issues
search
jarbus
/
evo
MIT License
1
stars
0
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
todo: change create_next_pop back once tests are done
#149
jarbus
opened
1 year ago
0
todo: profile update_pop!
#148
jarbus
closed
1 year ago
1
todo: reduce day length
#147
jarbus
closed
1 year ago
0
todo: make agents less adept at collecting a resource
#146
jarbus
closed
1 year ago
0
todo: rm pick/place counts on bc
#145
jarbus
closed
1 year ago
0
bug: archive bc visualization can disappear in a later gen
#144
jarbus
closed
1 year ago
0
bug: agents can place more than they pick
#143
jarbus
closed
1 year ago
0
todo: bc includes total fitness
#142
jarbus
opened
1 year ago
2
todo: make standalone program to generate image of trade env?
#141
jarbus
opened
1 year ago
0
todo: change metric collector to collect individual rewards and update bc accordingly
#140
jarbus
closed
1 year ago
0
todo: buffer output for trade env renderer
#139
jarbus
closed
1 year ago
1
todo: add massing integration test for 2 pops for a single gen
#138
jarbus
closed
1 year ago
0
todo: make eval!()
#137
jarbus
closed
1 year ago
0
todo: README
#136
jarbus
opened
1 year ago
0
todo: create metric to dictate strategy used for each agent
#135
jarbus
closed
1 year ago
0
idea: remove mutations as well as add
#134
jarbus
opened
1 year ago
0
todo: confirm correctness of reversed-agent rollout
#133
jarbus
closed
1 year ago
0
bug: exchange still emerges somehow when playing against previous elite?
#132
jarbus
closed
1 year ago
1
todo: add env_pick and other_pick metric
#131
jarbus
closed
1 year ago
1
perf: compression statistics
#130
jarbus
closed
1 year ago
0
bug: min(exchange_0, exchange_1) != mut_exchanges for pop
#129
jarbus
closed
1 year ago
3
todo: pytest trade env
#128
jarbus
opened
1 year ago
0
perf: timings for groups
#127
jarbus
closed
1 year ago
2
x:1x1 notes
#126
jarbus
closed
1 year ago
5
todo: Collect *some* metrics from population every eval_gen intervals.
#125
jarbus
closed
1 year ago
1
x:trade-only, 3x3 4a
#124
jarbus
closed
1 year ago
10
todo: stabilize times for x/trade-only/2a-trade
#123
jarbus
closed
1 year ago
2
todo: fix visualizer
#122
jarbus
closed
1 year ago
0
todo: debug pycall error, will need to make interactive setting for trade env
#121
jarbus
closed
1 year ago
1
todo: switch to normal logger, which will allow log messages inside EvoTrade
#120
jarbus
closed
1 year ago
0
todo: clean up x/gatrade.jl, remove legacy tests
#119
jarbus
closed
1 year ago
0
todo: change rollout_groups_size and num_agents to be same
#118
jarbus
closed
1 year ago
0
todo: every eval_gen, cache prefix on all processors and identify it by it's hashed string. all populations are then sent out with this hash
#117
jarbus
closed
1 year ago
1
todo: add idxes of relevant elites along side the seeds, cache results for those elites during pmap
#116
jarbus
closed
1 year ago
0
lesson: writing to disk can big a major bottleneck
#115
jarbus
closed
1 year ago
0
lesson: figure out the right amount of logging, and implement it before you need it. anticipate what details you might need
#114
jarbus
closed
1 year ago
0
todo: speed up cache_elites on cluster at all costs
#113
jarbus
closed
1 year ago
6
todo: test with smaller num of gpu nodes
#112
jarbus
closed
1 year ago
1
bug: cache_elites takes more time than pmapping, investigate why. it might just be doing its job to reduce the time it takes to parallel map.
#111
jarbus
closed
1 year ago
3
bug: fix std plots
#110
jarbus
closed
1 year ago
0
bug: key not found error
#109
jarbus
closed
1 year ago
0
lesson: actually solve the toy problem before starting anything more complex
#108
jarbus
closed
1 year ago
0
todo: separate colors for elites
#107
jarbus
closed
1 year ago
0
todo: explore why novelty crashed for maze-test/full-adaptive
#106
jarbus
closed
1 year ago
8
bug: fitness and rew_light dont match up
#105
jarbus
closed
1 year ago
0
bug: elites suck for trade vis
#104
jarbus
closed
1 year ago
5
todo: solve maze
#103
jarbus
closed
1 year ago
1
todo: add archive size to metrics
#102
jarbus
closed
1 year ago
0
bug: MR trends towards zero as agents can't find the food->light solution
#101
jarbus
closed
1 year ago
0
bug: eval agents aren't moving
#100
jarbus
closed
1 year ago
0
Next