issues
search
newmanne
/
open_spiel
Apache License 2.0
0
stars
0
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
add gym to requirements
#55
gregdeon
opened
2 years ago
3
pip "Double requirement given"
#54
gregdeon
opened
2 years ago
0
Add total activity (or activity fraction) as features?
#53
newmanne
closed
2 years ago
0
Add a torso to recurrent models
#52
newmanne
closed
2 years ago
0
Sweep over MLP learning rates
#51
newmanne
closed
2 years ago
0
Support marginally increasing valuations
#50
newmanne
closed
2 years ago
0
Database too many connections?
#49
newmanne
closed
2 years ago
2
We still sometimes have 0 nash convs because our networks can't find any better exploitations
#48
newmanne
closed
2 years ago
0
Enhance pricing bonus to be player-targeted
#47
newmanne
opened
2 years ago
0
Compute efficiency metric
#46
newmanne
closed
2 years ago
0
Look at Boltzman DQN
#45
newmanne
opened
2 years ago
0
Push boundaries
#44
newmanne
opened
2 years ago
0
Add last layer clustering to UI
#43
newmanne
closed
2 years ago
0
Complete code for NashConv bokeh graphs on server UI
#42
newmanne
closed
2 years ago
0
Add allocation heatmaps
#41
newmanne
closed
2 years ago
0
Tell if experiments failed -> needs a better solution
#40
newmanne
opened
2 years ago
2
Add per-product clock prices to the UI
#39
newmanne
closed
2 years ago
1
Should iterations be set per-player?
#38
newmanne
closed
2 years ago
0
Fix python path on clusters
#37
newmanne
closed
2 years ago
0
DQN reporting iteration number - checfk
#36
newmanne
closed
2 years ago
2
Build omniscien mode
#35
newmanne
opened
2 years ago
0
Add %
#34
newmanne
closed
2 years ago
1
Autosort by Num Plays
#33
newmanne
closed
2 years ago
0
Q values in opening explorer
#32
newmanne
closed
2 years ago
0
Sort order for Runs is silly
#31
newmanne
closed
2 years ago
1
Feature normalization
#30
newmanne
closed
2 years ago
0
If we give up on LSTMs, make information state tensors less complicated?
#29
newmanne
closed
2 years ago
1
What can we say about non-credible threats?
#28
newmanne
opened
2 years ago
0
Deal with the learning rate issues through some normalization
#27
newmanne
closed
2 years ago
1
Capture additional information in the reward pickles
#26
newmanne
closed
2 years ago
0
Allow comparison of recurrent ideas to MLPs with past N observations
#25
newmanne
closed
2 years ago
0
Repackage clusters with stronger machines
#24
newmanne
closed
2 years ago
0
RNN caching
#23
newmanne
opened
2 years ago
0
Within-epsiode caching for RNN
#22
newmanne
closed
2 years ago
2
Weight iters in supervised network of NFSP
#21
newmanne
closed
2 years ago
1
Best response should log loss
#20
newmanne
closed
2 years ago
1
Transformers crash and burn on best response
#19
newmanne
opened
2 years ago
1
Option in graphviz to collapse chance infosets into expectations
#18
newmanne
opened
2 years ago
0
Consider modelling the grace period
#17
newmanne
opened
3 years ago
0
Look at activity rule bug
#16
newmanne
closed
3 years ago
0
Implement actual tie-breaking
#15
newmanne
closed
2 years ago
0
Investigate issues w/ MCCFR
#14
newmanne
closed
3 years ago
2
Code up a restricted nash response
#13
newmanne
opened
3 years ago
3
Add back undersell to multi_item games
#12
newmanne
closed
3 years ago
0
Understand why this game is too big to solve
#11
newmanne
closed
3 years ago
1
TODO: Understand storage requirements for actions
#10
newmanne
closed
3 years ago
1
Sharpen solution should not put 0's on probabilities for invalid actions
#9
newmanne
closed
2 years ago
1
Add warm starting (from pickle) to main script
#8
newmanne
closed
2 years ago
0
Error in prob column?
#7
newmanne
closed
3 years ago
1
Skip players that only have 1 action choice
#6
newmanne
closed
3 years ago
1
Next