Skip to content
Permalink
Branch: master
Commits on Jul 9, 2019
  1. DOC Add MNIST torchvision example (#490)

    thomasjpfan committed Jul 9, 2019
Commits on Jul 8, 2019
  1. BUG Adds eps to loss to prevent nans (#491)

    thomasjpfan authored and BenjaminBossan committed Jul 8, 2019
    Adds eps to loss to prevent nans
Commits on May 13, 2019
Commits on May 2, 2019
  1. [MRG] ENH Uses checkpoint internally (#463)

    thomasjpfan authored and BenjaminBossan committed May 2, 2019
    Fix a bug in TrainEndCheckpoint that prevented it from being cloned
    
    TrainEndCheckpoint now wraps Checkpoint instead of inheriting from it. This way, there is no longer a mismatch in the init parameters.
Commits on May 1, 2019
  1. [MRG] Adds Batch Count in History (#445)

    thomasjpfan authored and BenjaminBossan committed May 1, 2019
    Adds batch_count to history
    
    The batch_count is recorded during the fit_loop. Also adds batch_count support for lr_scheduler.
Commits on Apr 23, 2019
Commits on Feb 13, 2019
  1. DOC Adds FAQ entry on pytorch's Dataset (#439)

    thomasjpfan authored and BenjaminBossan committed Feb 13, 2019
    DOC Adds FAQ entry on pytorch's Dataset
Commits on Jan 21, 2019
  1. [MRG] Update transform learning to use class in NeutralNet.__init__ (#…

    thomasjpfan authored and ottonemo committed Jan 21, 2019
    …425)
    
    * DOC: Uses class based init
    
    * DOC: Fixes bugs to update
Commits on Jan 8, 2019
  1. Trigger CI

    thomasjpfan committed Jan 8, 2019
Commits on Jan 7, 2019
  1. STY: Typo

    thomasjpfan committed Jan 7, 2019
Commits on Jan 4, 2019
  1. DOC: Adds more details

    thomasjpfan committed Jan 4, 2019
Commits on Dec 12, 2018
  1. [MRG] Adds test for pytorch 1.0 (#404)

    thomasjpfan authored and ottonemo committed Dec 12, 2018
    * ENH: Tests pytorch 1.0.0
    
    * RFC: Move pytorch dep to matrix
    
    * BUG: Fix
    
    * ENH: Adds python 3.7
    
    * BLD: Test for pytorch 1.0.0
    
    * RFC: Tests script
    
    * BUG: Uses xenial for 3.7 support
    
    * TST: Support last two versions of pytorch
    
    * TST: Removes 0.4.1
    
    * RFC: Removes matrix
    
    * TST: Uses requirements file to fix versions
    
    * TST: Fix tabulate version
    
    * Revert "TST: Fix tabulate version"
    
    This reverts commit 77c314f.
    
    * REV: Undo
    
    * ENH: Removes tabulate from install
    
    * TST: Adjusts pip install
    
    * REV: Uses python setup directly
    
    * ENH: Uses pip to install
    
    * REV: Uses python setup install
    
    * REV: Uses pip to install
Commits on Dec 11, 2018
  1. [MRG] Warn user when using possibly wrong checkpoint monitor (#403)

    thomasjpfan authored and ottonemo committed Dec 11, 2018
    * ENH: Adds warning for checkpoint
    
    * RFC: Updates warning
    
    * DOC: Adds to changelog
    
    * TST: LoadInitState must come before scoring
    
    * DOC: Adds description
Commits on Nov 29, 2018
  1. RFC: Changes default fn_prefix in TrainEndCheckpoint to train_end_ (#391

    thomasjpfan authored and benjamin-work committed Nov 29, 2018
    )
    
    * RFC: Changes default prefix for TrainEndCheckpoint from final_ to train_end_
Commits on Nov 9, 2018
  1. DOC: Minor fixes to save load doc

    thomasjpfan committed Nov 9, 2018
Commits on Nov 8, 2018
  1. ENH: Adds TeeGenerator for named_parameters (#379)

    thomasjpfan authored and benjamin-work committed Nov 8, 2018
    * ENH: Less overhead caused by named_parameters in on_grad_computed.
Commits on Nov 7, 2018
  1. [MRG]: Puts deployment into a script (#380)

    thomasjpfan authored and benjamin-work committed Nov 7, 2018
    Use "./scripts/deploy.sh stage" or "./scripts/deploy.sh live" to upload pypi package.
  2. TST: Asserts deprecation to remove warnings

    thomasjpfan committed Nov 3, 2018
Commits on Nov 6, 2018
  1. ENH: Early exit for train_loss_score and valid_loss_score already in …

    thomasjpfan authored and benjamin-work committed Nov 6, 2018
    …history (#381)
    
    * ENH: Early exit for metrics already in history to reduce overhead.
Commits on Nov 5, 2018
  1. BUG: Disables cache

    thomasjpfan committed Nov 5, 2018
Commits on Nov 1, 2018
  1. DOC: Adds colab support to MNIST tutorial (#373)

    thomasjpfan authored and benjamin-work committed Nov 1, 2018
    * DOC: Adds colab support to MNIST
Commits on Oct 29, 2018
  1. DOC: Add links to colab in notebook (#375)

    thomasjpfan authored and benjamin-work committed Oct 29, 2018
Commits on Oct 26, 2018
  1. DOC: Adds advanced usage to Colab (#371)

    thomasjpfan authored and ottonemo committed Oct 26, 2018
    * DOC: Adds support for colab
    
    * DOC: Adds link into tutorials page
    
    * RFC: Less diffs
Commits on Oct 25, 2018
  1. DOC: Basic Usage - Adds support for colab (#368)

    thomasjpfan authored and ottonemo committed Oct 25, 2018
    * DOC: Adds support for colab
    
    * Document change in changelog
Commits on Oct 24, 2018
  1. [MRG]: Updates transfer learning notebook in Colab (#366)

    thomasjpfan authored and ottonemo committed Oct 24, 2018
    * DOC: Fixes doc for google colab
    
    * DOC: Convert to Freezer
    
    * DOC: Uses master skorch
    
    * DOC: Fixes grammer
Commits on Oct 22, 2018
  1. [MRG] Adds optimizer and history to save/load_params (#360)

    thomasjpfan authored and benjamin-work committed Oct 22, 2018
    * Adds optimizer and history to save_params and load_params. Deprecates using f.
    * Adds the optimizer, basedir, fn_prefix to Checkpoint.
    *  Deprecates save_history and load_history in favor of save_params and load_params to save history. This simplifies the skorch api to only needing save_params and load_params to save/load state.
    *  Deprecates using the positional argument in favor of keywords.
    *  Adds history serialization (in the History class)
    *  Adds loading params from checkpoint. net.load_params(checkpoint=cp)
    *  Adds LoadInitState to load checkpoint during training.
    * RFC: Adds event_name parameter (uses None to disable placing events)
    * ENH: Adds history serialization
    * Adds TrainEndCheckpoint
Commits on Oct 16, 2018
  1. DOC: Updates seg tutorial to use freezer (#359)

    thomasjpfan authored and ottonemo committed Oct 16, 2018
    * DOC: Updates seg tutorial to use freezer
    
    * DOC: Rename file
Commits on Oct 14, 2018
  1. BUG: Fixes ReduceLROnPlateau when mode == max

    thomasjpfan committed Oct 14, 2018
Commits on Oct 4, 2018
  1. Merge remote-tracking branch 'upstream/master' into flakey

    thomasjpfan committed Oct 4, 2018
  2. RFC: Uses flaky

    thomasjpfan committed Oct 4, 2018
  3. TST: Fixes deprecation warnings (#351)

    thomasjpfan authored and ottonemo committed Oct 4, 2018
  4. TST: Less flakey tests

    thomasjpfan committed Oct 4, 2018
Commits on Sep 18, 2018
  1. RFC: Adapts filter_parameters to use iterator

    thomasjpfan committed Sep 18, 2018
Commits on Sep 17, 2018
  1. BUG: Uses rm -f when removing files in CI when caching (#345)

    thomasjpfan authored and benjamin-work committed Sep 17, 2018
    * BUG: Sudos when removing files
    * TST: Removes sudo uses -f
Older
You can’t perform that action at this time.