diff --git "a/dNFQT4oBgHgl3EQfjDY0/content/tmp_files/load_file.txt" "b/dNFQT4oBgHgl3EQfjDY0/content/tmp_files/load_file.txt"
new file mode 100644--- /dev/null
+++ "b/dNFQT4oBgHgl3EQfjDY0/content/tmp_files/load_file.txt"
@@ -0,0 +1,1583 @@
+filepath=/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf,len=1582
+page_content='Sentence Identification with BOS and EOS Label Combinations  Takuma Udagawa, Hiroshi Kanayama, Issei Yoshida  IBM Research - Tokyo, Japan  Takuma.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='Udagawa@ibm.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='com, {hkana,issei}@jp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='ibm.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='com  Abstract  The sentence is a fundamental unit in many  NLP applications.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Sentence segmentation is  widely used as the first preprocessing task,  where an input text is split into consecutive  sentences considering the end of the sentence  (EOS) as their boundaries.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This task formula-  tion relies on a strong assumption that the in-  put text consists only of sentences, or what we  call the sentential units (SUs).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' However, real-  world texts often contain non-sentential units  (NSUs) such as metadata, sentence fragments,  nonlinguistic markers, etc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' which are unrea-  sonable or undesirable to be treated as a part  of an SU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To tackle this issue, we formulate a  novel task of sentence identification, where the  goal is to identify SUs while excluding NSUs  in a given text.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To conduct sentence iden-  tification, we propose a simple yet effective  method which combines the beginning of the  sentence (BOS) and EOS labels to determine  the most probable SUs and NSUs based on  dynamic programming.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To evaluate this task,  we design an automatic, language-independent  procedure to convert the Universal Dependen-  cies corpora into sentence identification bench-  marks.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Finally, our experiments on the sen-  tence identification task demonstrate that our  proposed method generally outperforms sen-  tence segmentation baselines which only uti-  lize EOS labels.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  1  Introduction  The sentence, which we refer to as the sentential  unit (SU), is a fundamental unit of processing in  many NLP applications including syntactic pars-  ing (Dozat and Manning, 2017), semantic parsing  (Dozat and Manning, 2018), and machine transla-  tion (Liu et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2020).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Existing works mostly rely  on sentence segmentation (a.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='k.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='a.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' sentence bound-  ary detection) as the first preprocessing task, where  we predict the end of the sentence (EOS) to split a  text into consecutive SUs (Kiss and Strunk, 2006;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Gillick, 2009).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This approach relies on a strong  assumption that the text only consists of SUs;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' how-  ever, real-world texts like web contents often con-  tain non-sentential units (NSUs) such as the meta-  data of attachments embedded in the email body,  repetition of symbols for separating texts, irregular  series of nouns, etc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' (just to name a few).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Such  NSUs may cause detrimental or unexpected results  in the downstream tasks if considered as parts of  the SUs and are more desirable to be distinguished  from SUs in the first preprocessing step.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To tackle this problem, we formulate a novel  task of sentence identification, where the goal is  to identify SUs while excluding NSUs in a given  text (§3).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This can be regarded as an SU span ex-  traction task, where each SU span is represented  by the beginning of the sentence (BOS) and the  EOS labels.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1 We illustrate the difference between  sentence segmentation and sentence identification  in Table 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In sentence segmentation, the text frag-  ment of an embedded file (“- TEXT.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='htm << File:  TEXT.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='htm >>”) needs to be considered as a part  of an SU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In contrast, sentence identification can  regard it as an NSU and exclude it for downstream  applications such as dependency parsing.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To conduct sentence identification, we propose  a simple method which effectively combines the  BOS and EOS probabilities to determine both SUs  and NSUs (§4).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To be specific, we first train  the BOS and EOS labeling models based on ei-  ther the sentence identification dataset (with SUs  and NSUs) or sentence segmentation dataset (only  SUs).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Then, we search for the most probable spans  of SUs and NSUs using a simple dynamic program-  ming framework.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Theoretically, our method can be  considered as a natural generalization of existing  sentence segmentation algorithms.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To evaluate this task, we design an automatic pro-  1For simplicity, we assume that the input text can be seg-  mented into consecutive, non-overlapping units of SUs and  NSUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This way, we can also represent and evaluate SU  extraction as an equivalent BIO labeling task (§5-§7).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  arXiv:2301.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='13352v1  [cs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='CL]  31 Jan 2023  Input Text  Thank you.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' - TEXT.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='htm << File: TEXT.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='htm >> I was thinking of converting it to a hover  (from EWT)  vehicle.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' I might just sell the car and get you to drive me around all winter.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  E  Sentence  Thank you.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' - TEXT.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='htm << File: TEXT.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='htm >> I was thinking of converting it to a hover  Segmentation  E  E  vehicle.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' I might just sell the car and get you to drive me around all winter.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  B  E  B  Sentence  Thank you.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' - TEXT.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='htm << File: TEXT.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='htm >> I was thinking of converting it to a hover  Identification  E B  E  vehicle.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' I might just sell the car and get you to drive me around all winter.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Table 1: Illustration of sentence segmentation and sentence identification.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In sentence segmentation, EOS labels  (E) are used to segment the input text into consecutive SUs (in blue).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In sentence identification, only the spans  bracketed by the BOS (B) and EOS labels are extracted as SUs, while the rest can be excluded as NSUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  cedure to convert the Universal Dependencies (UD)  corpora (de Marneffe et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2021) into sentence  identification benchmarks (§5).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To be specific, (i)  we use the original sentence boundaries in UD as  the unit (SU and NSU) boundaries and (ii) classify  each unit as an SU iff it contains at least one clausal  predicate with a core/non-core argument.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Impor-  tantly, our classification rule follows the definition  of lexical sentence in linguistics (Nunberg, 1990),  is easily customizable with language-independent  rules, and makes reasonable classification within  the scope of our experiments.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To conduct our experiments, we focus on the  English Web Treebank (Silveira et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2014) as the  primary benchmark for sentence identification and  train the BOS/EOS labeling models by finetuning  RoBERTa (Liu et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2019) (§6).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We also propose  techniques to develop these models using a stan-  dard sentence segmentation dataset, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' the Wall  Street Journal corpus (Marcus et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 1993), which  only contains clean, edited SUs without any NSUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Based on our experimental results, we demon-  strate that our proposed method generally outper-  forms sentence segmentation baselines which only  utilize EOS labels (§7).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' These results highlight  the importance of combining the BOS labels in  addition to the EOS labels for accurate sentence  identification under various conditions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  2  Background  Sentence segmentation, a.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='k.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='a.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' sentence boundary  detection, is the task of segmenting an input text  into the unit of sentences.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Despite the long his-  tory of study (Riley, 1989) and its importance in  the entire NLP pipeline (Walker et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2001), this  area has received relatively little attention.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' For  one reason, the task has been recognized as “long  solved” (Read et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2012) with the most recent  approach reporting 99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8% F1 score on the standard  English Wall Street Journal (WSJ) dataset (Wicks  and Post, 2021).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Their state-of-the-art method ER-  SATZ combines (i) a regular-expression based de-  tector of candidate sentence boundaries, followed  by (ii) a Transformer-based (Vaswani et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2017)  binary classifier which predicts whether the can-  didate boundary is EOS based on the local con-  text, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' surrounding few words.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This modern  context-based approach has been shown to outper-  form competitive, widely used baselines such as  SPLITTA (Gillick, 2009), PUNKT (Kiss and Strunk,  2006), and MOSES (Koehn et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2007).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  However, two important aspects are not fully ad-  dressed in the current literature.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' First is the cover-  age of diverse domains, genres, and writing styles.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Existing works (including Wicks and Post, 2021)  focus on formal/edited text and assume the exis-  tence of sentence ending punctuations (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' full  stops) at the sentence boundaries.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' However, social  media texts often lack such punctuations and con-  tain various types of non-linguistic noise, which  can lead to a substantial degradation in the seg-  mentation performance (Read et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2012;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Rudra-  pal et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2015).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Speech transcription texts also  usually contain disfluent, ungrammatical, or frag-  mented structures and lack both punctuations and  casing (Wang et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2019;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Rehbein et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2020).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Considering the amount of such informal or non-  standard texts in the real world, it is compelling  to expand the capability of sentence segmentation  beyond formal, standardized text.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  The second aspect is the coverage of multiple  languages.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Different languages involve different  complexities in sentence segmentation, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Chi-  nese requires the disambiguation of commas as the  sentence ending punctuation (Xue and Yang, 2011)  and Thai does not mark EOS with any type of punc-  tations (Aroonmanakun et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2007;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Zhou et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=',  2016).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To advance NLP from a multilingual per-  spective, it is crucial to develop and evaluate mod-  els in multiple languages: Wicks and Post (2021)  make an important step in this direction, proposing  a language-agnostic, unified sentence segmentation  model covering a total of 87 languages.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Based on these observations, we first propose to  extend the task of sentence segmentation to sen-  tence identification, which expands the capability  of sentence segmentation beyond formal, standard-  ized text (§3, §4).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Secondly, we propose a cross-  lingual method of benchmarking sentence identifi-  cation based on the UD corpora, considering every  word or character as the candidate boundary to  cover diverse domains, genres, and languages that  lack sentence ending punctuations (§5).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Finally,  we follow Wicks and Post (2021) to develop mod-  ern neural-based models that require no language-  specific engineering and can be developed for dif-  ferent languages in a unified manner (§6).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  3  Task Formulation  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  Sentence Segmentation Task  First, we introduce a precise (re-)formulation  of the sentence segmentation task.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Let W =  (w0, w1, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', wN−1) represent the input text, where  each wi denotes a word (but can also be a sub-  word or character).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We also define the text span  W [i : j] = (wi, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', wj−1), their concatenation  W [i : j] ⊕ W [j : k] = W [i : k], and SU bound-  ary indices B = (b0, b1, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', bM) where b0 = 0,  bM = N, and �M  i=1 W [bi−1 : bi] = W (i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' the  concatenation of all SUs recovers the input text).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Next, we introduce the SU probability pSU(W [i:  j]) which corresponds to the probability of the text  span W [i:j] being an SU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Based on this probabil-  ity, the task of sentence segmentation can be for-  malized as searching for the boundaries B which  maximize the following probability:2  arg max  B  M  �  i=1  pSU(W [bi−1 :bi])  (1)  The most standard approach is to define pSU(W [i:  j]) based on a pretrained EOS labeling model, as  we describe in §4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' However, our (re-)formulation  2M is a variable and need not be fixed during the search.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  as Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' (1) is more general and permits other defini-  tions of SU probability as well.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  Sentence Identification Task  In sentence identification, we consider the input  text W can be segmented into consecutive, non-  overlapping units of SUs and NSUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Hence, we  regard B = (b0, b1, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', bM) as the unit (SU and  NSU) boundaries and define the unit indicators  A = (a1, a2, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', aM) for each unit as follows:  ai =  �  1  if W [bi−1 :bi] is an SU  0  if W [bi−1 :bi] is an NSU  Next,  we  introduce  the  NSU  probability  pNSU(W [i : j]) which corresponds to the prob-  ability of the text span W [i : j] being an NSU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Based on pSU and pNSU, we can formalize the  task of sentence identification as searching for the  unit boundaries B and unit indicators A which  maximize the following probability:  arg max  B,A  M  �  i=1  pSU(W [bi−1 :bi])ai pNSU(W [bi−1 :bi])1−ai  (2)  Note that this strictly generalizes the sentence seg-  mentation task in Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' (1), which is a special case  where ai = 1, ∀ai ∈ A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Based on this task formu-  lation, we discuss how we can define pSU(W [i:j])  and pNSU(W [i:j]) to derive our sentence identifi-  cation algorithm in §4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  4  Methods  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  Sentence Segmentation Method  In the most standard approach, sentence segmenta-  tion employs an EOS labeling model pEOS to define  the SU probability pSU in Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' (1).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To be specific,  let pEOS(wi|W ;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' θ) denote the EOS labeling model,  which computes the probability of wi being EOS  in W (θ denotes the model parameters).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Typically,  it is straightforward to train this model in a super-  vised learning setup using a dataset annotated with  gold EOS boundaries (Wicks and Post, 2021).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' For  brevity, we use the notation pEOS(wi) as a short-  hand for pEOS(wi|W ;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' θ), i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' we omit W and θ  (unless required) in the rest of this paper.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Based on the pretrained model pEOS, we can  define the SU probability as pSU(W [i : j]) =  pEOS(wj−1) �  i≤k<j−1(1 − pEOS(wk)), which re-  quires the last word wj−1 to be EOS and all other  words to be non-EOS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' By substituting this defini-  tion, we can decompose Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' (1) as follows:  (1) = arg max  B  M  �  i=1  log pSU(W [bi−1 :bi])  = arg max  B  M  �  i=1  �  log pEOS(wbi−1) +  �  bi−1≤j<bi−1  log (1 − pEOS(wj))  �  = arg max  B  �  i∈BEOS  log pEOS(wi) +  �  i/∈BEOS  log (1 − pEOS(wi))  (3)  where BEOS = {bi − 1 | i ∈ (1, 2, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', M)} repre-  sents all the EOS indices defined by B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  This is a trivial optimization problem where we  can simply choose BEOS = {i ∈ (0, 1, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', N −  1) | pEOS(wi) ≥ 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5} to maximize Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' (3).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This  also shows that sentence segmentation can be con-  ducted by predicting the EOS independently for  each wi based on pEOS(wi).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In contrast, sentence  identification involves a more complex optimiza-  tion problem which we solve using dynamic pro-  gramming (§4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  Sentence Identification Method  We extend the method of sentence segmentation  (§4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1) to conduct sentence identification.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To be spe-  cific, we employ pretrained BOS and EOS labeling  models pBOS, pEOS to define the SU and NSU prob-  abilities pSU, pNSU in Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' (2).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' As a first step, we  need to train the BOS and EOS labeling models:  this can be conducted in a supervised manner using  a dataset containing gold BOS and EOS labels, as  we explain in §6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Based on the pretrained BOS and EOS labeling  models, we can define the SU and NSU probabili-  ties as follows:  pSU(W [i:j]) = pBOS(wi)  �  i<k≤j−1  (1 − pBOS(wk))  × pEOS(wj−1)  �  i≤k<j−1  (1 − pEOS(wk))  pNSU(W [i:j]) =  �  i≤k≤j−1  (1 − pBOS(wk)) ×  �  i≤k≤j−1  (1 − pEOS(wk))  In the SU probability pSU, the first word wi is  required to be BOS, the last word wj−1 to be EOS,  and all other words to be neither BOS nor EOS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Note that this definition of pSU is a natural gener-  alization from §4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1 which only relies on the EOS  probability pEOS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  In contrast, the NSU probability pNSU requires  all words to be neither BOS nor EOS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Notably, this  definition does not distinguish contiguous NSUs in  the sense that pNSU(W [i:k]) = pNSU(W [i:j]) ×  pNSU(W [j :k]) if W [i:j] ⊕ W [j :k] = W [i:k].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  This is convenient as we are only interested in the  extraction of SUs and do not need to seek the exact  boundaries between consecutive NSUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  By substituting these definitions of pSU and pNSU,  we can decompose Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' (2) as follows:  (2) = arg max  B,A  M  �  i=1  �  ai log pSU(W [bi−1 :bi])  + (1 − ai) log pNSU(W [bi−1 :bi])  �  = arg max  B,A  �  i∈BA  BOS  log pBOS(wi) +  �  i/∈BA  BOS  log (1 − pBOS(wi))  +  �  i∈BA  EOS  log pEOS(wi) +  �  i/∈BA  EOS  log (1 − pEOS(wi))  (4)  where BA  BOS = {bi−1 | i ∈ (1, 2, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', M), ai = 1}  denotes the BOS indices and BA  EOS = {bi − 1 | i∈  (1, 2, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', M), ai = 1} denotes the EOS indices,  both defined by B and A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Therefore, our goal is to choose BA  BOS and BA  EOS  which maximize Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' (4).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To this end, we need  to consider the restrictions that (i) the first label  should be BOS, (ii) the last label should be EOS,  and (iii) BOS and EOS labels need to appear alter-  nately.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' These restrictions can be incorporated in  our dynamic programming framework to find the  argmax of Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' (4).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' For the precise algorithm, we  refer the readers to Appendix A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  5  Evaluation  Due to the novelty of the task, currently there exists  no benchmark for evaluating sentence identifica-  tion.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To address this issue, we propose a fully  automatic procedure to convert the Universal De-  pendencies (UD) corpora (de Marneffe et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2021)  into sentence identification benchmarks.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Concretely speaking, we conduct the following  two steps based on the gold UD annotation: (i) the  detection of unit (SU and NSU) boundaries and (ii)  the classification of each unit into SU or NSU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' As  for (i), we simply use the original sentence bound-  aries in the UD annotation, where UD uses the term  sentence in a broader sense including both SUs and  NSUs (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' sentence fragments).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Note that the ex-  act boundaries between consecutive NSUs (which  we call NSU–NSU boundaries) do not need to be  accurate or consistent, since we are only interested  in extracting the spans of SUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' However, we do  expect that the original boundaries are generally  reliable in all other cases (SU–SU and SU–NSU  boundaries), which seems to be the case.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  The main problem is (ii), i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' how to classify  ∼*∼*∼*∼*∼*∼*∼*∼*∼*  **********NOTE**********  By video conference from _______  Excerpt:  02/13/2001 08:02 PM  5:00 PT ** 6:00 MT ** 7:00 CT ** 8:00 ET  Time: 11:30am / 1:30pm Central / 2:30pm Eastern  Sunshine Coast, British Columbia, Canada  UnleadedStocks.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='pdf  t r u t h o u t — Perspective  ( Answered, 2 Comments )  The federal sites of Washington, DC.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  From Madrid to Seville to Barcelona an Valencia.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Table 2: Examples of gold NSUs in the English Web  Treebank (EWT) identified based on our procedure.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Each line corresponds to one example of NSU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  each unit as an SU or NSU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To this end, we fol-  low the notion of lexical sentence in linguistics  (Nunberg, 1990) which defines an SU based on the  dependencies among the lexical items, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' a group  of words that contain a subject and predicate.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In  this work, we build upon the UD dependency re-  lations and define an SU as a unit that contains at  least one clausal predicate with a core or non-core  argument.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3 Here, a clause expresses an event or  proposition which we regard as an essential aspect  of SUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' A clausal predicate and a core argument  form the backbone of a clause, while a non-core  argument modifies it (de Marneffe et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2021).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Note that our current definition excludes noun  phrases appearing by themselves, since they only  consist of the nominal dependent relations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' How-  ever, we can flexibly customize the definition of  SUs to include or exclude such phrases.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Due to the reliance on UD, our conversion proce-  dure can be applied to a wide variety of languages  supported in UD (currently over 100 languages).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  However, as a first set of experiments, we focus on  the English Web Treebank (EWT) (Silveira et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=',  2014) as the primary benchmark of sentence identi-  fication.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This dataset comprises five genres of web  media texts: namely weblogs, newsgroup threads,  emails, product reviews, and Q&A websites.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Con-  sequently, the dataset contains formal SUs, infor-  mal SUs (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' without capitalization or punctua-  tions) as well as a wide variety of NSUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  We show some examples of NSUs in Table 2  (and more in Appendix B) identified based on our  3To check this condition, we simply need to verify whether  there is at least one core argument (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' nsubj, obj, ccomp) or  non-core dependent (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' obl, advcl, aux).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' For a full list of the  UD relations, see https://universaldependencies.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='org/u/dep/.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Train  Dev  Test  Total SUs  10,356  1,523  1,490  Total NSUs  2,187  478  587  Word-  Level  B-Label  10,356  1,523  1,490  I-Label  160,127  18,791  18,222  O-Label  6,939  1,302  1,822  Character-  Level  B-Label  10,356  1,523  1,490  I-Label  773,223  92,309  88,441  O-Label  47,107  9,925  13,232  Table 3: EWT dataset statistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  procedure.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' As shown by the results, our procedure  can identify various NSUs including nonlinguistic  markers, timestamps, lists, contact information, etc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  We can also see that noun/prepositional phrases  are classified as NSUs based on our criteria.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' By  excluding such NSUs and identifying SUs, we can  clearly separate the portions of the text that are  worth sophisticated linguistic analyses, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' based  on dependency parsing or manual inspection.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Finally, we summarize the dataset statistics of  EWT in Table 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Overall, 17∼28% of the units  were classified as NSUs, with the test set containing  the highest proportion of NSUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We also regard SU  extraction as a word-level or character-level BIO  labeling task and report the number of gold BIO  labels in Table 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4 At the word-level, we can see  that the proportion of O-labels (indicating NSUs) is  only 4∼8% and much smaller than the proportion  of NSUs in terms of units: this is because NSUs  are usually short and contain only a few words.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  At the character-level, the proportion of O-labels  is slightly larger (6∼13%): this is because NSUs  often contain extraordinarily long words like URLs  and long sequences of nonlinguistic symbols.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Overall, we could verify that there exists a non-  negligible amount of NSUs in the EWT dataset,  which we aim to exclude with sentence identifica-  tion in our experiments.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  6  Experimental Setup  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  Model Setup  As we discussed in §4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2, our sentence identification  method requires pretrained BOS and EOS labeling  models to identify SUs and NSUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To develop  these models, we simply finetune RoBERTaBASE  4B = Beginning of SU, I = Inside of SU, and O = Outside  of SU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Details of how we assign the gold BIO labels (at the  word-level and character-level) are provided in Appendix C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  by adding a binary BOS/EOS classifier on top of  the encoder.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To enable our models to handle various lengths  of the input texts, we concatenate the consecutive  L units of gold SUs and NSUs as the input during  training, where L is sampled from a geometric  distribution with parameter pCC.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5 However, the  RoBERTa encoder has the restriction that the input  text size cannot exceed 512 subwords.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Therefore, if  the input text size is too large, we replace L with the  maximum L′ < L which satisfies this restriction.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Note that this is a common procedure to sample  variable (instead of fixed) lengths of concatenated  units (Joshi et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', 2020).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Assuming the existence of the in-domain sen-  tence identification dataset (EWT Train/Dev), it  is straightforward to train the BOS/EOS labeling  models based on our unit concatenation procedure.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  However, we may not always have the gold anno-  tation of SUs and NSUs for the target domain.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To  take such cases into account, we also consider a  setup where we only have the standard sentence  segmentation dataset (WSJ Train/Dev) to train the  BOS/EOS labeling models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  When using the sentence segmentation dataset  (WSJ), we need to apply the unit concatenation  procedure using only clean, edited SUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Unfor-  tunately, this can yield the following data priors  which do not actually hold in a sentence identifi-  cation dataset (EWT): (i) an SU (almost) always  starts with a capitalization and ends with punctua-  tion, (ii) the first word of the input is always BOS  and the last word is always EOS, and (iii) BOS  always directly follows EOS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To address (i) and (ii), we propose a simple data  augmentation technique to alleviate the discrepancy  in the data priors.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To address (iii), we propose  an ensembling technique with the unidirectional  (instead of bidirectional) models which are agnostic  to this data prior.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  Data Augmentation (+AUG)  To address (i), we conduct a unit-level data aug-  mentation, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' we modify each unit based on the  following rules with a small probability pDA:  Convert all words in the unit to lower-case,  upper-case, or title-case (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' “hello world”,  5With parameter pCC ∈ (0, 1], the probability mass func-  tion of the geometric distribution is p(L = l) = (1 −  pCC)l−1pCC where l ∈ {1, 2, 3, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='}.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' As pCC decreases,  the distribution gets more skewed towards larger L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' With  pCC = 0, we consider p(L = ∞) = 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Orig.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  B  E B  Joe went to school.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' After that he .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  (i) Unit  B  E  B  Aug.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Joe went to school  AFTER THAT HE .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  (ii) Unit  B  E  B  Trunc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Joe went to school  AFTER THAT HE .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Table 4: Illustration of our data augmentation tech-  nique.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  In (i) unit-level augmentation, we randomly  change the casing or remove the last punctuations of  each unit.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In (ii) unit truncation, we randomly truncate  the first and last units of the input (and regard them as  NSUs).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  “HELLO WORLD”, or “Hello World”).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Remove sentence ending punctuations based  on a regular-expression matcher (following  ERSATZ, Wicks and Post, 2021).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  After the unit-level augmentation, we can apply the  unit concatenation in the exact same manner.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Finally, to address (ii), we randomly apply a unit  truncation to the first and last units of the concate-  nated input.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To be specific, we choose a random  word in the first (last) unit and remove all words  prior (posterior) to it with a small probability pT R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  If the truncation is conducted, we regard the unit as  an NSU and fix the gold BOS/EOS labels accord-  ingly.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' See Table 4 for an illustration.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Based on this procedure, we can expect to allevi-  ate the data priors (i) and (ii).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' For more details, we  refer the readers to Appendix D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  Unidirectional Model (+UNI)  Simply concatenating SUs (without NSUs) yields  the data prior (iii), i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' BOS always directly fol-  lows EOS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This prior can be easily captured by the  bidirectional models pBOS(wi|W ), pEOS(wi|W )  conditioned on the whole input W , including our  RoBERTa-based models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' For instance, as shown in  Figure 1, the model may predict EOS at the end of  the first unit (w2 = #) just because the next word  (w3 = This) is likely predicted as BOS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To alleviate this issue, we propose to combine  the predictions of the unidirectional models for  BOS and EOS labeling.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To be precise, let W ≤i =  (w0, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', wi) and W ≥i = (wi, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', wN−1).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Then,  we can represent the unidirectional BOS model as  pUni  BOS(wi|W ≥i) (looking the context right-to-left)  and EOS model as pUni  EOS(wi|W ≤i) (looking left-to-  right).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' As illustrated in Figure 1, these models are  agnostic to the data prior (iii).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In practice, we can  Figure 1: Illustration of the bidirectional EOS model  (left) and the unidirectional EOS model (right).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  simply use different attention masks and share the  encoder parameters (except the last classifier) for  the unidirectional and bidirectional models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  We can utilize these unidirectional models by  taking a linear intepolation with the bidirectional  models as follows:  p+Uni  BOS (wi|W ) = λ · pUni  BOS(wi|W ≥i) + (1−λ) · pBOS(wi|W )  p+Uni  EOS (wi|W ) = λ · pUni  EOS(wi|W ≤i) + (1−λ) · pEOS(wi|W )  Then, we can use p+Uni  BOS and p+Uni  EOS in place of  pBOS and pEOS (respectively) to conduct sentence  identification, as described in §4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Finally, we compare our proposed methods  against sentence segmentation baselines which only  utilize EOS labels.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6 As for the baselines, we use  the EOS labeling model developed in the same  manner to segment the input text based on EOS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Note that we can optionally force the last word in  the input to be EOS: in this case, the result will  only contain SUs since all segments will end with  EOS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' By default, we do not force the last EOS: in  this case, the segment after the last EOS (if exists)  is considered as an NSU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  As a default configuration, we use pCC = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5,  pDA = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3, pT R = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1, and λ = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5 in our ex-  periments.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To ensure reproducibility, we report  more details on the hyperparameters and model  setup in Appendix D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' For the precise procedure on  how we convert between the word-, character-, and  subword-level labels (for RoBERTa), we refer the  readers to Appendix C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  Evaluation Setup  In the evaluation phase, we consider three ways of  assembling the input texts on which we conduct  sentence identification.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Firstly, we can apply the  same unit concatenation procedure as described in  6This EOS-only method is the most reasonable baseline to  quantify the precise advantage from combining BOS labels in  addition to EOS, which is proposed in our methods.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  §6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To be specific, we use pCC =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5 (same as the  training phase) and pCC = 0 (which concatenates  the units up to the maximal length) to simulate both  shorter and longer lengths of the input texts.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  However, this approach is relatively synthetic in  the sense that we take the gold unit boundaries for  granted.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' They are usually unavailable at the infer-  ence time, so we should consider a more realistic  setting for evaluating the methods without relying  on the gold unit boundaries.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To this end, we propose to evaluate sentence  identification as a postprocessing of sentence seg-  mentation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To be specific, we first apply the state-  of-the-art method ERSATZ (Wicks and Post, 2021)  on the raw text of EWT and then apply sentence  identification to each segmented text.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Note that ER-  SATZ has high precision but still predicts false EOS  which can fragment a gold SU: in such cases, we  consider the fragmented SUs as NSUs and fix the  labels accordingly (just as we did in unit truncation,  cf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' §6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1 and Table 4).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  As for the evaluation metrics, we convert the  predictions of our methods into word/character-  level BIO labels (cf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Appendix C) and compute  the F1 score for each label prediction.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Then, we  summarize the results as the macro average F1  and weighted average F1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We also compute the  F1 score of the exact SU span extraction at the  word/character-level.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Finally, we run each exper-  iment (from model training to testing) five times  with different random seeds and report the average  and standard deviation as the final results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  7  Results  Table 5 summarizes the word-level evaluation re-  sults.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' The results for the character-level evaluation  show similar tendencies, so we put them in Ap-  pendix E.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' The F1 score for each BIO label predic-  tion is also available in Appendix E.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Firstly, we take a look at the results when we  have the in-domain sentence identification dataset  (EWT Train/Dev) for model development.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In this  setup, we can verify that our proposed method  (BOS&EOS) significantly outperforms the base-  lines (EOS-Only) in all metrics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' For instance, our  method achieves consistently high performance of  84∼89% F1 for the exact SU span extraction, both  at the word- and character-level.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This is a very  promising result that demonstrates the effective-  ness of our method when we can leverage the gold  SUs and NSUs from the target domain.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Peos (W2/W) = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='65  (W2/W≤2) = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='08  Unidirectional Model  Bidirectional Model  (Left-to-Right)  This  #  This  #  #  #  is  #  #  is  wo  W1  W2  wo  W1  W2  W3  W4  W3  W4Train/Dev  Datasets  Model  EWT Test (pCC = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5)  EWT Test (pCC = 0)  EWT Test (Postprocess)  BIO  BIO  Span  BIO  BIO  Span  BIO  BIO  Span  Macro  Weighted  Macro  Weighted  Macro  Weighted  EWT  Train/Dev  EOS-Only  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  58.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  81.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  EOS-Only (force last)  58.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  60.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  62.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  BOS&EOS  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  91.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  96.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  84.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  92.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  96.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  88.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  WSJ  Train/Dev  EOS-Only  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  88.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  48.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  67.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  EOS-Only (force last)  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  53.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  55.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  48.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  58.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  61.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  EOS-Only (+AUG)  66.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  88.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  58.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  54.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  88.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  66.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  BOS&EOS  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  89.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  48.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  68.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  BOS&EOS (+UNI)  70.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  88.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  60.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  63.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  53.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  70.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  68.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  BOS&EOS (+UNI +AUG)  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  89.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  66.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  89.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  63.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  74.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  89.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  Table 5: Overall Results (Word-Level).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We report the macro/weighted average F1 of the BIO labeling task and  the F1 score of the exact SU span extraction task.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Details of our experimental setup are discussed in §6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Secondly, we focus on the results where we  only utilize the standard sentence segmentation  dataset (WSJ Train/Dev) for model development.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  In this setup, we also report the results of applying  our data augmentation (+AUG) and unidirectional  model (+UNI) techniques from §6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  Due to the data discrepancy between WSJ and  EWT, we find a natural drop in performance com-  pared to the previous setup using in-domain EWT  Train/Dev.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' However, we can verify that our tech-  niques (+AUG, +UNI) generally help to alleviate  this issue, and our proposed method performs on  par or slightly better than the EOS-only baselines  when applying these techniques.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' It is especially  worth noting the improvement in the exact SU  span extraction task (reaching 64∼72% F1), where  the advantage of our method is the most conspic-  uous and consistent in both word- and character-  level evaluation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This improvement can also be  explained by the higher performance in the B-label  prediction with our method (Appendix E), which  is a prerequisite for accurate SU span extraction.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Finally, we note that the EOS-only baseline with-  out forcing the last EOS can be quite competitive  with shorter inputs (pCC = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5 and postprocessing)  but performs considerably worse when the input  texts are longer (pCC = 0).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This is because the  baseline can only predict the last segment of the  input as an NSU, which is less problematic when  the input texts are shorter but becomes increasingly  problematic with longer inputs (since most NSUs  will not be able to be removed).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In contrast, our  proposed method performs much more robustly  under various input lengths.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  7We did not observe any improvement from applying these  techniques to the in-domain dataset (EWT Train/Dev), which  is consistent with our motivation and expectation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Through further experiments and analyses, we  found that (i) the results are stable across different  hyperparameter choices, (ii) predictions are reason-  able especially when using the in-domain dataset  (EWT Train/Dev) for model development, and (iii)  our methods do not sacrifice performance on the  formal/edited texts of the sentence segmentation  dataset (WSJ Test).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' These detailed evidences can  be found in Appendix F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  8  Conclusion  In this paper, we introduced a novel task of sen-  tence identification, where we aim to identify SUs  while excluding NSUs in a given text (§3).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Through  sentence identification, we can clearly distinguish  the portions of the text that are appropriate (or not)  for the prediction and evaluation of sophisticated  linguistic analyses, such as dependency parsing,  semantic role labeling, etc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To conduct sentence identification, we proposed  a simple yet effective method of combining the  BOS and EOS labeling models to determine the  SUs and NSUs (§4).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To evaluate sentence iden-  tification, we designed an automatic, language-  independent procedure to convert the UD corpora  into sentence identification benchmarks (§5).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  In our experiments, we developed the BOS/EOS  labeling models by finetuning pretrained RoBERTa  (§6).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Based on the experimental results, we showed  that our proposed method combining the BOS and  EOS labels outperforms sentence segmentation  baselines which only utilize EOS labels in all of  the considered settings (§7).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Overall, we expect  sentence identification to be a fundamental frame-  work for the preprocessing of noisy, informal, or  non-standard texts in the real world.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Limitations  Firstly, our current experiments are limited to En-  glish and cover only five domains of web media  texts in EWT.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' However, our task formulation (§3),  method (§4), and evaluation framework (§5) are  fully agnostic to the language and domain.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Hence  it is straightforward to conduct experiments in dif-  ferent languages or domains (as long as they are  supported in the UD).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' While we expect similar re-  sults with different languages/domains, we leave  further investigation as a future work.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Secondly, while our method performs reliably  when the in-domain dataset is available, there is  still a huge room left for improvement without re-  lying on such resources (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' only using the stan-  dard sentence segmentation dataset).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To make our  method fully practical, we still need to improve on  the accuracy and robustness in such cross-domain  scenarios.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' One potential approach is to refine the  definitions of SU and NSU probabilities from §4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  to make sentence identification more robust.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' For  instance, we can incorporate span-level scores in-  stead of only using word-level BOS/EOS probabil-  ities to define the SU/NSU probabilities.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We leave  further improvement and extension of our approach  as an important future work.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Finally, our methods are currently evaluated on  the (exact) SU span extraction task.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Ideally, we  should also evaluate the methods on downstream  applications such as POS tagging, syntactic pars-  ing, semantic role labeling, etc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' However, we still  expect that the (exact) SU span extraction will play  a primary role in the evaluation, since accurate  (say human-level) identification of SUs/NSUs will  likely provide unprecedented benefits on a wide va-  riety of NLP applications dealing with real-world  texts.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' While we leave the precise analyses on down-  stream applications as future work, our contribu-  tions make the first foundational step towards ex-  panding the capability of the long-established sen-  tence segmentation task.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  References  Wirote Aroonmanakun et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2007.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Thoughts on word  and sentence segmentation in thai.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proceedings  of the Seventh Symposium on Natural language Pro-  cessing, pages 85–90.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Marie-Catherine de Marneffe, Christopher D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Man-  ning, Joakim Nivre, and Daniel Zeman.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Uni-  versal Dependencies.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Computational Linguistics,  47(2):255–308.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Timothy Dozat and Christopher D Manning.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2017.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Deep biaffine attention for neural dependency pars-  ing.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' of ICLR.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Timothy Dozat and Christopher D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Manning.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Simpler but more accurate semantic dependency  parsing.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proceedings of the 56th Annual Meet-  ing of the Association for Computational Linguis-  tics (Volume 2: Short Papers), pages 484–490, Mel-  bourne, Australia.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Association for Computational  Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Dan Gillick.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2009.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Sentence boundary detection and  the problem with the U.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proceedings of Human  Language Technologies: The 2009 Annual Confer-  ence of the North American Chapter of the Associa-  tion for Computational Linguistics, Companion Vol-  ume: Short Papers, pages 241–244, Boulder, Col-  orado.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Association for Computational Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Mandar Joshi, Danqi Chen, Yinhan Liu, Daniel S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Weld, Luke Zettlemoyer, and Omer Levy.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  SpanBERT: Improving pre-training by representing  and predicting spans.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Transactions of the Associa-  tion for Computational Linguistics, 8:64–77.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Tibor Kiss and Jan Strunk.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2006.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Unsupervised mul-  tilingual sentence boundary detection.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Computa-  tional Linguistics, 32(4):485–525.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Philipp Koehn, Hieu Hoang, Alexandra Birch, Chris  Callison-Burch, Marcello Federico, Nicola Bertoldi,  Brooke Cowan,  Wade Shen,  Christine Moran,  Richard Zens, Chris Dyer, Ondˇrej Bojar, Alexandra  Constantin, and Evan Herbst.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2007.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Moses: Open  source toolkit for statistical machine translation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In  Proceedings of the 45th Annual Meeting of the As-  sociation for Computational Linguistics Companion  Volume Proceedings of the Demo and Poster Ses-  sions, pages 177–180, Prague, Czech Republic.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' As-  sociation for Computational Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Yinhan Liu, Jiatao Gu, Naman Goyal, Xian Li, Sergey  Edunov, Marjan Ghazvininejad, Mike Lewis, and  Luke Zettlemoyer.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Multilingual denoising  pre-training for neural machine translation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Transac-  tions of the Association for Computational Linguis-  tics, 8:726–742.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Man-  dar Joshi, Danqi Chen, Omer Levy, Mike Lewis,  Luke Zettlemoyer, and Veselin Stoyanov.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  RoBERTa: A robustly optimized BERT pretraining  approach.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' arXiv preprint arXiv:1907.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='11692.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Mitchell P.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Marcus, Beatrice Santorini, and Mary Ann  Marcinkiewicz.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 1993.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Building a large annotated  corpus of English: The Penn Treebank.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Computa-  tional Linguistics, 19(2):313–330.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Geoffrey Nunberg.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 1990.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' The linguistics of punctua-  tion.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 18.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Center for the Study of Language (CSLI).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Jonathon Read, Rebecca Dridan, Stephan Oepen, and  Lars Jørgen Solberg.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2012.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Sentence boundary de-  tection: A long solved problem?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proceedings of  COLING 2012: Posters, pages 985–994, Mumbai,  India.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' The COLING 2012 Organizing Committee.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Ines  Rehbein,  Josef  Ruppenhofer,  and  Thomas  Schmidt.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Improving sentence boundary detec-  tion for spoken language transcripts.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proceedings  of the Twelfth Language Resources and Evaluation  Conference, pages 7102–7111, Marseille, France.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  European Language Resources Association.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Michael D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Riley.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 1989.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Some applications of tree-  based modelling to speech and language.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Speech  and Natural Language: Proceedings of a Workshop  Held at Cape Cod, Massachusetts, October 15-18,  1989.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Dwijen Rudrapal, Anupam Jamatia, Kunal Chakma,  Amitava Das, and Björn Gambäck.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2015.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Sen-  tence boundary detection for social media text.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In  Proceedings of the 12th International Conference  on Natural Language Processing, pages 254–260,  Trivandrum, India.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' NLP Association of India.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Natalia Silveira,  Timothy Dozat,  Marie-Catherine  de Marneffe, Samuel Bowman, Miriam Connor,  John Bauer, and Chris Manning.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2014.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' A gold stan-  dard dependency corpus for English.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proceedings  of the Ninth International Conference on Language  Resources and Evaluation (LREC’14), pages 2897–  2904, Reykjavik, Iceland.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' European Language Re-  sources Association (ELRA).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob  Uszkoreit, Llion Jones, Aidan N Gomez, Ł ukasz  Kaiser, and Illia Polosukhin.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2017.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Attention is all  you need.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' of NeurIPS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Daniel J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Walker, David E.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Clements, Maki Darwin,  and Jan W.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Amtrup.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2001.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Sentence boundary de-  tection: a comparison of paradigms for improving  MT quality.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proceedings of Machine Translation  Summit VIII, Santiago de Compostela, Spain.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Xiaolin Wang, Masao Utiyama, and Eiichiro Sumita.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Online sentence segmentation for simultane-  ous interpretation using multi-shifted recurrent neu-  ral network.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proceedings of Machine Translation  Summit XVII: Research Track, pages 1–11, Dublin,  Ireland.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' European Association for Machine Transla-  tion.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Rachel Wicks and Matt Post.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' A unified approach  to sentence segmentation of punctuated text in many  languages.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Proceedings of the 59th Annual Meet-  ing of the Association for Computational Linguistics  and the 11th International Joint Conference on Nat-  ural Language Processing (Volume 1: Long Papers),  pages 3995–4007, Online.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Association for Computa-  tional Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Nianwen Xue and Yaqin Yang.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2011.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Chinese sen-  tence segmentation as comma classification.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' In Pro-  ceedings of the 49th Annual Meeting of the Associ-  ation for Computational Linguistics: Human Lan-  guage Technologies, pages 631–635, Portland, Ore-  gon, USA.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Association for Computational Linguis-  tics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Nina Zhou, AiTi Aw, Nattadaporn Lertcheva, and Xu-  ancong Wang.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 2016.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' A word labeling approach to  Thai sentence boundary detection and POS tagging.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  In Proceedings of COLING 2016, the 26th Inter-  national Conference on Computational Linguistics:  Technical Papers, pages 319–327, Osaka, Japan.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' The  COLING 2016 Organizing Committee.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  A  Dynamic Programming Algorithm  To find the maximum value (and the argmax) of  Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 4 from §4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2, we rely on a simple dynamic  programming framework.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To be specific, we con-  sider the partial labeling of BOS and EOS up to  W ≤k = (w0, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=', wk), where k ≤ N − 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Then,  we aim to compute the maximum log probability  of Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 4 based on the partial labeling, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' using  W ≤k in place of W .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Since the labeling is partial, W ≤k may end in-  side the SU (i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' the last label is BOS) or outside the  SU (i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' the last label is EOS).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Let log pIS(k+1) de-  note the maximum log probability when W ≤k ends  inside the SU and log pOS(k + 1) the maximum  log probability when W ≤k ends outside the SU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Then,' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' we can initialize log pIS(0) = log 0 = −∞,' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  log pOS(0) = log 1 = 0 (since we always start out-  side the SU) and iteratively update the two values  as follows:  log p′  IS(i) = max { log pIS(i) + log (1−pBOS(wi)),' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  log pOS(i) + log pBOS(wi) }  log p′  OS(i) = log pOS(i) + log (1−pBOS(wi))  log pIS(i+1) = log p′  IS(i) + log (1−pEOS(wi))  log pOS(i+1) = max { log p′  IS(i) + log pEOS(wi),' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  log p′  OS(i) + log (1−pEOS(wi)) }  (5)  Note that we first update pIS(i) → p′  IS(i) and  pOS(i) → p′  OS(i) based on the BOS probability  pBOS(wi).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Then, we update p′  IS(i)→pIS(i+1) and  p′  OS(i)→pOS(i+1) based on the EOS probability  pEOS(wi).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8 The iterative procedure is illustrated in  Figure 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Finally, we can compute the log probability  log pOS(N) (since we always end outside the SU),  which corresponds to the maximum value of Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To obtain the argmax, we can simply incorporate  backtracking during the iterative updates of Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Through this dynamic programming framework,  we can ensure that the restrictions from §4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2 are  satisfied: namely, (i) the first label should be BOS,  (ii) the last label should be EOS, and (iii) BOS and  EOS labels need to appear alternately.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  In practice, we can limit the candidates of BOS  indices to the subset where pBOS(wi) is higher  than a certain threshold c.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This can be efficiently  implemented by simply skipping the updates of  p′  IS(i) and p′  OS(i), i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' using p′  IS(i) = pIS(i) and  8Note that if a single word wi is labeled as both BOS and  EOS at the same time, we can extract it as a single SU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  p′  OS(i) = pOS(i), if pBOS(wi) < c.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9 Likewise, we  can limit the candidates of EOS indices by skip-  ping the updates of pIS(i + 1) and pOS(i + 1) if  pEOS(wi) < c.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Generally speaking, this leads to  a more efficient algorithm: therefore, we use the  candidate threshold of c = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1 for restricting both  BOS and EOS indices throughout our experiments.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  B  SU and NSU Examples  In Table 6, we provide more examples of SUs and  NSUs identified based on our procedure described  in §5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' As for the SUs, we can verify that EWT con-  tains clean, formal SUs with appropriate capitaliza-  tion and punctuation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We can also verify that EWT  contains various types of informal SUs, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' that  lack capitalization/punctuation, use non-standard  casing, end with emoticons, include spelling errors,  concatenate consecutive SUs without a space, etc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  C  Label Assignment and Conversion  In this section, we explain the precise procedure  on how we (i) assign the gold character-level  labels, (ii) convert the character-level labels to  word/subword-level labels, and (iii) convert the  subword-level labels to character/word-level labels.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  We limit our explanation to BIO labels, since it is  straightforward to convert them to the combination  of BOS and EOS labels (and vice versa).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Firstly, we can assign the gold character-level  labels from the UD annotation by taking the  character-level alignment, which determines the  exact spans of SUs and NSUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' From the character-  level labels, we can assign the word- or subword-  level labels based on the following rule:  If the word (or subword) contains a character  with the B-label, assign it the B-label.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Else if it contains a character with the I-label,  assign the I-label.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Otherwise assign the O-label.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  For instance, this procedure is used to create the  subword-level labels for training our BOS/EOS  labeling models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  To evaluate our methods, we need to convert  the subword-level labels produced by our methods  into the character-level labels, which can then be  converted into the word-level labels (based on the  9This is equivalent to forcing wi to be non-BOS, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' set-  ting pBOS(wi) = 0 in Eq.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Figure 2: Illustration of the dynamic programming procedure.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  SUs  President Bush on Tuesday nominated two individuals to replace retiring jurists on federal courts in the Washington area.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Unfortunately, Mr.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Lay will be in San Jose, CA participating in a conference, where he is a speaker, on June 14.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  “In 1972, there was an enormous glut of pilots,” Campenni says.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  PS – There is a happy hour tonight at Scudeiros on Dallas Street (just west of the Met Garage) beginning around 5:00.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  2) Your vet would not prescribe them if they didn’t think it would be helpful.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  BUT EVERYONE HAS THERE OWN WAY!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' !' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  The motel is very well maintained, and the managers are so accomodating, it’s kind of like visiting family each year!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' ;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='-)  where can I find the best tours to the Mekong Delta at reasonable prices?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  it seems like its healthier too, but its prolly not.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  I have wifi at my house, but thats just at my house.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='is there anyway i can buy some card to make the ipod itself have wifi?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  NSUs  —->===}*{===<—-  Lisa_coverletter.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='doc << File: Lisa_coverletter.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='doc >>  Thur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Sept.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 28 - Paris (Versailles or Fontainbleu - half day side trip)  9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3m - Number of US unemployed in April 2004.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Game 1: Monday, May 28 @ 2:00PM vs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Los Angeles SPARKS  Mixed Tempura.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='25 Shrimp or vegetable tempura & salad.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Infinity stereo, bucket seats, nerf bars, tool box, bed liner, camper tow package, 5 speed manual.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  printing, printing, copies, printing, copies, printing,  A++++ !' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' !' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' !' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Dear Sir / Madam,  Table 6: Examples of gold SUs and NSUs in the English Web Treebank (EWT) identified based on our procedure  (§5).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Each line corresponds to one example of SU or NSU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  previous procedure).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To convert a subword-level  label into a sequence of character-level labels, we  apply the following rule (where n denotes the num-  ber of characters in the subword):  If the subword has the B-label, the character-  level labels are 1 B-label followed by n − 1  I-labels.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  If the subword has the I-label, the character-  level labels are n I-labels.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  If the subword has the O-label, the character-  level labels are n O-labels.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  D  Details on the Model Setup  As discussed in §6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1, we finetune the pretrained  RoBERTaBASE publicly available on the Hugging-  Face model hub10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We add a binary BOS/EOS  classifier on top of the encoder, which is a single-  layer MLP with a hidden size of 768.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We share the  10https://huggingface.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='co/models  encoder parameters and use different classifiers for  the BOS/EOS predictions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' The BOS/EOS models  are trained jointly by summing their losses.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  When we combine the unidirectional models  (+UNI), we take the same approach and use differ-  ent classifiers for the unidirectional/bidirectional  models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Again, the encoder parameters are shared  and all models are trained jointly.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  As for the training data preparation, we apply the  unit concatenation and data augmentation (+AUG)  on the fly, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' we see different concatenation and  augmentation of the units in each iteration.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' The  same procedure is applied on the validation set.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  During data augmentation, we remove the last  sentence ending punctuation based on the following  regular-expression, similar to the candidate bound-  ary detector in ERSATZ (Wicks and Post, 2021):  (.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='∗PeP ∗) where P denotes the set of punctua-  tions and Pe ⊂ P denotes the sentence ending  punctuations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Since our experiments are conducted on English,  Update based on PBos(wi)  Update based on peos(wi)  In-sentence: pis(i)  pis(i) = max (pis(i) * (1 - pBos(wi),  Pis(i + 1) = pis(i) * (1 -peos (wi))  1 - PBos(wi)  1 - PEos(wi)  Pos(i) * pBos(wi) )  PEos(wi)  PBos(wi)  Text  Out-of-sentence: pos(i)  Pos(i + 1) = max (pis(i) * peos(wi)  pos(i) = pos(i) * (1 - pBos (wi)  1 - PBos(wi)  1 - peos(wi)  ps(i) * (1 - peos(wi)) )  1-m .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Tm "om Jae (ns apisul=) s u!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' buiaq jo K!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='qeoud :(2)sid  Return: Pos(N)  Start: pis(O) = 0, Pos(0) = 1  Pos(i): probability of being in OS (=outside SU) after Wo, Wi, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='.Wi-1we use P = {.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='?!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' ")’} and Pe = {.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='?!' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' }.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Finally, all models are implemented in Pytorch  and trained on a single Tesla V100-SXM2-32GB  GPU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We use a batch size of 8, accumulate the  gradients for 32 batches, and apply the gradient  clipping at 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0 before updating the model weights.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  As for the optimizer, we use Adam with the initial  learning rate of 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0001 and exponentially decay the  learning rate by γ = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='95 after each epoch.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We  check the validation loss every 200 batches and  stop the training early if there is no improvement  for 5 consecutive evaluations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  E  The Full Experimental Results  In this section, we report the full results of our  experiments which did not fit in §7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Table 7 shows  the word-level F1 scores for each B-, I-, and O-  label prediction.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Table 8 shows the overall results  for the character-level evaluation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Generally speaking, we can confirm the same  results as observed in §7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Firstly, our proposed  method significantly outperforms the baselines  when we use the EWT Train/Dev dataset for model  development.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Secondly, our method performs  slightly better than (or at least on par with) the  baselines when developed on the WSJ Train/Dev  dataset.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Finally, the baseline without forcing the  last EOS is competitive with shorter inputs (pCC =  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5 and postprocessing) but performs considerably  worse when the input texts are longer (pCC = 0).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  F  Further Experiments and Analyses  In this section, we provide further experiments  and analyses to complement our study.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To be spe-  cific, we provide discussions on the effect of the  choice of hyperparameters (F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1), qualitative anal-  yses based on example model outputs (F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2), and  evaluation of sentence identification based on the  sentence segmentation dataset (F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  Effect of Hyperparameters  As a default configuration, we used pDA  =  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3, pT R =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1 for the data augmentation (+AUG)  and λ = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5 for the unidirectional model ensem-  bling (+UNI).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' To examine the effect of the choice  of these hyperparameters, we conducted further  experiments by changing these default hyperpa-  rameters.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Note that all evaluation results in this  subsection are based on BOS&EOS (+UNI +AUG)  developed on WSJ Train/Dev.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Firstly, we focus on the data augmentation and  report the results of our method trained with dif-  ferent sets of pDA and pT R (with λ fixed at 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Since increasing pDA leads to higher recall (and  lower precision) of SU extraction and increasing  pT R leads to higher precision (and lower recall),  we used a fixed ratio of pDA : pT R = 3 : 1 which  seemed to make a good trade-off.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' As shown in  Table 9, the results are generally stable with the  different choices of the hyperparameters.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' However,  more data augmentation (with larger values of pDA  and pT R) tends to slightly improve the performance,  especially for the exact SU span extraction.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Secondly, we focus on the unidirectional model  ensembling and report the results of changing the  linear interpolation rate λ ∈ [0, 1], where λ = 0 is  equivalent to using only the bidirectional models  and λ = 1 only the unidirectional models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We fix  pDA =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3 and pT R =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1 and only change λ at the  inference time without retraining the unidirectional  or bidirectional models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' As shown in Figure 3, we  found that unidirectional and bidirectional models  generally have complementary benefits, and choos-  ing the intermediate value of λ leads to the best  performance.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' The results also indicate that we may  be able to obtain further improvement by tuning  λ on the validation set, although we simply fixed  λ = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5 throughout our experiments.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  Qualitative Analyses  In Table 10 and 11, we show the actual predictions  made by our proposed method developed on EWT  Train/Dev and WSJ Train/Dev.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' For the latter, we  applied +UNI and +AUG with the default hyperpa-  rameters.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  In the first example (Table 10), we can verify  that both models identify the correct SU span while  removing the non-sentential header as the NSU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  This is a relatively easy example, since the start of  the SU is capitalized and less ambiguous.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  In the second example (Table 11), we can ob-  serve that our method using in-domain data (EWT  Train/Dev) extracts the correct SU span, while our  method developed on out-of-domain data (WSJ  Train/Dev) incorrectly excludes a part of an SU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  This seems to be a relatively difficult example,  since the start of the SU is not capitalized and more  ambiguous.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' It is worth noting that such SUs can  be reliably extracted when we can leverage the in-  domain annotation of gold SUs and NSUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Train/Dev  Datasets  Model  EWT Test (pCC = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5)  EWT Test (pCC = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0)  EWT Test (Postprocess)  B-Label  I-Label  O-Label  B-Label  I-Label  O-Label  B-Label  I-Label  O-Label  EWT  Train/Dev  EOS-Only  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  66.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  78.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  90.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  EOS-Only (force last)  79.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  77.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  81.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  BOS&EOS  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  81.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  WSJ  Train/Dev  EOS-Only  78.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  42.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  37.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  EOS-Only (force last)  76.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  81.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  EOS-Only (+AUG)  79.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  24.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  78.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  82.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  35.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  BOS&EOS  79.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  40.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  35.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  BOS&EOS (+UNI)  79.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  37.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  76.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  20.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  34.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  BOS&EOS (+UNI +AUG)  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  38.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  39.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  41.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  Table 7: BIO Labeling Results (Word-Level).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We report the F1 scores for each B-, I- and O-label prediction.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Train/Dev  Datasets  Model  EWT Test (pCC = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5)  EWT Test (pCC = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0)  EWT Test (Postprocess)  BIO  BIO  Span  BIO  BIO  Span  BIO  BIO  Span  Macro  Weighted  Macro  Weighted  Macro  Weighted  EWT  Train/Dev  EOS-Only  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  92.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  58.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  81.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  58.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  81.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  EOS-Only (force last)  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  81.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  60.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  80.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  58.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  79.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  62.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  BOS&EOS  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  92.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  96.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  84.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  96.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  88.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  WSJ  Train/Dev  EOS-Only  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  80.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  48.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  73.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  67.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  EOS-Only (force last)  56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  80.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  53.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  54.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  80.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  48.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  79.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  61.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  EOS-Only (+AUG)  64.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  81.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  54.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  69.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  84.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  66.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  BOS&EOS  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  81.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  48.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  68.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  BOS&EOS (+UNI)  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  65.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  52.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  68.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  BOS&EOS (+UNI +AUG)  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  66.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  63.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  73.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  Table 8: Overall Results (Character-Level).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We report the macro/weighted average F1 of the BIO labeling task  and the F1 score of the exact SU span extraction task.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  Evaluation on the Sentence Segmentation  Dataset  Finally, we report the results of sentence identifica-  tion on the standard sentence segmentation dataset  (WSJ Test).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  In Table 12, we summarize the WSJ dataset  statistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Note that WSJ only contains SUs and  do not contain any NSUs (O-labels).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' However, we  can still evaluate the performance using the same  metrics, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' the macro/weighted average F1 of the  BIO labeling task and the F1 of the exact SU span  extraction task.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='11  Table 13 summarizes the word-level evaluation  results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' Since we are evaluating on WSJ Test, the  performance is naturally better when the models  are trained on WSJ Train/Dev rather than EWT  Train/Dev (which is now out-of-domain).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  When the models are trained on EWT, we found  that the baseline (EOS-Only) forcing the last EOS  performs the best.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This is natural, since this base-  line better reflects the nature of the sentence seg-  mentation dataset where all units are SUs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' How-  11Since the O-label does not exist, we report the macro  average F1 as the average F1 scores of the B-label and I-label  predictions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  ever, our method (BOS&EOS) is still comparable  to this baseline and do not (or minimally) sacrifice  performance on such datasets.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  When the models are trained on WSJ, we found  that our method without +UNI or +AUG performs  the best.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' This is most likely because we can lever-  age the knowledge of BOS to predict EOS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' When  we apply the data augmentation (+AUG) and uni-  directional model ensembling (+UNI), we observe  a slight decrease in performance compared to our  vanilla method.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' However, the results are still com-  parable and even outperforms the baselines in some  metrics (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' the exact SU span extraction task).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Overall, we can conclude that our methods do  not sacrifice the performance on the the clean,  edited texts of the sentence segmentation dataset.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Evaluation  Augmentation Rates  EWT Test (pCC = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5)  EWT Test (pCC = 0)  EWT Test (Postprocess)  BIO  BIO  Span  BIO  BIO  Span  BIO  BIO  Span  Macro  Weighted  Macro  Weighted  Macro  Weighted  Word-Level  pDA =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='15, pT R =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='05  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  89.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  65.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  88.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  62.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  73.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  89.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  pDA =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3, pT R =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  89.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  66.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  89.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  63.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  74.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  89.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  pDA =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='45, pT R =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='15  73.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  90.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  67.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  73.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  89.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  64.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  75.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  90.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  Character-  Level  pDA =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='15, pT R =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='05  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  65.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  73.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  62.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  73.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  pDA =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3, pT R =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  66.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  63.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  73.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  pDA =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='45, pT R =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='15  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  67.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  64.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  73.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  Table 9: Effect of Data Augmentation Rates (Word/Character-Level).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  We use different data augmentation  rates (pDA and pT R) and evaluate BOS&EOS (+UNI +AUG) developed on WSJ Train/Dev.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  We report the  macro/weighted average F1 of the BIO labeling task and the F1 score of the exact SU span extraction task.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  B  Developed  .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 06/04/2001 05:54 PM Can you pass this along to Elizabeth to ensure Sanders  on EWT  E  is on board as well?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  B  Developed  .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' 06/04/2001 05:54 PM Can you pass this along to Elizabeth to ensure Sanders  on WSJ  E  is on board as well?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Table 10: Example Outputs (Both Correct).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We show the predictions made by our proposed method (BOS&EOS)  developed on EWT Train/Dev (top) or WSJ Train/Dev (bottom).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We can verify that both methods identify the  correct SU span while removing the non-sentential header as the NSU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  B  Developed  with my breakfast I like bacon and sausage when I having a big breakfast like  on EWT  E  a grand slam with pancakes and the works.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  B  Developed  with my breakfast I like bacon and sausage when I having a big breakfast like  on WSJ  E  a grand slam with pancakes and the works.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Table 11: Example Output with One Incorrect Case.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We show the predictions made by our proposed method  (BOS&EOS) developed on EWT Train/Dev (top) or WSJ Train/Dev (bottom).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We can verify that the former  extracts the correct SU span, while the latter incorrectly excludes the first prepositional phrase as an NSU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Train  Dev  Test  Total SUs  37,447  2,021  7,442  Total NSUs  0  0  0  Word-Level  B-Label  37,447  2,021  7,442  I-Label  805,387  44,354  163,132  O-Label  0  0  0  Character-Level  B-Label  37,447  2,021  7,442  I-Label  4,308,729  236,798  876,461  O-Label  0  0  0  Table 12: WSJ dataset statistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  (a) EWT Test (pCC =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5), BIO Macro (b) EWT Test (pCC =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5), BIO Weighted  (c) EWT Test (pCC =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5), Span  (d) EWT Test (pCC =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0), BIO Macro (e) EWT Test (pCC =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0), BIO Weighted  (f) EWT Test (pCC =0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0), Span  (g) EWT Test (Postproc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' ), BIO Macro  (h) EWT Test (Postproc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' ), BIO Weighted  (i) EWT Test (Postproc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' ), Span  Figure 3: Effect of the Unidirectional Model Interpolation Rate (Word-Level).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We change λ ∈ [0, 1] and report  the macro/weighted average F1 of the BIO labeling task and the F1 score of the exact SU span extraction task.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Interpolated results are shown in blue and non-interpolated results (i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' λ = 0) shown in red.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' The line shows the  mean and the shade shows the standard deviation from the five experimental runs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  Train/Dev  Datasets  Model  WSJ Test (pCC = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5)  WSJ Test (pCC = 0)  BIO  BIO  Span  BIO  BIO  Span  Macro  Weighted  Macro  Weighted  EWT  Train/Dev  EOS-Only  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  EOS-Only (force last)  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  BOS&EOS  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  86.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  WSJ  Train/Dev  EOS-Only  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  92.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  90.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  EOS-Only (force last)  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  92.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  90.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  EOS-Only (+AUG)  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  92.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6±2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  87.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  BOS&EOS  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  93.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  BOS&EOS (+UNI)  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  92.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='9±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='5  91.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  BOS&EOS (+UNI +AUG)  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  94.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='7  98.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1±0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='3  91.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8±1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='1  Table 13: Overall Results on WSJ Test (RoBERTa, Word-Level).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content=' We report the macro/weighted average F1 of  the BIO labeling task and the F1 score of the exact SU span extraction task.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='90  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='89  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='88  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='87  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='86  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  入0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='73  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='72  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='71  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='70  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='69  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='68  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='67  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='00.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='73  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='72  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='71  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='70  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='69  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='68  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  入0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='90  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='89  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='88  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='87  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='86  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='85  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='00.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='66  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='64  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='62  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='60  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='58  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='56  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='54  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  入0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='74  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='72  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='70  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='68  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='66  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='64  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='62  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='60  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='00.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='90  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='88  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='86  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='84  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='82  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  入0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='65  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='60  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='55  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='50  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='45  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='00.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='77  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='76  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='75  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='74  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='73  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='72  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='71  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}
+page_content='0' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/dNFQT4oBgHgl3EQfjDY0/content/2301.13352v1.pdf'}