Commit graph

Select branches

Hide pull requests

main

ruby/studentstuff

ruby/subsets

#1

#2

Mono Color

• 1c16496e61
  
  O1: allow audit to pull from training set main Akemi Izuko 2024-12-08 23:35:53 -0700
• 99ba0b3f6d
  
  O1: multithread k-search Akemi Izuko 2024-12-07 17:46:07 -0700
• 70d4e4dfdc
  
  O1: slight cleanup Akemi Izuko 2024-12-07 17:37:19 -0700
• f407827ac1
  
  O1: add wrn2 architecture Akemi Izuko 2024-12-07 14:00:39 -0700
• 5da8c44743
  
  O1: add simple convnet Akemi Izuko 2024-12-07 14:00:06 -0700
• 7b77748dcd
  
  O1: wrn2 fixes Akemi Izuko 2024-12-07 13:59:39 -0700
• 2586c351d9
  
  O1: insert attack points Akemi Izuko 2024-12-06 19:11:54 -0700
• e239602148
  
  O1: update theoretical plots Akemi Izuko 2024-12-06 18:57:02 -0700
• ce3a848eb7
  
  O1: add fast training code Akemi Izuko 2024-12-06 18:56:47 -0700
• 86d16e53d7
  
  O1: fix epochs Akemi Izuko 2024-12-06 17:12:38 -0700
• ebfbd88332
  
  O1: with student model Akemi Izuko 2024-12-05 01:04:35 -0700
• a697d4687c
  
  O1: new data splitting Akemi Izuko 2024-12-05 00:13:50 -0700
• 3c07d6fe6b O1: update student to use pure train dataloader ruby/studentstuff akemi 2024-12-04 23:36:36 -0700
• 1200907c31 student stuff ARVP 2024-12-04 23:32:30 -0700
• c6f91352d9 added if statement to either keep the datapoint in our out ruby/subsets ARVP 2024-12-04 07:45:06 -0700
• 3fda6dd727 made to use student/teacher subsets that are half data each ARVP 2024-12-04 07:25:56 -0700
• 369249ce69
  
  O1: add epsilon audit Akemi Izuko 2024-12-03 23:26:50 -0700
• 5d6f7e2916
  
  O1: get audit vectors Akemi Izuko 2024-12-03 23:02:55 -0700
• 4692502763
  
  O1: return target point labels Akemi Izuko 2024-12-03 22:35:41 -0700
• d606245ad1
  
  O1: save starting model Akemi Izuko 2024-12-03 19:43:05 -0700
• 36deb4613b
  
  O1: fix inout dataloader Akemi Izuko 2024-12-03 16:53:33 -0700
• e9af7cacf1
  
  O1: fix dataloader batch size Akemi Izuko 2024-12-03 13:01:38 -0700
• 0d67830f7e
  
  O1: add training code Akemi Izuko 2024-12-02 23:48:50 -0700
• 2eef211415
  
  Wres: add conda env file Akemi Izuko 2024-12-02 19:01:26 -0700
• ad666283c5
  
  Wres: clean up student a bit Akemi Izuko 2024-12-02 18:45:04 -0700
• 524576db31 fixed to use student model ARVP 2024-12-02 17:54:29 -0700
• 2790f3b0fe fixed to use student model ARVP 2024-12-02 17:54:29 -0700
• e3ccfbcf76
  
  Wres: args for students Akemi Izuko 2024-12-02 17:31:11 -0700
• 5be312bf18 changed to same data loaders as train.py and added saving student model Ruby 2024-12-01 15:33:03 -0700
• 7208c16efc
  
  Wres: epsilon in args Akemi Izuko 2024-12-01 14:49:13 -0700
• aa190cd4f1
  
  Wres: toggle non-dp training Akemi Izuko 2024-12-01 13:58:50 -0700
• 0eb26f8979
  
  Wres: add distillation code Akemi Izuko 2024-12-01 13:38:32 -0700
• 424cb01a15
  
  Wres: cuda and norm flags Akemi Izuko 2024-11-30 23:54:48 -0700
• e4b5998dbb
  
  Wres: dp training Akemi Izuko 2024-11-30 20:28:25 -0700
• c1561c7d55
  
  Wres: wresnet with audit settings Akemi Izuko 2024-11-30 15:41:11 -0700
• c163e3062c
  
  Torchlira: cite repositories Akemi Izuko 2024-11-30 13:36:40 -0700
• 944ff9d5cd
  
  Torchlira: add wresnet-16 Akemi Izuko 2024-11-30 13:35:38 -0700
• 2b865a5f58
  
  Torchlira: add additional networks Akemi Izuko 2024-11-30 13:34:05 -0700
• 1b099f4dad
  
  Torchlira: attempt microbatch Akemi Izuko 2024-11-29 19:43:38 -0700
• c7eee3cdc2
  
  Torchlira: remove secure mode Akemi Izuko 2024-11-29 18:30:26 -0700
• fe7fcb2761
  
  Torchlira: update gitignore Akemi Izuko 2024-11-29 17:39:44 -0700
• e95444fa74
  
  Pytorch version of lira Akemi Izuko 2024-11-29 17:16:09 -0700
• 3467c25882
  
  Lira: remove deadcode Akemi Izuko 2024-11-23 23:32:39 -0700
• bffecb459c
  
  Lira: train run shadow models Akemi Izuko 2024-11-23 23:19:01 -0700
• 91c61df0a8
  
  Lira: save load shadow models Akemi Izuko 2024-11-23 22:37:07 -0700
• 18426c7552
  
  Init Akemi Izuko 2024-11-20 12:11:10 -0700