diff --git a/training/scripts/prepare_dataset.py b/training/scripts/prepare_dataset.py index e7313aa..b10230e 100755 --- a/training/scripts/prepare_dataset.py +++ b/training/scripts/prepare_dataset.py @@ -62,11 +62,11 @@ def prepare_dataset(input_file, output_file, test_size=0.05): test_data = formatted[split_point:] # Save - with open(output_file.parent / "train.jsonl", "w") as f: + with open(output_dir / "train.jsonl", "w") as f: for item in train_data: f.write(json.dumps(item) + "\n") - with open(output_file.parent / "test.jsonl", "w") as f: + with open(output_dir / "test.jsonl", "w") as f: for item in test_data: f.write(json.dumps(item) + "\n")