linmou/cup

Code for CUP: Curriculum Learning based Prompt Tuning for Implicit Event Argument Extraction

cup's Issues

EEQA的问题

start_token = start_token[0]
IndexError: list index out of range

请问您在复现2020年EMNLP会议论文模型EEQA时，这个问题怎么解决的呀，刚刚看到了您在 原作者下面 的问题提问，所以前来冒昧打扰！ 期待您的恢复！

evt_triggers not found in dataset

Hi, I'm trying to get the results described in your paper, and facing this issue while running the code. But 'evt_triggers' cannot be found in WikiEvents dataset, may I know how can I solve this issue?

82 args are discarded due to the limitation of context len.
461 args in total. Avg doc_len =175.31929824561402
Traceback (most recent call last):
File "CUP/train.py", line 626, in
curriculum_learnig(args)
File "CUP/train.py", line 474, in curriculum_learnig
plm, tokenizer, model_config, WrapperClass, template, dataset, train_dataloader, amr_dataloader, validation_dataloader, test_dataloader = setup_all(
File "CUP/train.py", line 201, in setup_all
dataset['AmrGraph_train'] = AmrGraphProcesser().get_train_examples(
File "CUP/openprompt/data_utils/data_processor.py", line 100, in get_train_examples
return self.get_examples(data_dir, "train",**kwargs)
File "CUP/DataProcessers.py", line 687, in get_examples
event_type = instance["evt_triggers"][0][2][0][0]
KeyError: 'evt_triggers'

file not found

can't find the python file indicated in RAMStest.py in test_WikiEvents.sh

No module named 'models.constrained_gen'

Hello! I am also trying to reproduce the results of your paper, but there is a "No module named 'models.constrained_gen'" problem. Is the part of models missing from the code you uploaded? Can you help me solve this problem, thank you!