issues
search
raharth
/
PyMatch
A collection of different PyTorch wrappers for training neural networks and reinforcement algorithms
MIT License
13
stars
2
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Create a module that updates the agent, so that the learnign process can be easily changed
#116
raharth
opened
2 years ago
0
Create MulitInstanceEnvironment
#115
raharth
closed
3 years ago
0
All `to(device)` should return self
#114
raharth
opened
3 years ago
0
Make the network update a callable?
#113
raharth
opened
3 years ago
0
One could use the overall uncertainty to determin when it is useful to sample new trajectories
#112
raharth
closed
3 years ago
1
catch exception in train script or the experiment and write it to some text file, for later evaluation
#111
raharth
opened
3 years ago
0
Find working breakout configuration
#110
raharth
opened
3 years ago
2
Remove unsused temperature parameter from params.json file
#109
raharth
closed
3 years ago
0
Create Ensemble Predictor
#108
raharth
closed
3 years ago
0
How to incorporate entropy as uncertainty measure?
#107
raharth
closed
3 years ago
0
Introduce Entropy based uncertainty hat
#106
raharth
closed
3 years ago
0
Replace Lists in Memory with tensors
#105
raharth
closed
3 years ago
0
Setup for second GPU machine
#104
raharth
closed
3 years ago
0
Compare results of exp 168,169,171
#103
raharth
closed
3 years ago
0
Favoring uncertain states over certain ones when sampling actions
#102
raharth
opened
3 years ago
0
Memory dropping based on uncertainty
#101
raharth
opened
3 years ago
0
Non-uniform memory sampling
#100
raharth
closed
3 years ago
3
Why is the prob evaluation of exp 148 as bad
#99
raharth
closed
3 years ago
0
Priority Memory
#98
raharth
closed
3 years ago
0
Run ensemble on Lunar Lander DQN
#97
raharth
closed
3 years ago
0
Use episode sampler instead of fixed size memory for DQN
#96
raharth
closed
3 years ago
0
Read paper on uncertainty with lunar lander
#95
raharth
closed
3 years ago
1
Chose good hyperparameters for PG on Lunar Lander
#94
raharth
closed
3 years ago
0
Evaluate impact of tau for DDQN on the stability and volatility
#93
raharth
closed
3 years ago
0
Find good tau for DDQN using MCD
#92
raharth
closed
3 years ago
0
MCD with eternal memory PG
#91
raharth
closed
3 years ago
0
MCD with eternal memory DDQN
#90
raharth
closed
3 years ago
0
Ensemble with eternal memory PG
#89
raharth
closed
3 years ago
0
Ensemble with eternal memory DDQN
#88
raharth
closed
3 years ago
0
Ensemble with eternal memory DQN
#87
raharth
closed
3 years ago
0
Boosting with eternal memory PG
#86
raharth
closed
3 years ago
0
Boosting with eternal memory DDQN
#85
raharth
closed
3 years ago
0
Boosting with eternal memory DQN
#84
raharth
closed
3 years ago
0
Bericht DoE
#83
raharth
closed
3 years ago
0
Rerun exp_75/72/73 two times as independent repetitions
#82
raharth
closed
3 years ago
0
Compare forgetfulness and volatility from agent average to ensemble/boosting
#81
raharth
closed
3 years ago
0
change font in plots
#80
raharth
closed
3 years ago
0
Give numeric evidence that an ensemble improves the performance while boosting does not
#79
raharth
closed
3 years ago
0
Estimate density over action-state space
#78
raharth
closed
3 years ago
1
Adding memory saving and loading to agents memory class
#77
raharth
closed
3 years ago
0
Compare individual and shared memories for boosting and model averaging
#76
raharth
closed
3 years ago
0
Add walltime evaluations for training and inference to report
#75
raharth
closed
3 years ago
1
Create automatic add, commit, and push to github, when bumping version
#74
raharth
closed
3 years ago
0
Experiments do not terminate after finishing training
#73
raharth
closed
3 years ago
0
Stability
#72
raharth
closed
3 years ago
0
exp_75
#71
raharth
closed
3 years ago
0
Aggregate results of different experiments to compare MC-Dropout, Boosting, Ensemble
#70
raharth
closed
3 years ago
0
Hardware Monitoring during experiment
#69
raharth
closed
3 years ago
0
Run exp_58 10 times
#68
raharth
closed
3 years ago
0
Run exp_55 10 times
#67
raharth
closed
3 years ago
0
Next