xiaoxinhe / tape Goto Github PK

Hi, I find that the num_classes of arxiv_2023 dataset is not 40, class 0 and 19 should be missing.

Thanks for your great work!

when I run the command

WANDB_DISABLED=True TOKENIZERS_PARALLELISM=False CUDA_VISIBLE_DEVICES=0 python -m core.trainLM dataset cora

I encountered a data loading error:

 File "/home/TAPE/core/data_utils/load_cora.py", line 93, in get_raw_text_cora
    with open(path+fn) as f:
FileNotFoundError: [Errno 2] No such file or directory: 'dataset/cora_orig/mccallum/cora/extractions/http:##www.sdsc.edu#~tbailey#papers#ismb96.ps'

Hello,
I'm getting some errors when running trainLM.py:

Traceback (most recent call last):
  File "/root/anaconda3/envs/TAPE/lib/python3.8/site-packages/transformers/utils/import_utils.py", line 1472, in _get_module
    return importlib.import_module("." + module_name, self.__name__)
  File "/root/anaconda3/envs/TAPE/lib/python3.8/importlib/__init__.py", line 127, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 1014, in _gcd_import
  File "<frozen importlib._bootstrap>", line 991, in _find_and_load
  File "<frozen importlib._bootstrap>", line 975, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 671, in _load_unlocked
  File "<frozen importlib._bootstrap_external>", line 843, in exec_module
  File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed
  File "/root/anaconda3/envs/TAPE/lib/python3.8/site-packages/transformers/training_args.py", line 75, in <module>
    from .trainer_pt_utils import AcceleratorConfig
  File "/root/anaconda3/envs/TAPE/lib/python3.8/site-packages/transformers/trainer_pt_utils.py", line 37, in <module>
    from torch.optim.lr_scheduler import LRScheduler
ImportError: cannot import name 'LRScheduler' from 'torch.optim.lr_scheduler' (/root/anaconda3/envs/TAPE/lib/python3.8/site-packages/torch/optim/lr_scheduler.py)

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "trainLM.py", line 1, in <module>
    from LMs.lm_trainer import LMTrainer
  File "/workspace/core/LMs/lm_trainer.py", line 4, in <module>
    from transformers import AutoTokenizer, AutoModel, TrainingArguments, Trainer, IntervalStrategy
  File "<frozen importlib._bootstrap>", line 1039, in _handle_fromlist
  File "/root/anaconda3/envs/TAPE/lib/python3.8/site-packages/transformers/utils/import_utils.py", line 1462, in __getattr__
    module = self._get_module(self._class_to_module[name])
  File "/root/anaconda3/envs/TAPE/lib/python3.8/site-packages/transformers/utils/import_utils.py", line 1474, in _get_module
    raise RuntimeError(
RuntimeError: Failed to import transformers.training_args because of the following error (look up to see its traceback):
cannot import name 'LRScheduler' from 'torch.optim.lr_scheduler' (/root/anaconda3/envs/TAPE/lib/python3.8/site-packages/torch/optim/lr_scheduler.py)

Dear Mr. He,

Firstly, I'd like to express my gratitude for sharing the code related to your remarkable research. I've encountered an issue during the reconstruction process. Specifically, I couldn't locate certain papers in the "extraction" folder.

Could you please let me know if there's an updated version of the Cora dataset available for public use? Alternatively, I might have misunderstood the instructions provided in the README.md.

To replicate the issue I faced, you can run the following code:

if __name__ == '__main__':
    data, data_citeid = get_cora_casestudy()
    data, text = get_raw_text_cora(use_text=True)
    print(data)
    print(data_citeid)

Executing the above should help you reproduce the error I encountered.

Thank you for your time and assistance.

Best regards,
csh

Hi, Thanks for the nice work and sharing the code. I want to know where the shallow embedding are from, or Is it consistent with the dataset in pyg?

Why is cuda incorrect according to the environment you installed, torch. cuda. is_ Availalabel() is False

Hello, I would like to ask about the result. I used the prediction result of LLM (cora dataset) you provided to calculate the accuracy. Whether using the original LLM response file or the csv file you processed, the accuracy rate is only about 16%. No such results appear for other datasets (pubmed, arxiv, and products).
Thanks so much for answering my questions!

Seem like you haven't uploaded the full file yet. When i training the GNNs

python -m core.trainEnsemble gnn.model.name GCN

It causes error No such file or directory: 'prt_lm/cora/microsoft/deberta-base-seed0.emb'
at line 47 in gnn_trainer.py

Hi Xiaoxin. In the paper, the topK LLM predictions are used as node features. Do we have ablation study on the accuracy of the LLM predictions? Thanks a lot!

Hi！Thank you for your open source and excellent work！

But I found that the title and abstract in some of the Cora data you provided are missing. Is this the issue I dealt with?

It seems that the cora_orig file is missing some files

Hi,

Thanks for the nice work and sharing the code.
I'm trying to run your code but have some problems.

1. When I runWANDB_DISABLED=True TOKENIZERS_PARALLELISM=False CUDA_VISIBLE_DEVICES=0,1 python -m core.trainLM dataset cora seed 0 >> cora_lm.out, I got data missing error: FileNotFoundError: [Errno 2] No such file or directory: 'dataset/cora_orig/mccallum/cora/extractions/http:##www.cs.ucl.ac.uk#staff#t.yu#ep97.ps'
2. When I run WANDB_DISABLED=True TOKENIZERS_PARALLELISM=False CUDA_VISIBLE_DEVICES=0,1 python -m core.trainLM dataset ogbn-arxiv seed 0 >> ogbn-arxiv_lm.out, I got File "/home/xxx/TAPE/core/data_utils/load_arxiv.py", line 24, in get_raw_text_arxiv data.edge_index = data.adj_t.to_symmetric() AttributeError: 'Tensor' object has no attribute 'to_symmetric'. I guess it's because of the OGB version conflict, could you tell me your OGB version or update your code? My OGB version is 1.3.6.

D:\anaconda3\envs\TAPE\lib\site-packages\accelerate\accelerator.py:432: FutureWarning: Passing the following arguments to Accelerator is deprecated and will be removed in version 1.0 of Accelerate: dict_keys(['dispatch_batches', 'split_batches', 'even_batches', 'use_seedable_sampler']). Please pass an accelerate.DataLoaderConfiguration instead:
dataloader_config = DataLoaderConfiguration(dispatch_batches=None, split_batches=False, even_batches=True, use_seedable_sampler=True)
warnings.warn(
Traceback (most recent call last):
File "E:\PycharmProjects\TAPE-main\core\trainLM.py", line 24, in
run(cfg)
File "E:\PycharmProjects\TAPE-main\core\trainLM.py", line 12, in run
trainer.train()
File "E:\PycharmProjects\TAPE-main\core\utils.py", line 80, in wrapper
ret = func(*args, **kw)
File "E:\PycharmProjects\TAPE-main\core\LMs\lm_trainer.py", line 119, in train
self.trainer.train()
File "D:\anaconda3\envs\TAPE\lib\site-packages\transformers\trainer.py", line 1624, in train
return inner_training_loop(
File "D:\anaconda3\envs\TAPE\lib\site-packages\transformers\trainer.py", line 1653, in _inner_training_loop
train_dataloader = self.get_train_dataloader()
File "D:\anaconda3\envs\TAPE\lib\site-packages\transformers\trainer.py", line 852, in get_train_dataloader
return self.accelerator.prepare(DataLoader(train_dataset, **dataloader_params))
File "D:\anaconda3\envs\TAPE\lib\site-packages\torch\utils\data\dataloader.py", line 238, in init
assert prefetch_factor > 0
TypeError: '>' not supported between instances of 'NoneType' and 'int'

Hi, after tuning the LLM by runing trainLM.py, on cora dataset, when I set feture_type is "TA", all the values of the features is 0. Moreover, the performance of models is low.

Hi, in the paper, you mentioned constructing the subset of ogbn-products dataset, but I don't find the code of the above function, could you please provide the code?

Thank you!

The README.md does not contain the links for downloading OGBN-Products(subset) dataset. Could you please provide it if available?

Within ‘ensemble_trainer.py’, the absence of a set seed in ‘load_data’ could lead to inconsistencies between the data split during the training process and the data split during the final ensemble evaluation.