Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

mask fixes #78

Merged
merged 19 commits into from
Sep 6, 2023
Merged

mask fixes #78

Changes from 1 commit
Commits
Show all changes
19 commits
Select commit Hold shift + click to select a range
c58809e
add some noise to masks to prevent entropy collapse
kywch Aug 30, 2023
64c53f0
tweaked give gold target mask to prevent entropy collapse
kywch Aug 30, 2023
d3febd2
tweaked action masks during combat
kywch Aug 30, 2023
ec22970
mask tweak
kywch Aug 30, 2023
3614560
disallow no-op for attack and buy to jitter
kywch Aug 31, 2023
9851ed3
fixed tests
kywch Aug 31, 2023
eecee94
quick mask tweak
kywch Aug 31, 2023
240dc0e
move mask tweak
kywch Aug 31, 2023
377ff22
removed mask jitter other than attack and move
kywch Aug 31, 2023
41ac731
mask no-op in move, attack when there are valid options
kywch Aug 31, 2023
7aabad2
removed attack mask jitter
kywch Sep 1, 2023
04026a1
put back move no-op
kywch Sep 2, 2023
0769c4b
added exp to entity ds, tweaked exp threshold
kywch Sep 3, 2023
dce5c0a
added resilient agents to help training
kywch Sep 4, 2023
e771e89
added gain exp predicate
kywch Sep 4, 2023
f35c9eb
fixed error with custom pred arg
kywch Sep 4, 2023
1b864fe
fix water exp error
kywch Sep 4, 2023
6a92799
correct skill type hint
kywch Sep 4, 2023
3a1ebb0
Merge pull request #119 from CarperAI/exp-ds
jsuarez5341 Sep 6, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
put back move no-op
  • Loading branch information
@kywch
kywch committed Sep 2, 2023
commit 04026a1a3fa71e8e560cf38e35a3b22671d2fcc9
10 changes: 1 addition & 9 deletions nmmo/core/observation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -217,17 +217,9 @@ def _make_move_mask(self):
return mask

# pylint: disable=not-an-iterable
mask = np.array([self.tile(*d.delta).material_id in material.Habitable.indices
return np.array([self.tile(*d.delta).material_id in material.Habitable.indices
for d in action.Direction.edges], dtype=np.int8)

if sum(mask) <= 1:
# if only the stay (no-op) is possible, then allow all actions
mask[:] = 1
# Mask the no-op option, since there should be at least one allowed move
mask[-1] = 0

return mask

def _make_attack_mask(self):
# NOTE: Currently, all attacks have the same range
# if we choose to make ranges different, the masks
Expand Down