Skip to content

Commit 2eca2d2

Browse files
remove tasks from workarena, reduce num steps to 30
1 parent 945c524 commit 2eca2d2

File tree

2 files changed

+1
-4
lines changed

2 files changed

+1
-4
lines changed

browsergym/experiments/src/browsergym/experiments/benchmark/configs.py

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -254,7 +254,7 @@
254254
level="dg",
255255
task_category_filter=None,
256256
meta_seed=42, # meta seed for evaluation curriculum
257-
max_steps=50,
257+
max_steps=30,
258258
curriculum_type="agent",
259259
seeds_l1=n_repeats,
260260
),

browsergym/experiments/src/browsergym/experiments/benchmark/metadata/workarena.csv

Lines changed: 0 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -714,9 +714,6 @@ workarena.servicenow.infeasible-navigate-and-sort-incident-list-l3,l3,contextual
714714
workarena.servicenow.infeasible-navigate-and-sort-change-request-list-l3,l3,contextual_understanding_infeasible_tasks,test
715715
workarena.servicenow.infeasible-navigate-and-sort-hardware-list-l3,l3,contextual_understanding_infeasible_tasks,test
716716
workarena.servicenow.infeasible-navigate-and-sort-service-catalog-item-list-l3,l3,contextual_understanding_infeasible_tasks,test
717-
workarena.servicenow.order-apple-watch,dg,service catalog,test
718-
workarena.servicenow.order-developer-laptop,dg,service catalog,test
719-
workarena.servicenow.order-ipad-pro,dg,service catalog,test
720717
workarena.servicenow.assign-role-to-user-admin,dg,role,test
721718
workarena.servicenow.assign-roles-to-user-explicit,dg,role,test
722719
workarena.servicenow.assign-roles-to-user-implicit,dg,role,test

0 commit comments

Comments
 (0)