pixeli99 / OwLore

Official Pytorch Implementation of "OwLore: Outlier-weighed Layerwise Sampled Low-Rank Projection for Memory-Efficient LLM Fine-tuning" by Pengxiang Li, Lu Yin, Xiaowei Gao, Shiwei Liu
https://arxiv.org/abs/2405.18380
28 stars 6 forks source link

Error in loading json file #2

Open TianjinYellow opened 3 months ago

TianjinYellow commented 3 months ago

Hi, when i run bash owlore_scripts/run_owlore_low_rank.sh mmlu, i encountered the below error, do you know how to solve it. BTW, is there a script for downloading your datasets? Best

en=False, disable_conversation_eos_token=False, conversation_template='empty') single file /gpfs/home2/huangti/Projects/OwLore/data/gsm/gsm8k.json Traceback (most recent call last): File "/gpfs/home2/huangti/Projects/OwLore/examples/finetune.py", line 62, in main() File "/gpfs/home2/huangti/Projects/OwLore/examples/finetune.py", line 54, in main dataset = Dataset(data_args) ^^^^^^^^^^^^^^^^^^ File "/gpfs/home2/huangti/Projects/OwLore/src/lmflow/datasets/dataset.py", line 82, in init json_data = json.load(fin) ^^^^^^^^^^^^^^ File "/home/huangti/miniconda3/envs/ft/lib/python3.11/json/init.py", line 293, in load return loads(fp.read(), ^^^^^^^^^^^^^^^^ File "/home/huangti/miniconda3/envs/ft/lib/python3.11/json/init.py", line 346, in loads return _default_decoder.decode(s) ^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/huangti/miniconda3/envs/ft/lib/python3.11/json/decoder.py", line 337, in decode obj, end = self.raw_decode(s, idx=_w(s, 0).end()) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/huangti/miniconda3/envs/ft/lib/python3.11/json/decoder.py", line 355, in raw_decode raise JSONDecodeError("Expecting value", s, err.value) from None json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0) Traceback (most recent call last):

pixeli99 commented 3 months ago

Hi, @TianjinYellow ,

Sorry for the late reply,

Did you download the dataset from here, https://huggingface.co/datasets/pengxiang/OwLore_Dataset? This Hugging Face space contains all the datasets we use.

h0ngxuanli commented 1 month ago

Directly git clone dataset from huggingface may lead to empty dataset. Manually downloading could solve this issue.