diff --git a/training/scripts/train.py b/training/scripts/train.py index 4d996fd..7f473dc 100755 --- a/training/scripts/train.py +++ b/training/scripts/train.py @@ -61,10 +61,10 @@ def train(config_path): from datasets import load_dataset import os - # Get dataset paths relative to script location - script_dir = Path(__file__).parent.parent - train_path = str(script_dir / ".." / ".." / "training" / "data" / "train.jsonl") - test_path = str(script_dir / ".." / ".." / "training" / "data" / "test.jsonl") + # Get dataset paths - dataset is in training/data/ relative to repo root + repo_root = Path(__file__).parent.parent.parent + train_path = str(repo_root / "training" / "data" / "train.jsonl") + test_path = str(repo_root / "training" / "data" / "test.jsonl") dataset = load_dataset( "json",