Commit Graph

66 Commits

Author SHA1 Message Date
lfy79001
acc2d41bdb add mixtral cogagent 2024-03-17 22:27:59 +08:00
Timothyxxx
e156a20e3d Update new func 2024-03-17 22:25:13 +08:00
lfy79001
505e772463 claude3_agent_code 2024-03-16 11:57:49 +08:00
lfy79001
684b4a1b7b claude3_agnet_code 2024-03-16 11:27:09 +08:00
lfy79001
3b13046745 add claude3 agent code 2024-03-16 01:40:41 +08:00
lfy79001
017dde8966 add claude3 agent code 2024-03-16 01:37:42 +08:00
David Chang
57f2257254 ver Mar15th
fixed bugs about infeasible task evaluation
2024-03-15 22:49:35 +08:00
David Chang
6face585f3 Merge branch 'zdy' 2024-03-15 22:48:40 +08:00
David Chang
e166106b6a ver Mar15th
added an option to keep buttons without text information but with an
image for SoM setting
2024-03-15 22:46:14 +08:00
Timothyxxx
1ad4527e8b Change SoM input and output 2024-03-15 22:10:35 +08:00
Timothyxxx
4db207fc27 Merge remote-tracking branch 'origin/main'
# Conflicts:
#	mm_agents/agent.py
#	run.py
2024-03-15 21:10:32 +08:00
Timothyxxx
5cbf1b28ca Fix bugs 2024-03-15 21:06:50 +08:00
Jason Lee
815c7ab67c filter unfinished examples and add timer to ensure upper limit of each example 2024-03-15 16:52:17 +08:00
David Chang
f6b96165e2 Merge branch 'zdy' 2024-03-14 22:40:27 +08:00
Timothyxxx
44ff027801 Refactor experiments and agent implementation 2024-03-14 22:32:49 +08:00
Timothyxxx
71ca8fbe1c refactor on exp code 2024-03-14 19:25:25 +08:00
Timothyxxx
26d52a7231 Code clean 2024-03-14 11:52:38 +08:00
Timothyxxx
741e26c3f8 Update 2024-03-13 23:35:04 +08:00
Timothyxxx
c2aa009ed8 Update server script, baseline and running script 2024-03-13 15:04:19 +08:00
David Chang
0c9c2f214a ver Mar11thv2
minor adjustment
2024-03-11 22:45:16 +08:00
David Chang
e95e8e55ea ver Mar11th
updated filter_nodes
2024-03-11 12:33:47 +08:00
David Chang
f08fa4912c ver Mar10th
changed AT element filtering
2024-03-10 18:03:02 +08:00
Timothyxxx
030574e316 Improve on mmagents prompts; initialize online tasks from Mind2Web 2024-02-22 22:01:22 +08:00
Timothyxxx
068c6f5769 122324154 2024-02-02 14:36:53 +08:00
Timothyxxx
32bcdd0937 Modify the logic of SoM agent 2024-02-01 18:58:22 +08:00
Timothyxxx
c31c9f4e7d Merge remote-tracking branch 'origin/main'
# Conflicts:
#	mm_agents/gpt_4v_agent.py
2024-02-01 16:57:01 +08:00
Timothyxxx
59e2417a08 Add Mistral, Qwen, Gemini support; Fix minor bugs 2024-02-01 16:55:38 +08:00
David Chang
5d436a6b66 ver Feb1st
human evaluation and SoM experiments on Thunderbird
2024-02-01 11:38:46 +08:00
Timothyxxx
606fab4cfa Fix minor bug when get a11y tree and linearize for agent 2024-01-31 00:29:51 +08:00
David Chang
da306376da ver Jan30th
updated function to get AT on Windows
2024-01-30 20:06:58 +08:00
David Chang
9e91b8a5a8 ver Jan29thv2
check som implementation
2024-01-30 00:25:00 +08:00
David Chang
d8a497a417 ver Jan29th
updated the position of SoM marks
2024-01-29 21:49:53 +08:00
Timothyxxx
cc21c3a6b1 Fix some errors found in calc examples 2024-01-28 21:19:18 +08:00
David Chang
8525825fb2 Merge branch 'zdy' 2024-01-27 23:18:33 +08:00
David Chang
5a486b6b37 ver Jan27th
debugged at+screenshot implementation, no issues found
fixed a little bugs
2024-01-27 23:10:48 +08:00
Timothyxxx
909aa868f3 Improve on agent codes; add auto-running experiments code; Fix some examples 2024-01-27 19:47:47 +08:00
David Chang
eef5158663 ver Jan26thv3
fixed bug caused by an empty node.text
remove nodes whose name and text are all empty
2024-01-26 23:49:15 +08:00
David Chang
73de0e387a Merge branch 'zdy' 2024-01-26 23:31:41 +08:00
Timothyxxx
6952b45de4 Improve on agent and tasks configs 2024-01-26 23:30:04 +08:00
David Chang
773c5ed40c ver Jan26thv4
updated linearized_accessibility_tree to add a column of "text"
removed replacement chars like uFFFC in thunderbird
2024-01-26 23:29:09 +08:00
David Chang
8d358d63ed ver Jan26thv3
updated agent history handling
2024-01-26 22:07:38 +08:00
Timothyxxx
6f27c5bf50 Wrap up SeeAct implementation 2024-01-20 19:19:37 +08:00
Timothyxxx
f88331416c Refactor baselines code implementations 2024-01-20 18:55:21 +08:00
Timothyxxx
09f3e776ae Initialize all baselines: screenshot, a11y tree, both, SoM, SeeAct 2024-01-20 00:13:46 +08:00
Timothyxxx
46bd3386dd Support input screenshot and a11y tree altogether 2024-01-19 20:34:47 +08:00
Timothyxxx
20b1d950a0 FIx corner cases (val connection in chrome when using playwright, and action parsing for agent, and accessibility tree xml handling) 2024-01-16 22:00:01 +08:00
Timothyxxx
186bf2e97c Implement heuristic cutting on the accessibility tree to get the important nodes; Finish accessibility tree text agent 2024-01-16 16:43:32 +08:00
Timothyxxx
48a86d36cf Minor updates 2024-01-16 12:15:21 +08:00
Timothyxxx
8efa692951 Add raw accessibility-tree based prompting method (but the tokens are too large); Minor fix some small bugs 2024-01-16 11:58:23 +08:00
Timothyxxx
493b719821 Add gemini agent implementation; Add missed requirements; Minor fix some small bugs 2024-01-15 21:58:33 +08:00