The dataset used to fine-tune the model.
Only one dataset is allowed per custom model, which is split 80-20 for training and validating. You must provide the dataset in a JSON Lines (JSONL) file. Each line in the JSONL file must have the format:
{@code {"prompt": "", "completion": ""}}
The dataset used to fine-tune the model.
Only one dataset is allowed per custom model, which is split 80-20 for training and validating. You must provide the dataset in a JSON Lines (JSONL) file. Each line in the JSONL file must have the format:
{@code {"prompt": "", "completion": ""}}