taishan1994 / bert-bilstm-crf Goto Github PK

大佬好，我在准备好自己的数据做微调时，数据报错：

`
Traceback (most recent call last):
File "main.py", line 190, in
main(data_name)
File "main.py", line 182, in main
train.train()
File "main.py", line 45, in train
for step, batch_data in enumerate(self.train_loader):
File "/root/miniconda3/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 363, in next
data = self._next_data()
File "/root/miniconda3/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 989, in _next_data
return self._process_data(data)
File "/root/miniconda3/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 1014, in _process_data
data.reraise()
File "/root/miniconda3/lib/python3.8/site-packages/torch/_utils.py", line 395, in reraise
raise self.exc_type(msg)
RuntimeError: Caught RuntimeError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/root/miniconda3/lib/python3.8/site-packages/torch/utils/data/_utils/worker.py", line 185, in _worker_loop
data = fetcher.fetch(index)
File "/root/miniconda3/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py", line 47, in fetch
return self.collate_fn(data)
File "/root/miniconda3/lib/python3.8/site-packages/torch/utils/data/_utils/collate.py", line 74, in default_collate
return {key: default_collate([d[key] for d in batch]) for key in elem}
File "/root/miniconda3/lib/python3.8/site-packages/torch/utils/data/_utils/collate.py", line 74, in
return {key: default_collate([d[key] for d in batch]) for key in elem}
File "/root/miniconda3/lib/python3.8/site-packages/torch/utils/data/_utils/collate.py", line 55, in default_collate
return torch.stack(batch, 0, out=out)
RuntimeError: stack expects each tensor to be equal size, but got [512] at entry 0 and [511] at entry 2

`
所有的requirements都按照md文件里的版本安装，请问还有哪里的代码需要调整吗，谢谢！

D:\Anaconda\Anaconda\envs\pytorch\python.exe D:/python/BERT-BILSTM-CRF-main/BERT-BILSTM-CRF-main/main.py
['O', 'B-故障设备', 'I-故障设备', 'B-故障原因', 'I-故障原因']
{'O': 0, 'B-故障设备': 1, 'I-故障设备': 2, 'B-故障原因': 3, 'I-故障原因': 4}
Some weights of the model checkpoint at ./model_hub/chinese-bert-wwm-ext/ were not used when initializing BertModel: ['cls.seq_relationship.bias', 'cls.predictions.transform.LayerNorm.bias', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.decoder.weight', 'cls.predictions.transform.dense.weight', 'cls.seq_relationship.weight', 'cls.predictions.bias', 'cls.predictions.transform.dense.bias']

• This IS expected if you are initializing BertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
• This IS NOT expected if you are initializing BertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
  D:\Anaconda\Anaconda\envs\pytorch\lib\site-packages\torch\nn\modules\rnn.py:62: UserWarning: dropout option adds dropout after all but last recurrent layer, so non-zero dropout expects num_layers greater than 1, but got dropout=0.1 and num_layers=1
  warnings.warn("dropout option adds dropout after all but last "

大佬可以帮忙看一下不

大佬好，我用了你的数据训练模型没有问题，但我自己标了一些数据后生成的训练集总是报错，报错信息如下：
Traceback (most recent call last):
File "F:\education\aaa-education\BERT-BILSTM-CRF\main.py", line 191, in
main(data_name)
File "F:\education\aaa-education\BERT-BILSTM-CRF\main.py", line 150, in main
train_data = [json.loads(d) for d in train_data]
File "F:\education\aaa-education\BERT-BILSTM-CRF\main.py", line 150, in
train_data = [json.loads(d) for d in train_data]
File "C:\Users\lsc11232.conda\envs\deepke\lib\json_init_.py", line 357, in loads
return _default_decoder.decode(s)
File "C:\Users\lsc11232.conda\envs\deepke\lib\json\decoder.py", line 337, in decode
obj, end = self.raw_decode(s, idx=_w(s, 0).end())
File "C:\Users\lsc11232.conda\envs\deepke\lib\json\decoder.py", line 353, in raw_decode
obj, end = self.scan_once(s, idx)
json.decoder.JSONDecodeError: Expecting ',' delimiter: line 1 column 345 (char 344)

以下是我的部分数据：
{"id": "AT0001", "text": ["一", "个", "给", "定", "集", "合", "中", "的", "元", "素", "是", "互", "不", "相", "同", "的", "，", "也", "就", "是", "说", "，", "集", "合", "中", "的", "元", "素", "是", "不", "重", "复", "出", "现", "的"], "labels": ["O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O"]}
{"id": "AT0002", "text": ["集", "合", "论", "的", "基", "本", "理", "论", "创", "立", "于", "1", "9", "世", "纪", "，", "关", "于", "集", "合", "的", "最", "简", "单", "的", "说", "法", "就", "是", "在", "朴", "素", "集", "合", "论", "（", "最", "原", "始", "的", "集", "合", "论", "）", "中", "的", "定", "义", "，", "即", "集", "合", "是", "“", "确", "定", "的", "一", "堆", "东", "西", "”", "，", "集", "合", "里", "的", "“", "东", "西", "”", "则", "称", "为", "元", "素"], "labels": ["B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-KNOW", "I-KNOW"]}
{"id": "AT0003", "text": ["现", "代", "的", "集", "合", "一", "般", "被", "定", "义", "为", "：", "由", "一", "个", "或", "多", "个", "确", "定", "的", "元", "素", "所", "构", "成", "的", "整", "体"], "labels": ["O", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O"]}
{"id": "AT0004", "text": ["集", "合", "是", "指", "具", "有", "某", "种", "特", "定", "性", "质", "的", "具", "体", "的", "或", "抽", "象", "的", "对", "象", "汇", "总", "而", "成", "的", "集", "体"], "labels": ["B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O"]}
{"id": "AT0005", "text": ["其", "中", "，", "构", "成", "集", "合", "的", "这", "些", "对", "象", "则", "称", "为", "该", "集", "合", "的", "元", "素"], "labels": ["O", "O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "B-KNOW", "I-KNOW"]}
{"id": "AT0006", "text": ["集", "合", "中", "元", "素", "的", "数", "目", "称", "为", "集", "合", "的", "基", "数", "，", "集", "合", "A", "的", "基", "数", "记", "作", "c", "a", "r", "d", "(", "A", ")"], "labels": ["B-KNOW", "I-KNOW", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "B-KNOW", "I-KNOW", "O", "B-KNOW", "I-KNOW", "O", "B-KNOW", "I-KNOW", "O", "O", "B-KNOW", "I-KNOW", "O", "O", "O", "O", "O", "O", "O", "O", "O"]}

大佬好，我用了你的数据训练模型没有问题，但我自己标了一些数据后生成的训练集总是报错，报错信息如下：

Traceback (most recent call last): File "main.py", line 190, in <module> main(data_name) File "main.py", line 153, in main dev_data = [json.loads(d) for d in dev_data] File "main.py", line 153, in <listcomp> dev_data = [json.loads(d) for d in dev_data] File "/root/miniconda3/lib/python3.8/json/__init__.py", line 357, in loads return _default_decoder.decode(s) File "/root/miniconda3/lib/python3.8/json/decoder.py", line 337, in decode obj, end = self.raw_decode(s, idx=_w(s, 0).end()) File "/root/miniconda3/lib/python3.8/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)

我核对了我的文件格式，都是按照你的来写，也尝试过更换标准的json文件格式，但都是报错，我的部分数据是这种格式：

{"id": "TEX0001", "text":["6", ".", "为", "进", "一", "步", "加", "大", "增", "值", "税", "留", "抵", "退", "税", "政", "策", "实", "施", "力", "度", "，", "着", "力", "稳", "市", "场", "主", "体", "稳", "就", "业", "，", "现", "将", "扩", "大", "全", "额", "退", "还", "增", "值", "税", "留", "抵", "税", "额", "政", "策", "行", "业", "范", "围", "有", "关", "政", "策", "公", "告", "如", "下", "："], "labels":["O", "O", "O", "O", "O", "O", "O", "O", "B-税费种类", "I-税费种类", "I-税费种类", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-税费种类", "I-税费种类", "I-税费种类", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O"]} {"id": "TEX0003", "text":["8", ".", "（", "一", "）", "符", "合", "条", "件", "的", "批", "发", "零", "售", "业", "等", "行", "业", "企", "业", "，", "可", "以", "自", "2", "0", "2", "2", "年", "7", "月", "纳", "税", "申", "报", "期", "起", "向", "主", "管", "税", "务", "机", "关", "申", "请", "退", "还", "增", "量", "留", "抵", "税", "额", "。"], "labels":["O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-所属行业", "I-所属行业", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O"]} {"id": "TEX0004", "text":["9", ".", "（", "二", "）", "符", "合", "条", "件", "的", "批", "发", "零", "售", "业", "等", "行", "业", "企", "业", "，", "可", "以", "自", "2", "0", "2", "2", "年", "7", "月", "纳", "税", "申", "报", "期", "起", "向", "主", "管", "税", "务", "机", "关", "申", "请", "一", "次", "性", "退", "还", "存", "量", "留", "抵", "税", "额", "。"], "labels":["O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "B-所属行业", "I-所属行业", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O", "O"]}
请大佬指教

请问大佬，如果要改成多卡训练，使用DataParallel方法该如何修改代码呢？

原始数据如何处理成这样"spo_list": [{"h": {"name": "空调", "pos": [5, 7]}, "t": {"name": "制冷效果差", "pos": [7, 12]}, "relation": "部件故障"}]}

C:\Users\35845\Desktop\连铸\程序\模型构建\ner命名实体识别\BERT-BiLSTM-CRF\ner>python main.py
['O', 'B-故障设备', 'I-故障设备', 'B-故障原因', 'I-故障原因']
{'O': 0, 'B-故障设备': 1, 'I-故障设备': 2, 'B-故障原因': 3, 'I-故障原因': 4}
C:\Users\35845\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\modules\rnn.py:62: UserWarning: dropout option adds dropout after all but last recurrent layer, so non-zero dropout expects num_layers greater than 1, but got dropout=0.1 and num_layers=1
warnings.warn("dropout option adds dropout after all but last "
C:\Users\35845\AppData\Local\Programs\Python\Python38\lib\site-packages\transformers\optimization.py:429: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use the PyTorch implementation torch.optim.AdamW instead, or set no_deprecation_warning=True to disable this warning
warnings.warn(
Traceback (most recent call last):
File "main.py", line 189, in
main(data_name)
File "main.py", line 183, in main
report = train.test()
File "main.py", line 66, in test
self.model.load_state_dict(torch.load(os.path.join(self.output_dir, "pytorch_model_ner.bin"),map_location='cpu'))
File "C:\Users\35845\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\modules\module.py", line 1604, in load_state_dict
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for BertNer:
Unexpected key(s) in state_dict: "bert.embeddings.position_ids".
网上查了一圈还是不知道怎么解决，求大佬解惑

['O', 'B-故障设备', 'I-故障设备', 'B-故障原因', 'I-故障原因']
{'O': 0, 'B-故障设备': 1, 'I-故障设备': 2, 'B-故障原因': 3, 'I-故障原因': 4}
Some weights of the model checkpoint at ./model_hub/chinese-bert-wwm-ext/ were not used when initializing BertModel: ['cls.seq_relationship.bias', 'cls.predictions.transform.LayerNorm.bias', 'cls.seq_relationship.weight', 'cls.predictions.transform.dense.bias', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.transform.dense.weight', 'cls.predictions.decoder.weight', 'cls.predictions.bias']

• This IS expected if you are initializing BertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
• This IS NOT expected if you are initializing BertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
  F:\Program Files\Anaconda3\envs\pytorch\lib\site-packages\torch\nn\modules\rnn.py:62: UserWarning: dropout option adds dropout after all but last recurrent layer, so non-zero dropout expects num_layers greater than 1, but got dropout=0.1 and num_layers=1
  warnings.warn("dropout option adds dropout after all but last "
  Traceback (most recent call last):
  File "e:\硕士毕业\BERT-BLSTM-CRF\ner\main.py", line 189, in
  main(data_name)
  File "e:\硕士毕业\BERT-BLSTM-CRF\ner\main.py", line 159, in main
  model = BertNer(args)
  File "e:\硕士毕业\BERT-BLSTM-CRF\ner\model.py", line 24, in init
  self.crf = CRF(args.num_labels, batch_first=True)
  TypeError: init() got an unexpected keyword argument 'batch_first'
  大佬请问能看一下报错原因吗

test_loader=dev_loader,是故意这样设置还是设置错了？
直接改为test_loader=test_loader会影响运行吗？

大佬好，借助gpt，我将您的模型修改为处理英文数据的模型，
使用公开英文数据集，这个性能在80左右，使用W2NER模型性能在90左右，
一般论文里提到，使用该模型最高性能在94左右，是超参数设置问题，还是我修改原模型原因，
处理英文数据时，使用的预训练模型是bert-base-cased，跟预训练模型的选择也有关系嘛
希望大佬帮忙解答下，在此非常感谢！

这个项目的pytorch应该用什么版本呢

作者您好，我在运行main文件时，出现了如图错误，结合gpt解读之后，给出原因(见图2)为缺少参数target_names（说其列表为空），我大致添加修改参数后运行还是报错，麻烦作者指点一下

你好，第一个数据集已成功跑通，因自己数据是多类实体，故准备调试duie数据。在调试时，感觉是缺少数据集的，如下：self.train_file = self.data_path + "ori_data/duie_train.json"
self.dev_file = self.data_path + "ori_data/duie_dev.json"
self.test_file = self.data_path + "ori_data/duie_test2.json"
self.schema_file = self.data_path + "ori_data/duie_schema.json"
上述三个数据是不存在的嘛？
感谢大佬，或者我需要将自己数据处理成那种格式，在dgre的代码上，修改代码能实现多类实体识别，而不是两类实体识别。

目前项目已引入您的模型了，大佬，当前模型是否可以增量迭代训练,😄

Traceback (most recent call last):
File "D:/BERT/ner/main.py", line 189, in
main(data_name)
File "D:/BERT/ner/main.py", line 183, in main
report = train.test()
File "D:/BERT/ner/main.py", line 66, in test
self.model.load_state_dict(torch.load(os.path.join(self.output_dir, "pytorch_model_ner.bin")))
File "C:\Users\liu.conda\envs\bbc\lib\site-packages\torch\nn\modules\module.py", line 2153, in load_state_dict
raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for BertNer:
size mismatch for linear.weight: copying a param with shape torch.Size([5, 256]) from checkpoint, the shape in current model is torch.Size([13, 256]).
size mismatch for linear.bias: copying a param with shape torch.Size([5]) from checkpoint, the shape in current model is torch.Size([13]).
size mismatch for crf.start_transitions: copying a param with shape torch.Size([5]) from checkpoint, the shape in current model is torch.Size([13]).
size mismatch for crf.end_transitions: copying a param with shape torch.Size([5]) from checkpoint, the shape in current model is torch.Size([13]).
size mismatch for crf.transitions: copying a param with shape torch.Size([5, 5]) from checkpoint, the shape in current model is torch.Size([13, 13]).
labels文件加了几个标签之后出现的问题，估计就是喂的向量不是模型需要的，这块怎么改啊，还是要自己重新训练模型？求助大佬，找了好多设置文件没有能修改标签数量的地方。

Traceback (most recent call last):
File "D:\python\python-workspace\BERT-BILSTM-CRF\main.py", line 189, in
main(data_name)
File "D:\python\python-workspace\BERT-BILSTM-CRF\main.py", line 159, in main
model = BertNer(args)
File "D:\python\python-workspace\BERT-BILSTM-CRF\model.py", line 25, in init
self.crf = CRF(args.num_labels, batch_first=True)
TypeError: init() got an unexpected keyword argument 'batch_first'
大佬们，这个是什么问题呢？怎么解决啊

你好，我想请教一下，博主自带数据集dgre运行成功，但是再换成自己数据集时，自己的数据集格式应该调成什么样的格式呢？
以下是本人数据集中的一条实例：
{"id":65521,"text":"811：北京市发展和改革委员会发文字号：京发改〔2023〕293号公布日期：2023.03.03施行日期：2023.03.03时效性：现行有效效力位阶：地方规范性文件法规类别：能源综合规定节能管理北京市发展和改革委员会关于印发数据中心项目年可再生能源利用水平核实评价技术导则（试行）的通知（京发改〔2023〕293号）各有关单位：　　根据《关于进一步加强数据中心项目节能审查的若干规定》（京发改规〔2021〕4号）相关工作要求，为依据节能审查意见和节能报告做好取得节能审查批复的数据中心项目的年可再生能源利用水平核实评价工作，我们研究制定了《数据中心项目年可再生能源利用水平核实评价技术导则（试行）》，现予以印发。试行期间，如有问题和意见建议，请及时反馈。　　特此通知。北京市发展和改革委员会　　2023年3月3日　　附件：数据中心项目年可再生能源利用水平核实评价技术导则（试行）附件预览无相关内容","entities":[{"id":473,"label":"政策主体-政策制定者","start_offset":341,"end_offset":352},{"id":483,"label":"政策过程-政策评估","start_offset":261,"end_offset":267},{"id":5917,"label":"政策主体-政策制定者","start_offset":103,"end_offset":114}],"relations":[],"Comments":[]}。

想知道您代码的数据来源

['O', 'B-故障设备', 'I-故障设备', 'B-故障原因', 'I-故障原因']
{'O': 0, 'B-故障设备': 1, 'I-故障设备': 2, 'B-故障原因': 3, 'I-故障原因': 4}
Some weights of the model checkpoint at ./model_hub/chinese-bert-wwm-ext/ were not used when initializing BertModel: ['cls.predictions.decoder.weight', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.bias', 'cls.seq_relationship.bias', 'cls.predictions.transform.dense.bias', 'cls.predictions.transform.dense.weight', 'cls.seq_relationship.weight', 'cls.predictions.transform.LayerNorm.bias']

• This IS expected if you are initializing BertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
• This IS NOT expected if you are initializing BertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
  C:\Users\DELL\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\modules\rnn.py:62: UserWarning: dropout option adds dropout after all but last recurrent layer, so non-zero dropout expects num_layers greater than 1, but got dropout=0.1 and num_layers=1
  warnings.warn("dropout option adds dropout after all but last "
  C:\Users\DELL\AppData\Local\Programs\Python\Python38\lib\site-packages\transformers\optimization.py:391: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use the PyTorch implementation torch.optim.AdamW instead, or set no_deprecation_warning=True to disable this warning
  warnings.warn(
  Traceback (most recent call last):
  File "C:/Users/DELL/Desktop/BERT-BILSTM-CRF-main/main.py", line 190, in
  main(data_name)
  File "C:/Users/DELL/Desktop/BERT-BILSTM-CRF-main/main.py", line 182, in main
  train.train()
  File "C:/Users/DELL/Desktop/BERT-BILSTM-CRF-main/main.py", line 52, in train
  output = self.model(input_ids, attention_mask, labels)
  File "C:\Users\DELL\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\modules\module.py", line 1110, in _call_impl
  return forward_call(*input, **kwargs)
  File "C:\Users\DELL\Desktop\BERT-BILSTM-CRF-main\model.py", line 36, in forward
  loss = -self.crf(seq_out, labels, mask=attention_mask.bool(), reduction='mean')
  File "C:\Users\DELL\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\modules\module.py", line 1110, in call_impl
  return forward_call(*input, **kwargs)
  File "C:\Users\DELL\AppData\Local\Programs\Python\Python38\lib\site-packages\TorchCRF_init.py", line 102, in forward
  numerator = self.compute_score(emissions, tags, mask)
  File "C:\Users\DELL\AppData\Local\Programs\Python\Python38\lib\site-packages\TorchCRF_init.py", line 186, in _compute_score
  score = self.start_transitions[tags[0]]
  IndexError: tensors used as indices must be long, byte or bool tensors

Process finished with exit code 1

请问我可以将您的代码更改后重新发布在GitHub吗？因为注意到您没有选择开源协议，版权方面有些顾虑。

Traceback (most recent call last):
File "e:\git clone download\BERT-BILSTM-CRF\main.py", line 190, in
main(data_name)
File "e:\git clone download\BERT-BILSTM-CRF\main.py", line 182, in main
train.train()
File "e:\git clone download\BERT-BILSTM-CRF\main.py", line 52, in train
output = self.model(input_ids, attention_mask, labels)
File "E:\miniconda3\envs\cyclegan\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "e:\git clone download\BERT-BILSTM-CRF\model.py", line 27, in forward
bert_output = self.bert(input_ids=input_ids, attention_mask=attention_mask)
File "E:\miniconda3\envs\cyclegan\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "E:\miniconda3\envs\cyclegan\lib\site-packages\transformers\models\bert\modeling_bert.py", line 1013, in forward
encoder_outputs = self.encoder(
File "E:\miniconda3\envs\cyclegan\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "E:\miniconda3\envs\cyclegan\lib\site-packages\transformers\models\bert\modeling_bert.py", line 607, in forward
layer_outputs = layer_module(
File "E:\miniconda3\envs\cyclegan\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "E:\miniconda3\envs\cyclegan\lib\site-packages\transformers\models\bert\modeling_bert.py", line 497, in forward
self_attention_outputs = self.attention(
File "E:\miniconda3\envs\cyclegan\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "E:\miniconda3\envs\cyclegan\lib\site-packages\transformers\models\bert\modeling_bert.py", line 427, in forward
self_outputs = self.self(
File "E:\miniconda3\envs\cyclegan\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "E:\miniconda3\envs\cyclegan\lib\site-packages\transformers\models\bert\modeling_bert.py", line 286, in forward
mixed_query_layer = self.query(hidden_states)
File "E:\miniconda3\envs\cyclegan\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "E:\miniconda3\envs\cyclegan\lib\site-packages\torch\nn\modules\linear.py", line 103, in forward
return F.linear(input, self.weight, self.bias)
File "E:\miniconda3\envs\cyclegan\lib\site-packages\torch\nn\functional.py", line 1848, in linear
return torch._C._nn.linear(input, weight, bias)
RuntimeError: CUDA error: CUBLAS_STATUS_NOT_INITIALIZED when calling cublasCreate(handle)
请问这个问题该怎么解决呢？

大佬，我把device的使用语句改成如图的mps，但是训练的速度极慢，感觉甚至比cpu慢十倍，GPU的占用也很低，但是内存占用挺多，9G左右。我测试了其他的mps使用场景是正常的，比cpu速度快好几倍。想请教下是否是我的设置问题。

如何手动停止训练