Commits · 17-adding-new-datasets · Luc Giffon / bolsonaro · GitLab

Snippets Groups Projects

Dec 18, 2019
- - Add an option to not use the best hyperparameters file; · 880ff78f
  Charly Lamothe authored 5 years ago
  880ff78f
- - Replace the futures concurrence by joblib Parallel (and add optional tqdm progress bar); · 11017545
  Charly Lamothe authored 5 years ago
  11017545
Dec 01, 2019
- Fix few typos in new compute_hyperparams version. · 31d12659
  Charly Lamothe authored 5 years ago
  
  31d12659
- Remove useless Real import · 39d4540f
  Charly Lamothe authored 5 years ago
  
  39d4540f
- Rename 'use_different_seed_number' parameter to 'use_variable_seed_number' · 7494dbc4
  Charly Lamothe authored 5 years ago
  
  7494dbc4
- Compute best hyperparameters over k seeds, and select the most commons. · e319343d
  Charly Lamothe authored 5 years ago
  
  e319343d
- By default, compute the hyperparams of the whole dataset list. · 254e846e
  Charly Lamothe authored 5 years ago
  
  254e846e
- - Add standard dataset scaling for dataset normalization; · f61314f3
  Charly Lamothe authored 5 years ago
  f61314f3
Nov 22, 2019
- - Add missing LOG_PATH in two init files; · a86dff87
  Charly Lamothe authored 5 years ago
  a86dff87
- When training, look if there is bayesian search results, if yes use this.... · c66d117d
  Léo Bouscarrat authored 5 years ago
  c66d117d
Nov 20, 2019
- Add functions to do bayesian hyperparameters search · bf5803b6
  Léo Bouscarrat authored 5 years ago
  
  bf5803b6