turbid cloud
Hi there, thanks for your help, I couldn't find the answer to this anywhere:
When using the following command to fine tune a model "openai api fine_tunes.create -t training_dataset.jsonl -v validation_dataset.jsonl -m ada ....." for classification, The validation dataset is it used to actually finetune the model on each epoch or is it used like a held out (test) dataset just to measure the performance ?
icy ravine