NotImplementedError about bgca

[ntudy]

Hi, you need to pass the "train_by_pair" argument

[ntudy]

你好，所有的参数设置可以在setup.py里面查看详细的解释。

缺失模块的原因是因为调用main函数的时候，未从命令行传入"--train_by_pair", 具体调用方式可以参考以下脚本最后一行：
https://github.com/DAMO-NLP-SG/BGCA/blob/master/script/run_ate.sh#L34

[3xiaoxin]

Traceback (most recent call last):
File "C:\Users\孤星\Desktop\BGCA-master\code\main.py", line 329, in
runed_dirs = run_multiple_seeds(args, seed_list)
File "C:\Users\孤星\Desktop\BGCA-master\code\main.py", line 319, in run_multiple_seeds
main(args)
File "C:\Users\孤星\Desktop\BGCA-master\code\main.py", line 137, in main
tokenizer = AutoTokenizer.from_pretrained(args.model_name_or_path, use_fast=False)
File "C:\Program Files\Python39\lib\site-packages\transformers\models\auto\tokenization_auto.py", line 471, in from_pretrained
tokenizer_config = get_tokenizer_config(pretrained_model_name_or_path, **kwargs)
File "C:\Program Files\Python39\lib\site-packages\transformers\models\auto\tokenization_auto.py", line 332, in get_tokenizer_config
resolved_config_file = get_file_from_repo(
File "C:\Program Files\Python39\lib\site-packages\transformers\utils\hub.py", line 678, in get_file_from_repo
resolved_file = cached_path(
File "C:\Program Files\Python39\lib\site-packages\transformers\utils\hub.py", line 282, in cached_path
output_path = get_from_cache(
File "C:\Program Files\Python39\lib\site-packages\transformers\utils\hub.py", line 545, in get_from_cache
raise ValueError(
ValueError: Connection error, and we cannot find the requested files in the cached path. Please try again or make sure your Internet connection is on.
您好，我想问问您报的这个错误怎么解决呢？（非常感谢您的回答）

[ntudy]

你好，这个应该是huggingface的模型下载问题，可以参考https://huggingface.co/t5-base 进行下载。

如果是国内网络引起的话，可以考虑通过镜像下载或者提前下载至本地。

[3xiaoxin]

Traceback (most recent call last):
File "C:\Users\孤星\Desktop\dataset\BGCA-master\code\main.py", line 329, in
runed_dirs = run_multiple_seeds(args, seed_list)
File "C:\Users\孤星\Desktop\dataset\BGCA-master\code\main.py", line 319, in run_multiple_seeds
main(args)
File "C:\Users\孤星\Desktop\dataset\BGCA-master\code\main.py", line 137, in main
tokenizer = AutoTokenizer.from_pretrained(args.model_name_or_path)
File "C:\Users\孤星\AppData\Roaming\Python\Python39\site-packages\transformers\models\auto\tokenization_auto.py", line 528, in from_pretrained
return tokenizer_class.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
File "C:\Users\孤星\AppData\Roaming\Python\Python39\site-packages\transformers\tokenization_utils_base.py", line 1780, in from_pretrained
return cls._from_pretrained(
File "C:\Users\孤星\AppData\Roaming\Python\Python39\site-packages\transformers\tokenization_utils_base.py", line 1806, in _from_pretrained
slow_tokenizer = (cls.slow_tokenizer_class).from_pretrained(
File "C:\Users\孤星\AppData\Roaming\Python\Python39\site-packages\transformers\tokenization_utils_base.py", line 1915, in from_pretrained
tokenizer = cls(*init_inputs, **init_kwargs)
File "C:\Users\孤星\AppData\Roaming\Python\Python39\site-packages\transformers\models\t5\tokenization_t5.py", line 152, in init
self.sp_model.Load(vocab_file)
File "C:\Program Files\Python39\lib\site-packages\sentencepiece_init.py", line 367, in Load
return self.LoadFromFile(model_file)
File "C:\Program Files\Python39\lib\site-packages\sentencepiece_init.py", line 171, in LoadFromFile
return _sentencepiece.SentencePieceProcessor_LoadFromFile(self, arg)
TypeError: not a string

[ntudy]

请确认相关package版本一致： https://github.com/DAMO-NLP-SG/BGCA/blob/master/requirements.txt
并检查hugging face下载文件是否完整： https://huggingface.co/t5-base/tree/main

可以尝试在python shell / notebook中检查模型是否能正常加载，再尝试运行当前repo代码

[3xiaoxin]

您好，代码跑到中途的时候还是出现了这个问题：
Traceback (most recent call last):
File "C:\BGCA-master\BGCA-master\code\main.py", line 329, in
collate_seed_results(args, runed_dirs)
File "C:\BGCA-master\BGCA-master\code\main.py", line 291, in collate_seed_results
avg_n_seeds_by_pair(args.output_dir, runed_dirs, decode_txt, args.n_runs)
File "C:\BGCA-master\BGCA-master\code\eval_utils.py", line 391, in avg_n_seeds_by_pair
score_dict_i = json.load(open(f"{dir_}/score/test_{decode_txt}_score.json","r"))
FileNotFoundError: [Errno 2] No such file or directory: '../outputs/uabsa/cross_domain\0609_2051-experinments/seed-3/laptop-rest/score/test_greedy_score.json'

[ntudy]

你好，问题应该出自windows和Linux系统之间的路径差异， '../outputs/uabsa/cross_domain\0609_2051-experinments/seed-3/laptop-rest/score/test_greedy_score.json' 路径中“cross_domain\0609_2051"处被setup.py 105行处代码自动添加为了反斜，将该处代码改成以正斜字符串联路径应该可以解决

[3xiaoxin]

您好，源代码中只有正斜字符，我修改了很多次，还是跑不通，这是一部分源码

Set up output directory

output_dir = f"../outputs/{args.task}/{args.dataset}"
os.makedirs(output_dir, exist_ok=True)
# Create a timestamped directory
timestamp = datetime.now().strftime("%m%d_%H%M")
output_dir = os.path.join(output_dir, f"{timestamp}-{args.name}").replace("/", "\\")
os.makedirs(output_dir, exist_ok=True)
args.output_dir = output_dir
return args

以下是报错部分：（已经跑通了一段时间，但是跑到这里的时候就报错了）
Traceback (most recent call last):
File "C:\BGCA-master\BGCA-master\code\main.py", line 329, in
collate_seed_results(args, runed_dirs)
File "C:\BGCA-master\BGCA-master\code\main.py", line 291, in collate_seed_results
avg_n_seeds_by_pair(args.output_dir, runed_dirs, decode_txt, args.n_runs)
File "C:\BGCA-master\BGCA-master\code\eval_utils.py", line 391, in avg_n_seeds_by_pair
score_dict_i = json.load(open(f"{dir_}/score/test_{decode_txt}_score.json","r")).replace("/", "\")
FileNotFoundError: [Errno 2] No such file or directory: '..\outputs\uabsa\cross_domain\0611_1422-experinments/seed-3/laptop-rest/score/test_greedy_score.json'

[ntudy]

通过你提供的结果来看的话，output_dir = os.path.join(output_dir, f"{timestamp}-{args.name}").replace("/", "\") 会导致正反斜混用（如截图报错所示，路径前半段为双反斜，后半段为正斜），因为这处代码只涉及到输出文件夹，后续的具体存档文件路径处代码没有修改成双反斜。

简单的做法是可以尝试全部统一成正斜，具体是只将setup.py 105行处改为 output_dir = f"{output_dir}/{timestamp}-{args.name}"。

[3xiaoxin]

Traceback (most recent call last):
File "C:\BGCA-master\BGCA-master\code\main.py", line 329, in
collate_seed_results(args, runed_dirs)
File "C:\BGCA-master\BGCA-master\code\main.py", line 291, in collate_seed_results
avg_n_seeds_by_pair(args.output_dir, runed_dirs, decode_txt, args.n_runs)
File "C:\BGCA-master\BGCA-master\code\eval_utils.py", line 391, in avg_n_seeds_by_pair
score_dict_i = json.load(open(f"{dir_}/score/test_{decode_txt}_score.json","r"))
FileNotFoundError: [Errno 2] No such file or directory: '../outputs/uabsa/cross_domain/0612_1721-experinments/seed-3/laptop-rest/score/test_greedy_score.json'

[ntudy]

你好，请问在文件系统中能看到该文件吗？

[3xiaoxin]

文件系统中是指这一部分吗

[ntudy]

是的，里面有test_greed_score.json文件吗？

[3xiaoxin]

没有的，到这一步就点不开这个文件了

[ntudy]

你好，我注意到你的tag tokens 也是空的，能否上传你的运行脚本和log文件方便我查看

[3xiaoxin]

好的，可以的，等我有时间我在给您上传

[ntudy]

可以发送至我邮箱：[email protected]