19 commits

Author	SHA1	Message	Date
Yizhou Chi	9ba0d217fc	update mcts logic	2024-09-09 14:52:25 +08:00
Yizhou Chi	c0262bcd8f	1. add support to hf dataset ... 2. add support to datasets that have both train and test 3. create data folder 4. fix new instruction bug	2024-09-06 19:05:10 +08:00
Yizhou Chi	a6f71f4498	1. avoid circular reference ... 2. add greedy	2024-09-06 13:09:18 +08:00
Yizhou Chi	e07ed0df8b	fix mcts bug	2024-09-06 10:42:24 +08:00
Yizhou Chi	c16286a006	Refactor MCTS class to handle role running errors and improve error logging	2024-09-05 10:12:37 +08:00
Yizhou Chi	ab8a1d6824	format code	2024-09-04 17:52:02 +08:00
Yizhou Chi	fcd1ba66a6	增加try catch	2024-09-04 16:38:33 +08:00
Yizhou Chi	72bd1665b1	ensure experimenter not evaluating csv from other experiments	2024-09-04 10:41:10 +08:00
Yizhou Chi	6972afb755	update prompt (specify whether each set has target label)	2024-09-03 16:31:03 +08:00
Yizhou Chi	3ec6dcd5df	update readme	2024-09-03 13:40:23 +08:00
Yizhou Chi	a1668a1d9d	add custom experimenter	2024-09-02 20:00:46 +08:00
Yizhou Chi	f28908e4c5	1. role.py - 如果已经有plan，便不再重复生成 ... 2. 修改prompt，让predictions.csv生成的格式与原gt格式一样	2024-09-02 15:15:53 +08:00
Yizhou Chi	0e5db1c364	1. add ml module ... 2. fix bug	2024-09-02 14:07:14 +08:00
Yizhou Chi	27bbc927b0	1. Rewrite logger message ... 2. fix import	2024-09-02 10:21:30 +08:00
Yizhou Chi	16d1bf0da0	Rename insight generate to instruction generator	2024-09-02 09:59:47 +08:00
Yizhou Chi	ae6a195750	fix bug - make rollout more consistent	2024-08-30 20:35:17 +08:00
Yizhou Chi	32759f031c	add experimenter	2024-08-30 19:55:40 +08:00
Yizhou Chi	d14f07f9b1	1. change data.yaml to more generalized path ... 2. correct import	2024-08-30 16:59:38 +08:00
Yizhou Chi	95d7160c0e	add expo	2024-08-30 16:26:05 +08:00