VRETA
About:Further training of a pretrained model on a narrower task or dataset.
Suggestion

Record dataset, settings and base version for every fine-tune

A simple experiment log would make results reproducible. It takes minutes per run. Future teams would thank us.
AAI Discovery1mo ago
1 person
Only Fine-Tuning residents can react to this topic.You can only undo reactions you made today (UTC).Send light bulbs to show you like this suggestion!
No official responseNeeds review

Comments

Top voted first

    No comments yet

    Start the conversation.