issues
search
corl-team
/
xland-minigrid
JAX-accelerated Meta-Reinforcement Learning Environments Inspired by XLand and MiniGrid 🏎️
Apache License 2.0
194
stars
15
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
fixed render error for numpy>2.0
#38
Howuhh
closed
3 weeks ago
0
Update ruff version + recheck
#37
Howuhh
closed
2 months ago
0
Fix typo
#36
helpingstar
closed
2 months ago
3
fix typo with pip installation for baselines
#35
Howuhh
closed
2 months ago
0
add tqdm dependency
#34
Howuhh
closed
2 months ago
0
use new signature for jax.numpy.clip
#33
garymm
closed
2 months ago
3
Improving baselines: extended obs, better embedder, faster with bf16
#32
Howuhh
closed
3 months ago
0
updated see also links
#31
Howuhh
closed
3 months ago
0
added wandb report
#30
Howuhh
closed
3 months ago
0
fix action dtype bug in Memory env
#29
Howuhh
closed
3 months ago
0
Added configs from the paper
#28
Howuhh
closed
3 months ago
0
updated to jax new key naming
#27
Howuhh
closed
3 months ago
0
XLand-100B dataset readme mention
#26
Howuhh
closed
3 months ago
0
add direction and task rules and goals to the observation via the wrappers
#25
Howuhh
closed
3 months ago
0
add support for bf16 training
#24
Howuhh
closed
3 months ago
0
Obs wrappers to add direction and actual task rules
#23
Howuhh
closed
3 months ago
0
Do you support single thread for a single environment?
#22
alexxchen
closed
4 months ago
1
Fail to optimize single tasks in the given demo. Maybe intermediate reward is needed.
#21
alexxchen
closed
4 months ago
10
Time limits rework
#20
Howuhh
closed
6 months ago
0
small improvements
#19
Howuhh
closed
6 months ago
0
The rollout speed is slower than gymnasium
#17
alexxchen
closed
7 months ago
4
Change tile grid cell conversion in render function
#16
afspies
closed
7 months ago
1
observations of env when not RGB
#15
wcarvalho
closed
7 months ago
1
fix evaluation in standalone
#14
Howuhh
closed
7 months ago
0
Allow XLand benchmarks as a single task envs in single-task ppo
#13
Howuhh
closed
7 months ago
0
move baselines to separate repo and add MAML
#12
Howuhh
closed
3 months ago
7
add more objects and colours for extended benchmarks
#11
Howuhh
opened
8 months ago
0
add option to save rollout gif for manual_control
#10
Howuhh
closed
7 months ago
0
RGB image observation wrapper compatible with jit
#9
Howuhh
closed
7 months ago
2
implement jit-compatible rendering
#8
Howuhh
closed
7 months ago
1
add experimental support for new gymnasium FuncEnv
#7
Howuhh
opened
8 months ago
0
Full typings support
#6
Howuhh
closed
8 months ago
0
Fix issue #3.
#5
floringogianu
closed
9 months ago
2
Add check for type hints with pyright (pre-commit hook + github action)
#4
Howuhh
closed
8 months ago
0
`ValueError: mutable default…` on the library import for python > 3.10
#3
ntoxeg
closed
9 months ago
5
[Feature Request] Add Maze-like env
#2
carlosluis
opened
11 months ago
6
Roadmap to v1.0
#1
Howuhh
opened
11 months ago
0