lucasosouza fasterRL issues

lucasosouza / fasterRL

Reinforcement Learning library based on pytorch. Designed for research and experiments. Platform-agnostic (supports openai, marlo, more to be added)

MIT License

6 stars 1 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

use python native logger to implement logger class

#29 lucasosouza opened 5 years ago
0
check and implement ideas from udacity DRL course

#28 lucasosouza opened 5 years ago
0
create evaluator class

#27 lucasosouza opened 5 years ago
0
implement replay pre-fill for experience replay algorithms

#26 lucasosouza opened 5 years ago
0
implement online and offline evaluation

#25 lucasosouza opened 5 years ago
0
importance sampling

#24 lucasosouza closed 5 years ago
3
prioritized replay

#23 lucasosouza closed 5 years ago
1
implement tests

#22 lucasosouza opened 5 years ago
0
improve logger class architecture

#21 lucasosouza opened 5 years ago
0
skill sharing

#20 lucasosouza opened 5 years ago
0
implement methods in latest papers by DeepMind and OpenAI

#19 lucasosouza opened 5 years ago
0
hierarchical learning

#18 lucasosouza opened 5 years ago
0
discretization

#17 lucasosouza closed 5 years ago
1
test and assure convergence of continuous policy gradient algorithms (including A2C)

#16 lucasosouza opened 5 years ago
0
make a list of sources used for reference for the library creation

#15 lucasosouza opened 5 years ago
0
try to make overall implementation deep learning platform-agnostic

#14 lucasosouza opened 5 years ago
0
create benchmark for main algorithms and environments

#13 lucasosouza opened 5 years ago
1
support for deepmind platform

#12 lucasosouza opened 5 years ago
0
support for marlo platform

#11 lucasosouza closed 5 years ago
1
experiments wall-clock time tracking

#10 lucasosouza closed 5 years ago
1
distributional methods

#9 lucasosouza opened 5 years ago
0
TRPO and PPO

#8 lucasosouza opened 5 years ago
0
eligibility traces

#7 lucasosouza opened 5 years ago
0
multiagent experiments - parallel

#6 lucasosouza opened 5 years ago
0
multiagent experiments - sequential

#5 lucasosouza closed 5 years ago
1
experience sharing

#4 lucasosouza closed 5 years ago
1
curriculum learning

#3 lucasosouza opened 5 years ago
1
implement n-steps in buffer class, to be reused by different agents

#2 lucasosouza opened 5 years ago
1
create expected sarsa agent class

#1 lucasosouza opened 5 years ago
0