	support ppo score norm (trl 0.5.1.dev required)	2023-08-18 12:02:42 +08:00