search

ASSERT-KTH / sequencer

Sequence-to-Sequence Learning for End-to-End Program Repair (IEEE TSE 2019). Open-science repo. http://arxiv.org/pdf/1901.01808
81 stars 42 forks source link

error in training #30

Open kevinGit777 opened 3 years ago

kevinGit777 commented 3 years ago

hi, I have trouble using src/sequencer-train.sh

FULL ERROR OUTPUT

sequencer-train.sh start Starting data preprocessing Please backup existing pt files: /home/hxw170013/APR/chai/results/Golden/final.train.pt, to avoid overwriting them! Starting training Traceback (most recent call last): File "train.py", line 120, in main(opt) File "train.py", line 51, in main single_main(opt, 0) File "/home/hxw170013/APR/chai/src/lib/OpenNMT-py/onmt/train_single.py", line 114, in main first_dataset = torch.load(glob.glob(opt.data + '.train.pt')[0]) File "/home/hxw170013/miniconda3/lib/python3.8/site-packages/torch/serialization.py", line 595, in load return _legacy_load(opened_file, map_location, pickle_module, pickle_load_args) File "/home/hxw170013/miniconda3/lib/python3.8/site-packages/torch/serialization.py", line 764, in _legacy_load magic_number = pickle_module.load(f, pickle_load_args) _pickle.UnpicklingError: invalid load key, 'v'. sequencer-train.sh done

I think it might because of the outdated value in version of chai/results/Golden/final.train.0.pt. But I am not sure, Can you help me out on this? Thank you!

kevinGit777 commented 3 years ago

I also notice that my final.train.0.pt. is different than the one in your repository, any reason that could happen? This is my final.train.0.pt looks like: version https://git-lfs.github.com/spec/v1 oid sha256:211527bbb96082da15f55556e5c8de408a1ea18ca09e3e7e5bd581b3403f8259 size 357618988

Anurag-Swarnim-Yadav commented 3 years ago

@kevinGit777 Install git lfs and initialize the files .