Skip to content
Snippets Groups Projects

Fix error in trainer function

Merged Driss KHALIL requested to merge Fix_checkpoint_process_error into master
2 unresolved threads

This MR fixes #14 (closed).

Edited by André Anjos

Merge request reports

Loading
Loading

Activity

Filter activity
  • Approvals
  • Assignees & reviewers
  • Comments (from bots)
  • Comments (from users)
  • Commits & branches
  • Edits
  • Labels
  • Lock status
  • Mentions
  • Merge request status
  • Tracking
342 342 if epoch >= max_epoch:
343 343 checkpointer.save("model_final_epoch", **arguments)
344 344
345 return lowest_validation_loss
  • 554 556 samples, model, valid_losses, device, criterion
    555 557 )
    556 558
    557 checkpointer_process(
    559 lowest_validation_loss = checkpointer_process(
  • Driss KHALIL added 1 commit

    added 1 commit

    • 1df9c09e - Fixing check pointer documentation

    Compare with previous version

  • Driss KHALIL added 1 commit

    added 1 commit

    • e681701e - Add unitest to checkpointer process

    Compare with previous version

  • André Anjos changed title from Fix error in trainer function ( lowest_validation_loss wasn't changing) to Fix error in trainer function

    changed title from Fix error in trainer function ( lowest_validation_loss wasn't changing) to Fix error in trainer function

  • André Anjos changed the description

    changed the description

  • merged

  • André Anjos mentioned in commit c77ba5e3

    mentioned in commit c77ba5e3

  • Please register or sign in to reply
    Loading