Issues

• do not save model by default

  marrlab/DomainLab#903 · Filter by author smilesun opened on Mar 18, 2025

• CI pytest runs out of resources at test_mk_exp_jigen.py

  bug
  marrlab/DomainLab#901 · Filter by author smilesun opened on Dec 11, 2024
  

• list_mu with zero elements should only be deleted for ERM, find a nicer solution

  acomplish >3hdesign_patternenhancementfbopt
  marrlab/DomainLab#899 · Filter by author smilesun opened on Dec 5, 2024
  

• move model_name and commit info to later columns in result.csv

  marrlab/DomainLab#894 · Filter by author smilesun opened on Oct 18, 2024

• when ForceSetpointSchrinkOnce, setpoint based selected performance should not be -1

  marrlab/DomainLab#893 · Filter by author smilesun opened on Oct 18, 2024
  

• add validation vs test performance as metric

  marrlab/DomainLab#892 · Filter by author smilesun opened on Oct 17, 2024

• compare before train reg loss and after 1st epoch reg loss, choose init_setpoint

  marrlab/DomainLab#888 · Filter by author smilesun opened on Oct 11, 2024
  

• implmente final_setpoint_ratio to stop decrease setpoint

  marrlab/DomainLab#887 · Filter by author smilesun opened on Oct 11, 2024

• trainer1_trainer_2, if trainer_2 only has tr_epoch implemented, then in trainer_1, call trainer_2

  marrlab/DomainLab#886 · Filter by author smilesun opened on Oct 11, 2024
  

• let each trainer's regularization loss return the name of the regularization

  marrlab/DomainLab#885 · Filter by author smilesun opened on Oct 10, 2024
  

• add boolean to fbopt if current output Pareto worse than setpoint

  fbopt
  marrlab/DomainLab#791 · Filter by author smilesun opened on Mar 8, 2024

• flag_info for model selection should be a diciontary containing different boolen varianble

  paper_jmlr_joss
  marrlab/DomainLab#790 · Filter by author smilesun opened on Mar 8, 2024
  

• scientific number support in fbopt branch

  fbopt
  marrlab/DomainLab#779 · Filter by author smilesun opened on Feb 8, 2024

• latex code generation for sampled hyperparameters of diva fbopt

  fbopthelp wanted
  marrlab/DomainLab#753 · Filter by author smilesun opened on Jan 18, 2024
  2

  

• Use python markdown to generate examples in readme

  paper_jmlr_joss
  marrlab/DomainLab#696 · Filter by author smilesun opened on Dec 15, 2023

• domainlab from PyPI under Python 3.11

  paper_jmlr_joss
  marrlab/DomainLab#691 · Filter by author agisga opened on Dec 14, 2023
  2

• Let user choose top2acc or valPerf as model selection from commandline

  benchmarkfboptprioritysuccess
  marrlab/DomainLab#574 · Filter by author smilesun opened on Oct 17, 2023
  1

  

• tensorboard in feedforward scheduler and fixed mu (trainer_basic)

  fboptin progress
  marrlab/DomainLab#507 · Filter by author smilesun opened on Oct 9, 2023
  7

  

• Let user choose the neural network q_zx

  enhancement
  marrlab/DomainLab#346 · Filter by author smilesun opened on Aug 30, 2023
  

• merge args.split with split

  marrlab/DomainLab#309 · Filter by author smilesun opened on Aug 16, 2023
  1

• decouple visitor from exp so that observer can use visitor independently from exp

  in progress
  marrlab/DomainLab#288 · Filter by author smilesun opened on Aug 4, 2023
  1