diff --git a/tokenise+train.py b/tokenise+train.py
index 5c99e73..a299724 100644
--- a/tokenise+train.py
+++ b/tokenise+train.py
@@ -57,7 +57,7 @@ def train(ouputdir: Path, blocksize: int, vocabsize: int, num_steps: int, gpu: b
 	return "Done!"
 
 
-def encode(filepath: str, blocksize: int, vocabsize: int, ouputdir: Path, verbose: bool = False) -> str:
+def encode(filepath: str, blocksize: int, vocabsize: int, ouputdir: Path, linebyline: bool, verbose: bool = False) -> str:
 
 	f_path = Path(filepath)
 
@@ -88,10 +88,10 @@ def encode(filepath: str, blocksize: int, vocabsize: int, ouputdir: Path, verbos
 	print(dataset_fn)
 
 	if type(text) is str:
-		data = TokenDataset(file_path=text, tokenizer_file=tok_fn, block_size=blocksize,  line_by_line=True)
+		data = TokenDataset(file_path=text, tokenizer_file=tok_fn, block_size=blocksize,  line_by_line=linebyline)
 	else:
 		texts = [x.read_text() for x in text]
-		data = TokenDataset(texts=texts, tokenizer_file=tok_fn, block_size=blocksize, line_by_line=True)
+		data = TokenDataset(texts=texts, tokenizer_file=tok_fn, block_size=blocksize, line_by_line=linebyline)
 	data.save(cache_destination=dataset_fn)
 
 	return "encode success"
@@ -105,6 +105,7 @@ def main() -> int:
 	p.add_argument("-v", "--vocabsize", type=int, default=1000)
 	p.add_argument("--ouputdir", type=str, default="data/tokens+models/")
 	p.add_argument("--gpu", action="store_true")
+	p.add_argument("--line_by_line", action="store_true")
 
 	args = p.parse_args()
 
@@ -125,7 +126,7 @@ def main() -> int:
 	else: 
 		output_dir.mkdir()
 
-	encode(filepath=args.text, blocksize=args.blocksize, vocabsize=args.vocabsize, ouputdir=output_dir)
+	encode(filepath=args.text, blocksize=args.blocksize, vocabsize=args.vocabsize, ouputdir=output_dir, linebyline=args.line_by_line)
 	train(ouputdir=output_dir, blocksize=args.blocksize, vocabsize=args.vocabsize, num_steps=args.numsteps, gpu=args.gpu)