stanleylsx / entity_extractor_by_ner

基于Tensorflow2.3开发的NER模型,都是CRF范式,包含Bilstm(IDCNN)-CRF、Bert-Bilstm(IDCNN)-CRF、Bert-CRF,可微调预训练模型,可对抗学习,用于命名实体识别,配置后可直接运行。
390 stars 73 forks source link

start evaluate engines...之后报错 #73

Open M-YoungBWeN opened 2 months ago

M-YoungBWeN commented 2 months ago

您好,我遇到了一些问题

  1. 运行到start evaluate engines...之后报错 100%|██████████| 5796/5796 [1:35:50<00:00, 1.01it/s] start evaluate engines... 0%| | 0/1159 [00:00<?, ?it/s]WARNING:tensorflow:The parameters output_attentions, output_hidden_states and use_cache cannot be updated when calling a model.They have to be set to True/False in the config object (i.e.: config=XConfig.from_pretrained('name', output_attentions=True)). WARNING:tensorflow:The parameter return_dict cannot be set in graph mode and will always be set to True. 2024-06-16 20:29:07.033624: W tensorflow/core/grappler/optimizers/loop_optimizer.cc:907] Skipping loop optimization for Merge node with control input: cond/branch_executed/_10 进程已结束,退出代码为 -1073740791 (0xC0000409)

  2. checkpoint文件夹里没有输出东西 .# 在此处设置模型保存位置,代码支持在原始模型上继续训练,新数据或从头训练一定要改! checkpoints_dir=checkpoints/0616 .# 模型的名字 checkpoint_name=model0616

  3. 我把huggingface的东西下载到了本地,不知道这样直接写路径可不可以 use_pretrained_model=True pretrained_model=Bert .# Bert/AlBert huggingface_tag=E:\1_Code\Transformer\BERT_Data\bert230\bert-base-chinese .# huggingface上模型的名字 finetune=True

  4. 显存最多只能使用70%,不知道如何更改这个限制

M-YoungBWeN commented 2 months ago

问题1、2、3已经解决,问题4不知道咋弄

stanleylsx commented 1 week ago

问题1、2、3已经解决,问题4不知道咋弄

看看能不能再改大一点batch_size?