Commits · d09ddc698be6d5b6c8bb2682d707a33a6760f02f · Wang, Mia / MetaRL

Apr 13, 2022
- Add documentation framework using Sphinx · d09ddc69
  Sun Jin Kim authored 2 years ago
  
  d09ddc69
- Update docstrings in /autoaugment_learners/ · 2d0583db
  Sun Jin Kim authored 2 years ago
  
  2d0583db
Apr 11, 2022
- John: Add EasyNet · 63a7084c
  Sun Jin Kim authored 2 years ago
  
  63a7084c
- baseline code · 3a200057
  Sun Jin Kim authored 2 years ago
  
  3a200057
Apr 10, 2022
- Try increase toy dataset size and decrease child network complexity · 3625043d
  Sun Jin Kim authored 2 years ago
  
  3625043d
Apr 08, 2022
- mnist 0.01 randomsearch vs gru graph data · 929f2dd0
  Sun Jin Kim authored 2 years ago
  
  929f2dd0
- Add check_pickles.py · 0279c6ea
  Sun Jin Kim authored 2 years ago
  
  0279c6ea
- gru_learner also saves itself as pkl · 31fa2958
  Sun Jin Kim authored 2 years ago
  
  31fa2958
- GRU layer 1->2 · 7658134b
  Sun Jin Kim authored 2 years ago
  
  7658134b
- work on genetic learner with some results · 1a375295
  Max Ramsay King authored 2 years ago
  
  1a375295
- finishied GRU learner · 0a0c399c
  Sun Jin Kim authored 2 years ago
  
  0a0c399c
- Round up gru_learner's prob and mag values of operations · 2420ae80
  Sun Jin Kim authored 2 years ago
  
  2420ae80
- Implement gru_learner's generate_new_policy · 1d782627
  Sun Jin Kim authored 2 years ago
  
  1d782627
Apr 07, 2022
- cleaned up imports in /autoaugment_learners · 6e49635f
  Sun Jin Kim authored 2 years ago
  
  6e49635f
- Add new files where GRU_Learner will grow · 555ef1b4
  Sun Jin Kim authored 2 years ago
  
  555ef1b4
- translate_operation_tensor can now translate probabilistically. (Using tensor as multinomial pdf's) · 92d97847
  Sun Jin Kim authored 2 years ago
  
  92d97847
- add Mia's ac learner in-progress code · 7935e672
  Sun Jin Kim authored 2 years ago
  
  7935e672
Apr 06, 2022
- covariance method for the evolutionary strategy · e06e3cc0
  Max Ramsay King authored 2 years ago
  
  e06e3cc0
Apr 05, 2022
- Using test_autoaugment_policy to calcualte the accuracy of subpolicies. ES... · 2a450edb
  Max Ramsay King authored 2 years ago
  
  Using test_autoaugment_policy to calcualte the accuracy of subpolicies. ES learner now outputs similar list to randomsearcher
  2a450edb
Apr 04, 2022
- fixed bug · f5a10e7e
  Max Ramsay King authored 2 years ago
  
  f5a10e7e
- Made the get_full_policy function a method of the ES class rather than a feature of the network · e08b708c
  Max Ramsay King authored 2 years ago
  
  e08b708c
- Updated the ES learner to be more in line with the random search funcitons · 5f05e0d7
  Max Ramsay King authored 2 years ago
  
  5f05e0d7
Apr 03, 2022
- Turned aa_learner.py into an actually usable parent class · 45689337
  Sun Jin Kim authored 2 years ago
  
  45689337
- edit aa_learner · 8bdeaf68
  Sun Jin Kim authored 2 years ago
  
  8bdeaf68
- Alternative genetic algorithm · e75c71a4
  Max Ramsay King authored 2 years ago
  
  e75c71a4
Apr 01, 2022
- CP2_MAX: replace copy pasted code · 5fda1aca
  Sun Jin Kim authored 2 years ago
  
  5fda1aca
Mar 31, 2022
- Changed CP2_max to include an evolutionary_learner class · bd2ff5dd
  Max Ramsay King authored 2 years ago
  
  bd2ff5dd
- Add cuda to train_child_network · 9c101635
  Sun Jin Kim authored 2 years ago
  
  9c101635
Mar 29, 2022
- Copy randomsearch_learner code into aa_learner · debf9e2d
  Sun Jin Kim authored 2 years ago
  
  debf9e2d
- small type · 676994ba
  Sun Jin Kim authored 2 years ago
  
  676994ba
- Add 'with torch.no_grad()' to train_child_network() · 647d873d
  Sun Jin Kim authored 2 years ago
  
  647d873d
- random search learner first version complete · 37f9713f
  Sun Jin Kim authored 2 years ago
  
  37f9713f
- Almost done with Random Search Learner · 61aed9cb
  Sun Jin Kim authored 2 years ago
  
  61aed9cb
Mar 28, 2022
- work on randomsearch_learner · c16fe2bc
  Sun Jin Kim authored 3 years ago
  
  c16fe2bc
- Add 'seed' argument for 'create_toy' · cef7662c
  Sun Jin Kim authored 3 years ago
  
  cef7662c
Mar 27, 2022
- Using: 'train_dataset.transform = my_transform' · 4a934128
  Sun Jin Kim authored 3 years ago
  
  4a934128
- create_toy in main.py is now deterministic · 879f574a
  Sun Jin Kim authored 3 years ago
  
  879f574a
- train_child_network in main.py now logs accuracy · 86b2bf58
  Sun Jin Kim authored 3 years ago
  
  86b2bf58
- change nomenclature in autoaugment.py: 'policies' -> 'subpolicies' · 124d6915
  Sun Jin Kim authored 3 years ago
  
  124d6915
- Add some comments that will help developers use autoaugment.py to implement... · 6c3a28a4
  Sun Jin Kim authored 3 years ago
  
  Add some comments that will help developers use autoaugment.py to implement the 5-2 policy system in their own agents.
  6c3a28a4