zhangyiss/mlx-examples
1
0
Fork 0
mirror of https://github.com/ml-explore/mlx-examples.git synced 2025-06-24 17:31:18 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

LoRA: Split small function

Browse Source
This commit is contained in:
madroid 2024-09-20 11:06:06 +08:00
parent bfd4ba2347
commit 3f6a5f19fd
1 changed files with 55 additions and 44 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
llms/mlx_lm/tuner/datasets.py
View File

@ -117,8 +117,17 @@ def create_dataset(path: Path, tokenizer: PreTrainedTokenizer = None):
) )
def load_dataset(args, tokenizer: PreTrainedTokenizer): def create_local_dataset(args, tokenizer: PreTrainedTokenizer):
if getattr(args, "hf_dataset", None) is not None: names = ("train", "valid", "test")
data_path = Path(args.data)
train, valid, test = [
create_dataset(data_path / f"{n}.jsonl", tokenizer) for n in names
]
return train, valid, test
def create_hf_dataset(args, tokenizer: PreTrainedTokenizer):
import datasets import datasets
hf_args = args.hf_dataset hf_args = args.hf_dataset
@ -158,13 +167,15 @@ def load_dataset(args, tokenizer: PreTrainedTokenizer):
else: else:
test = [] test = []
else: return train, valid, test
names = ("train", "valid", "test")
data_path = Path(args.data)
def load_dataset(args, tokenizer: PreTrainedTokenizer):
if getattr(args, "hf_dataset", None) is not None:
train, valid, test = create_hf_dataset(args, tokenizer)
else:
train, valid, test = create_local_dataset(args, tokenizer)
train, valid, test = [
create_dataset(data_path / f"{n}.jsonl", tokenizer) for n in names
]
if args.train and len(train) == 0: if args.train and len(train) == 0:
raise ValueError( raise ValueError(
"Training set not found or empty. Must provide training set for fine-tuning." "Training set not found or empty. Must provide training set for fine-tuning."