diff --git "a/99FLT4oBgHgl3EQfCS7y/content/tmp_files/load_file.txt" "b/99FLT4oBgHgl3EQfCS7y/content/tmp_files/load_file.txt"
new file mode 100644--- /dev/null
+++ "b/99FLT4oBgHgl3EQfCS7y/content/tmp_files/load_file.txt"
@@ -0,0 +1,2656 @@
+filepath=/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf,len=2655
+page_content='Byte Pair Encoding for Symbolic Music  Nathan Fradet 1 2 Jean-Pierre Briot 1 Fabien Chhel 3 Amal El Fallah Seghrouchni 1 Nicolas Gutowski 4  Abstract  The symbolic music modality is nowadays mostly  represented as discrete and used with sequential  models such as Transformers, for deep learning  tasks.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Recent research put efforts on the tokeniza-  tion, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' the conversion of data into sequences  of integers intelligible to such models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' This can  be achieved by many ways as music can be com-  posed of simultaneous tracks, of simultaneous  notes with several attributes.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Until now, the pro-  posed tokenizations are based on small vocabular-  ies describing the note attributes and time events,  resulting in fairly long token sequences.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In this  paper, we show how Byte Pair Encoding (BPE)  can improve the results of deep learning models  while improving its performances.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We experiment  on music generation and composer classification,  and study the impact of BPE on how models learn  the embeddings, and show that it can help to in-  crease their isotropy, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', the uniformity of the  variance of their positions in the space.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Introduction  Deep learning tasks on symbolic music are nowadays mostly  tackled by sequential models1, such as the Transformers  (Vaswani et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2017).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' These models receive sequences of  tokens as input, and convert them to learned embedding  vectors.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' A token is an integer associated to a high level  element, such as a word or sub-word in natural language,  and both are linked in a vocabulary that acts as a look-up  table.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' An embedding represents the semantic information of  a token as a vector of fixed-size, and is learning contextually  by the model.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' To use such models for symbolic music, one  needs to tokenize the data, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', convert it to sequences of  tokens that can be decoded back.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' This can be achieved by  several ways, as music can be composed of simultaneous  tracks, of simultaneous notes with several attributes such as  1LIP6, Sorbonne University - CNRS, Paris, France 2Aubay,  Boulogne-Billancourt, France 3 ESEO-TECH / ERIS, Angers,  France 4University of Angers, Angers, France.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Correspondence to:  Nathan Fradet <nathan.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='fradet@lip6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='fr>.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  1Commonly referred as Language Models (LM)  their pitch and duration.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Recently, the token representation of symbolic music has  been extensively studied, with the goal to improve 1) the  results, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' the quality of generated results or the accuracy  of a certain Music Information Retrieval (MIR) task, and;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  2) the efficiency of the models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The former is tackled with  more expressive representations (Huang & Yang, 2020;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Ker-  marec et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2022), and the latter by representations based  on either token combinations (Payne, 2019;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Donahue et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  2019), or embedding pooling (Hsiao et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Zeng et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  2021;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Ren et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2020), which reduce the overall sequence  length.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Still, current tokenizations only use tokens represent-  ing the values of time and note attributes, such as Pitch or  Duration.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' This comes with a big limitation: these tokens do  not carry much information by themselves, and neither their  associated embeddings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' By analogy to natural language,  these tokens are closer to the characters than words.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Yet, the  expressive information carried by music is deduced by the  combinations of its notes and their attributes.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Considering  the infinite possible arrangements, deep learning models  may struggle to implicitly learn their common features.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In this paper, we study the application of Byte Pair En-  coding (BPE, described in Section 3) for symbolic music  generation, aiming to improve the two objectives mentioned  above, while making the models learn more isotropic em-  bedding representations in some cases.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' To the best of our  knowledge, BPE has yet not been studied for the symbolic  music modality, although it can be applied on top of any  music tokenization that do not perform embedding pooling.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  This work aims at closing this gap by shedding light on the  results and performance gains of using BPE:  We experiment on two public datasets (Wang et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  2020b;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Kong et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021), with two base tokenizations,  on which BPE is learned with several vocabulary sizes,  on the generation and composer classification tasks,  and show that it improves the results;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  We compare BPE with other sequence reduction tech-  niques introduced in recent research;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  We study the geometry of the learned embeddings, and  show that BPE can improve their isotropy;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  We show some limits of BPE, such as on the proportion  arXiv:2301.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11975v1  [cs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='LG]  27 Jan 2023  Byte Pair Encoding for Symbolic Music  of sampled tokens, and that the vocabulary size has to  be carefully chosen.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  The source code is provided for reproducibility: https:  //github.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='com/Natooz/BPE-Symbolic-Music  The paper is organised as follows: Section 2 reviews the  related work while Section 3 sheds light on the BPE tech-  nique.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Section 4 describes our experimental settings and  Section 5 describes the evaluation metrics that we use for  the experimental evaluation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Section 6 presents the results  and analysis.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Furthermore, Section 7 provides an additional  study on the impact of BPE on how the models learn the  embeddings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Finally, Section 8 presents our conclusion and  perspectives.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Related work  In this section we start by reminding research of specific  music representation of symbolic music generation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Then,  we present how recent works put efforts on different strate-  gies to reduce the sequence length.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Finally, we explain their  limitations which conduce us to propose our novel approach  that is to apply Byte Pair Encoding in the field of symbolic  music for reducing sequence length.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Representation of symbolic music  Most works on symbolic music generation from deep learn-  ing use a specific music representation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Early research in-  troduced representations specifically tied to the training  data being used, such as DeepBach (Hadjeres et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2017),  FolkRNN (Sturm et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2015) or BachBot (Liang et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  2017).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Non-sequential models such as MuseGAN (Dong  et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2018) often represent music as pianoroll matrices.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Since, more universal representations have been studied, al-  lowing to convert any sequence of (simultaneous) notes into  tokens (Oore et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2018;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Huang & Yang, 2020;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Hadjeres  & Crestel, 2021;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Fradet et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Some of them are  depicted in Figure 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Sequence reduction strategies  In more recent works, efforts have been put towards the  efficiency.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Indeed, most recent models are based on the  Transformer architecture (Vaswani et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2017).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The atten-  tion mechanism, at the heart of Transformers, has however  a time and space complexity that grows quadratically with  the input sequence length.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' This is a well known bottleneck,  that led researchers to work on more efficient attention esti-  mations (Tay et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021), down to linear complexity.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In the  field of symbolic music specifically, researchers worked on  strategies to reduce the sequence length in order to increase  1) the efficiency of the models;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 2) the scope of the attention  mechanism;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 3) the quality of the generated results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' These  Bar Pos.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 0  Pitch D3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 22Dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 7Pos.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 7  Pitch A3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 24Dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 7Pos.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 15  Pitch E4  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 24Dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 7Pos.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 27  Pitch G3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 16Dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 3 Bar Pos.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 0  Pitch A3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 20Dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 31  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 0  Pitch D3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 22Dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 7  Pitch A3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 24Dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 7  Pitch E4  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 24Dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 7  Pitch G3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 16Dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 3  Pitch A3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 20Dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 31  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 8  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 8  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 12  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 4  N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-On D3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 22  N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-On A3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 24  N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Off A3  N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-On E4  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 24  N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Off E4 N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-On G3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 16  N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-On A3  Vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 20  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 7  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 7  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 7  N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Off D3  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 3  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 3  N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Off G3  Ti.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Sh.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 31  N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Off A3  Music score  MIDI-Like  REMI  Structured  Figure 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' A sheet music and several token representations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  strategies can be split in two categories: 1) embedding pool-  ing strategies such as Compound Word (Hsiao et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021)  (CPWord), Octuple (Zeng et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021) or PopMag (Ren  et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2020);' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 2) token combination strategies such as in  MuseNet (Payne, 2019) or LakhNES (Donahue et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2019).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Embedding pooling consists in merging the embeddings of  several distinct tokens with a pooling operation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' This is  often done by concatenating the embeddings and projecting  the sequence, resulting in an aggregated embedding of fixed  size.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Token combinations is simply the use of a vocabu-  lary containing tokens that represent several values, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  Pitch-x Duration-y that represent both the pitch and  velocity information.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Limitations  However, these strategies show the following limitations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Embedding pooling: 1) requires a more complex training  procedure;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 2) for generation, inferring from such model  can be seen as sampling from a multivariate distribution,  which can be a delicate operation;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 3) the results can easily  degenerate if the pooling does not yield semantically rich  embeddings that represent the underlying tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' On the  other hand, token combinations of entire types of tokens can  lead to large vocabularies with unused tokens and potentially  non-optimized or unbalanced token distributions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  To the best of our knowledge, no work has been conducted  on applying BPE, introduced in Section 3, to symbolic mu-  sic generation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' A similar technique is used with Sympho-  nyNet (Liu et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2022), which does not rely on token adja-  cency but rather on the concurrence of multiple notes, and  they only experimented with a vocabulary size of 1k tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  The following section describes the Byte Pair Encoding  technique, its algorithm and depicts how it can be relevant  to use in the field of symbolic music.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  1  +Byte Pair Encoding for Symbolic Music  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Byte Pair Encoding  Byte Pair Encoding (BPE) (Gage, 1994) is a data com-  pression technique.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' It converts the most recurrent succes-  sive bytes (or in our case tokens) in a corpus into newly  created ones.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  For instance, in the character sequence  aabaabaacaa, the sub-sequence aa occurs three times  and is the most recurrent.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Learning and applying BPE on  this sequence would replace aa with a new symbol, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', d,  resulting in a reduced sequence dbdbdcd.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The latter can  be reduced again by replacing the db subsequence, giving  eedcd.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In the context of deep learning, BPE naturally in-  creases the size of the vocabulary, while reducing the overall  sequence lengths.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In practice BPE is learned on a corpus  until the vocabulary reaches a target size.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' BPE learning is  described by the pseudo-code of Algorithm 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Algorithm 1 Learning of BPE pseudo-code  Require: Base vocabulary V, target vocabulary size N,  dataset X  1: while |V|< N do  2:  Find s = {t1, t2} ∈ V2, from X, the most recurrent  token succession  3:  Add a new token t in V, mapping to s  4:  Substitute every occurrence of s in X with t  5: end while  6: return V  BPE is nowadays largely used in the NLP field as it allows  to encode rare words and segmenting unknown or com-  posed words as sequences of sub-word units (Sennrich et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  2016).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In symbolic music, notes are represented by successions  of tokens that represent the values of their attributes.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In  this context, BPE can allow to represent a note, or even a  succession of notes, that is very recurrent in the dataset, as  a single token.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For instance, a note that would be coded  as the succession of tokens Pitch D3, Velocity 60,  Duration 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0 could be replaced by a single new one.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Rare note (and attributes) would still be encoded as non-  BPE tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The same logic applies to time tokens, that can  also be associated to note tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Experimental settings  This section details the experimental protocol by describing  the models, the training and the datasets used along with the  specific tokenization processes.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Model and training  As we specifically focus on sequential models, we exper-  iment with the state of the art deep learning architecture  for most NLP tasks at the time of writing, the Transformer  (Vaswani et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2017) architecture.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The generator uses a  causal attention mask and is trained with teacher forcing,  while the classifier does not use attention mask and is first  pre-trained to retrieve randomized tokens then finetuned to  classify the input sequences.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' They are respectively similar  to GPT2 (Radford et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2019) and BERT (Devlin et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  2019).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The details of implementation, such as their sizes  and training, can be found in Appendix A  All models receive sequences between 384 and 460 tokens,  beginning with special BOS (Beginning of Sequence) and  ending EOS (End of Sequence) tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We split datasets  in two subsets: one only used for training and updating  the models, one for validation to monitor trainings, that is  also used to test the models after training.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' These subsets  represent respectively 65% and 35% of the original datasets.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Datasets  We experiment with two datasets: POP909 (Wang et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  2020b) and GiantMIDI (Kong et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  The POP909 dataset (Wang et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2020b) is composed of  909 piano tracks of Pop musics, with aligned MIDI and  audio versions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Each MIDI file contains three tracks: the  first is the lead melody, the second is secondary melodies  and bridges, the third is the arrangements with chords and  arpeggios.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For our experiments we merge all three tracks  into a single one.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  The GiantMIDI dataset (Kong et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021) is composed  of 10k piano MIDI files, transcribed from audio to MIDI  without downbeat and tempo estimation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Each file contains  a single track of non-interrupted piano music, often with  complex melodies and harmonies.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Considering the com-  plexity of its content, we make the assumption that it is a  difficult dataset for a model to learn from.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  We perform data augmentation on the pitch dimension on  both datasets.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Each MIDI file is augmented up and down to  two octaves.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Music tokenization  We experiment with Remi (Huang & Yang, 2020) and  TSD (for Time Shift Duration) as base tokenizations, on  which BPE will be applied on top.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Both tokenizations de-  scribe notes as a succession of the Pitch, Velocity and  Duration tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Remi represents time with Bar and  Position tokens, which respectively indicates when a  new bar is beginning and at which position within the time  is.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' TSD represents time with TimeShift tokens, indicat-  ing explicitly time movements.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  When tokenizing symbolic music, it is common to down-  sample continuous features to discrete sets of values.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For  instance, velocities can be downsampled from 128 to 32  Byte Pair Encoding for Symbolic Music  values.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' These sets should be sufficiently precise so that  the global information remains coherent (Huang & Yang,  2020;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Oore et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2018;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Hadjeres & Crestel, 2021).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Down-  sampling features helps models to learn more easily, as it  allows to reduce the perplexity of the predictions, especially  for values which are less commons in the training set.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The  details of our downsamplings can be found in Appendix B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  BPE is learned from tokenized corpuses, up to a maximum  of 1500 randomly picked files, to reduce the learning time.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  We choose to experiment with six vocabulary sizes.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' One  without BPE, and five where the original vocabulary size is  multiplied by 4, 10, 20, 50 and 100.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  To extend our analysis, we also experiment with a version  of TSD and Remi where Pitch and Velocity tokens  are merged (PVm), and one where Pitch, Velocity and  Duration are merged (PVDm).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' PVm is similar to the  strategy used with MuseNet (Payne, 2019).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We finally ex-  periment with the CPWord (Hsiao et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021) and Octuple  (Zeng et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021) embedding pooling strategies, that we  group with Remi in our experiments as they represent time  similarly.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We use the same pooling strategy, and sample  independently from the logits of each output modules.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For  implementation simplicity reasons, all embeddings have the  same size than the model dimension.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Evaluation metrics  Generative models are often evaluated with automatic met-  rics on the generated results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Image and audio models are  assessed with the Fr´echet Inception Distance (FID) (Heusel  et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2017) and Fr´echet Audio Distance (FAD) (Kilgour  et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2019), both comparing the distribution of original data  and generated results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Language models are often assessed  with BLEU (Papineni et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2002), ROUGE (Lin, 2004) or  other metrics that compare generated results with reference  sentences.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Automatic evaluation of symbolic music remains however  an open issue.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' It exists no reference-free metric measuring  its quality or fidelity.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Metrics with reference such as BLEU  may be suited for machine translation tasks, but remains  irrelevant for open-ended generation, such as in our case.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  We then perform both human and automatic evaluations, as  commonly done for symbolic music (Huang & Yang, 2020;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Huang et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2018;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Hsiao et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Our automatic met-  rics aim to measure the errors of prediction of the models,  and the similarity of some features.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Tokenization syntax error  Every tokenization has an underlying syntax of token type  and value successions, that can normally be made.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For  instance, if the last token of an input sequence is of type  Pitch, a tokenization could require that the next token to  predict must be of type Velocity.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We could also expect  a model to not predict more than once the same note at a  same moment, or to not go back in time.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Successions of incorrect token types can be interpreted as  errors of prediction.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' These errors can help us to measure  if a model has efficiently learned the music representation  and if it can yield coherent results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' With this motivation,  we introduce a new metric we called Tokenization Syntax  Errors (TSE).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Velocity  Pitch  Duration  Position  Bar  (a) REMI.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Velocity  Note-On  Note-Off  Time-Shift  (b) MIDI-Like  Figure 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Directed graphs of the token types succession (without  additional tokens) for a) REMI (Huang & Yang, 2020) and b)  MIDI-Like (Oore et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2018).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  We distinguish five categories of errors:  TSEtype: the predicted token does not have a type  that should follow the previous one.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For any tokeniza-  tion, we can draw a directed graph representing the  possible token types successions, such as in Figure 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  TSEtime: when using Position tokens, the pre-  dicted Position value is inferior or equal to the  current one, making the time goes backward.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  TSEdupn (duplicated note): when the model predicts  a note that has already been played at the current mo-  ment (by the same instrument).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  TSEnnof (no NoteOff): when using NoteOn and  NoteOff, and that a NoteOn token has been pre-  dicted with no NoteOff later to end it, or too distant  in time.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  TSEnnon (no NoteOn): when a NoteOff token is  predicted but the corresponding note has not been  played.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  For a given sequence of tokens, TSE measures the ratio,  scaled between 0 and 1, of errors for these five categories.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  A TSE of 0 means that there is no error in the sequence,  while a ratio of 1 means only errors were predicted.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Our  experiments are not concerned by the last two categories as  we do not use NoteOff tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Finally, we should mention that most of these errors can  be avoided by a ruled-based sampling.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' When predicting a  token, one can easily keep track of the time, notes played  and token types to automatically exclude invalid predictions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Byte Pair Encoding for Symbolic Music  In practice, this can be achieved by setting the invalid indices  of the predicted logits to −∞ before applying softmax.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Feature similarity  We expect models to generate continuations that keep the  features of the input prompt consistent.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For instance, it  should predict first notes within the same scale and with  the same velocity range.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We measure this similarity by  calculating the overlapping area of distributions of features,  for the prompt and the first 16 generated beats.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Previous works (Yang & Lerch, 2020;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Choi et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2020;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Mittal et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' von R¨utte et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2022) use the proba-  bility density function of the distributions, estimated with  kernel density estimations, and emphasizes that it smooths  and transforms the distributions into more general repre-  sentations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' While this method can be suited for continuous  modalities, it can lead to inaccuracies with categorical ones.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Here, pitch and duration features can be considered as dis-  crete.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Their distributions are both sparse, containing for  instance many white keys and fewer black keys, yet adja-  cent and corresponding to close integer values in the MIDI  format.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In order to be more accurate, we measure this simi-  larity with the histogram intersection of these features, as  described in Equation (1).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Similarity (D1, D2) = HI (Hist (D1) , Hist (D2))  HI(x, y) = �  i min(xi, yi), xi ≥ 0, yi ≥ 0  (1)  Hist : R|D| �→ Ne returns the normalized histogram of  a distribution of a feature with e elements, HI stands for  Histogram Intersection.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Human evaluations  For each experiment, we select 40 prompts of 8 beats.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For  each prompts, we generate continuations of 1k tokens with  the benchmarked models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Three musicians open the con-  tinuations as a MIDI file, allowing them to listen the tracks  and also visualize them as piano rolls.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Among the tracks,  they are asked to select the one: 1) with the highest fidelity  on pitch scale, velocity, note density and rhythm, with the  prompt;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 2) they subjectively prefer overall, considering its  correctness, structure and richness.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Results and analysis  We focus on how BPE is learned on the corpuses, then on its  benefits for music generation and composer classification.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' BPE learning  Figure 3 shows the distribution of token types combina-  tions of the learned BPE tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  We observe that the  majority of the combinations learned on the Remi tok-  enization represent notes, by their Pitch, Velocity and  4  10  20  50  100  BPE Factor  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Proportion  Pch-Vel-Dur  Pch-Vel-Dur-TimeShift  Vel-Dur-TimeShift  Vel-Dur  Pch-Vel-Dur-Pch-Vel-Dur  TimeShift-Pch  Other  (a) TSD  4  10  20  50  100  BPE Factor  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  Proportion  Pch-Vel-Dur  Pch-Vel-Dur-Pos  Vel-Dur  Pos-Pch-Vel-Dur  Pch-Vel-Dur-Pch-Vel-Dur  Pos-Pch  Other  (b) Remi  Figure 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Normalized distributions of the token types of the BPE  tokens, per BPE factor for the POP909 dataset.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  0  2k  4k  6k  8k  10k  12k  14k  Vocabulary size  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Avg.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' token combinations  POP909 TSD  POP909 REMI  GiantMIDI TSD  GiantMIDI REMI  0  2k  4k  6k  8k  10k  12k  14k  Vocabulary size  5  10  15  20  25  30  35  Max.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' token combinations  Figure 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Average (left) and maximum (right) number of token  combinations represented by BPE tokens in function of the vocab-  ulary size.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Duration attributes.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For TSD, the combinations also in-  clude TimeShift tokens early in the learning.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' This dif-  ference mostly comes from common TimeShift tokens  following notes, whereas for Remi the notes are distributed  at different Position(s).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' As the vocabulary grows, the  combinations tend to be more diverse.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The distribution for  the GiantMIDI dataset are showned in Appendix C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Figure 4 shows the evolution of the average number of non-  BPE token combinations represented by the BPE tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  At the beginning of the learning, the mean number of com-  binations grows more quickly as the most recurrent token  successions are often made of more than two tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The  POP909 dataset being smaller than GiantMIDI, it naturally  leads to a higher maximum number of combinations as the  latter is more diverse.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' When the vocabulary begins to con-  Byte Pair Encoding for Symbolic Music  Table 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Metrics of generated results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' TSE numbers are all scaled at e-3 for better readability.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Sim stands for similarity, the best results are  the closest to the datasets.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Hum.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Fidelity and Overall are the human evaluations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Data / Strategy  TSEtype (↓)  TSEdupn (↓)  TSEtime (↓)  Sim.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' pit.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Sim.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' vel.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Sim.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' dur.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Hum.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Fidelity (↑)  Hum.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Overall (↑)  POP909 TSD  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='66 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='84 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='69 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='14  No BPE  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  13.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 ± 8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='59 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='82 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='64 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  BPE×4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9  21.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 19.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='65 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='82 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='74 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='24  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='19  BPE×10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  13.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4 ± 14.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='64 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='78 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='74 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='53  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='42  BPE×20  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  12.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 11.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='62 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='79 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='70 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='31  BPE×50  22.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4 ± 24.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4 ± 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='56 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='70 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='62 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×100  21.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 40.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  35.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 ± 56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='54 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='66 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='14  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='63 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  PVm  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1 ± 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='59 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='78 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='73 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06  PVDm  23.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 ± 19.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='43 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='57 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='19  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='54 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  POP909 REMI  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='66 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='84 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='69 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='14  No BPE  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  115.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4 ± 33.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  74.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7 ± 26.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='61 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='85 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='72 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03  BPE×4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  65.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7 ± 21.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  154.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 27.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='55 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='77 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='70 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='27  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='34  BPE×10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  52.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 18.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  167.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1 ± 30.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='49 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='77 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='63 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='52  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='44  BPE×20  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  81.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 37.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  242.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 ± 46.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='46 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='71 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='61 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  BPE×50  37.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 35.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  128.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2 ± 22.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  324.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1 ± 21.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='30 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='56 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='55 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  BPE×100  83.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 78.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  136.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 32.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  324.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 ± 28.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='28 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='54 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='22  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='55 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  PVm  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  160.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0 ± 75.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  102.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7 ± 48.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='60 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='77 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='69 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04  PVDm  49.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 46.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 25.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  301.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 26.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='32 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='50 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='24  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='45 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  CPWord  331.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 33.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  144.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 46.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  99.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 16.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='57 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='85 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='73 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  Octuple 789.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 111.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  891.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 76.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='21  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='17  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  GiantMIDI TSD  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='49 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='17  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='74 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='52 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='23  No BPE  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='50 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='77 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='63 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='24  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='19  BPE×4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  15.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2 ± 18.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='51 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='75 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='62 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='14  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='33  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='27  BPE×10  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  35.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2 ± 45.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='51 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='68 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='17  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='65 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='29  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='37  BPE×20  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  17.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 29.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='52 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='73 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='65 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  BPE×50  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='50 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='70 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='64 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  BPE×100  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='46 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='63 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='17  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='53 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  PVm  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0 ± 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='46 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='69 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='67 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  PVDm  35.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 ± 56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='39 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='61 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='25 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  GiantMIDI REMI  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='49 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='17  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='74 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='52 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='23  No BPE  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 40.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  95.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1 ± 42.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='53 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='75 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='14  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='63 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  BPE×4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  44.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 23.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  82.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 36.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='46 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='71 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='62 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='41  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='43  BPE×10  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  31.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7 ± 20.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  175.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 ± 60.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='43 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='63 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='21  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='54 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='53  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='52  BPE×20  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  36.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 ± 29.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  221.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 66.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='33 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='65 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='16  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='46 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  BPE×50  34.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 11.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  80.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 53.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  316.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4 ± 54.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='36 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='58 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='30 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='23  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  BPE×100  476.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1 ± 148.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  159.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 60.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  285.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 31.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='19 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='59 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='19  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  PVm  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  53.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 47.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  181.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5 ± 56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='46 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='70 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='60 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='14  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  PVDm  31.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 63.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9  65.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 ± 28.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  285.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6 ± 32.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='33 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='14  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='58 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='19  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='29 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='17  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  CPWord  408.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9 ± 28.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  160.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1 ± 54.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  69.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 16.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='51 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='81 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='69 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  Octuple 763.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8 ± 134.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  894.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3 ± 62.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='19  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  tain BPE tokens with a large number of combinations, it  starts to specialize on very specific note successions that  may appear in few data samples.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In particular, big jumps  of maximum number of combinations, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' from 14 to 27  for POP909 Remi, indicate that two already big BPE tokens  represent the most recurrent succession.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' These numbers,  correlated with the model, dataset sizes and overall token  distribution of the dataset, might help to choose an optimal  vocabulary size.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Further analysis in Appendix C shows that BPE consider-  ably reduces the sequence length, and so the training and  generation time, at the cost of an increased tokenization  time.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Tokenization of data is however often performed once,  and the training time gain is very likely to be larger than the  tokenization time loss.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Generated results  For the generation task, we generate continuations of input  prompt from the validation subset.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The continuations are  autoregressively generated with 1024 steps, with nucleus  sampling (Holtzman et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2020), with p = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  The results of all metrics are reported in Table 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For TSD,  BPE allows to reduce both the token type and note dupli-  cation errors in most cases, while the time errors slightly  increase for Remi baselines.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' These results show that models  can easily scale to bigger vocabularies, up to a certain limit.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Here, starting from a BPE factor of 50, the TSE seems to  increase, as do the other results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' BPE tends to however  produce results with features slightly less similar, especially  with big vocabulary sizes.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  We gathered a total of 400 human evaluations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' They show  that BPE with factors of 4 and 10 significantly outperform  other baselines, in all experiments.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' BPE helps models to  Byte Pair Encoding for Symbolic Music  Table 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Number of tokens sampled and not sampled by generative  models, respectively right and left separated by |.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Strategy  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='POP909 TSD  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='POP909 Remi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='GiantMIDI TSD  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='GiantMIDI Remi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='No BPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='116 | 23 (16%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='141 | 11 (7%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='136 | 3 (2%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='151 | 1 (0%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='BPE×4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='454 | 102 (18%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='487 | 121 (19%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='456 | 100 (17%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='386 | 222 (36%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='BPE×10  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='479 | 911 (65%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='514 | 1006 (66%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='456 | 934 (67%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='618 | 902 (59%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='BPE×20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='592 | 2188 (78%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='552 | 2488 (81%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='478 | 2302 (82%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='504 | 2536 (83%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='BPE×50  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='521 | 6429 (92%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='249 | 7351 (96%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='401 | 6549 (94%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='155 | 7445 (97%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='BPE×100  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='521 | 13379 (96%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='244 | 14956 (98%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='281 | 13619 (97%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='89 | 15111 (99%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='321 | 426 (57%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='338 | 422 (55%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='342 | 405 (54%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='369 | 391 (51%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVDm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='391 | 13712 (97%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='144 | 13972 (98%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='252 | 13851 (98%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='166 | 13950 (98%)  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Table 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Average accuracy of classification models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Strategy  TSD (↑)  Remi (↑)  TSD Large (↑)  Remi Large (↑)  No BPE  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='196 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='031  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='169 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='021  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='208 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='033  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='175 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='022  BPE×4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='218 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='033  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='168 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='021  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='226 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='034  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='171 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='022  BPE×10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='226 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='038  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='190 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='030  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='228 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='037  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='201 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='034  BPE×20  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='236 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='038  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='195 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='026  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='240 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='039  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='210 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='029  BPE×50  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='199 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='027  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='207 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='032  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='247 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='041  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='216 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='035  BPE×100  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='122 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='009  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='119 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='008  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='243 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='037  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='126 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='010  PVm  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='199 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='027  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='150 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='016  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='213 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='029  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='188 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='025  PVDm  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='226 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='035  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='192 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='028  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='228 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='036  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='194 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='029  CPWord 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='204 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='28 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='214 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='024  Octuple 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='274 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='041 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='283 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='043  generate more correct and pleasant music.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We make the  assumption that having a larger set of learned embeddings  help the model to capture more easily the global melody, har-  mony and music structure, and in turn improve the generated  results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' These embedding, when well learned contextually,  may represent richer and more explicit information.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Table 2 shows that while models give high probabilities to  more unique tokens with BPE in absolute number, the pro-  portion of sampled tokens decreases.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Models tend to focus  on the sets of more recurrent tokens and omitting more rare  ones.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Beyond a BPE factor of 20 (or vocabulary size be-  tween 2k and 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5k tokens), the models are even focusing on  a more restricting sets of tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' These numbers highlight  the limitations of using a too large vocabulary size, as the  extra effort is unlikely to result in better results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Composer classification  Composer classification is performed with the top-10 most  present composers of the GiantMIDI dataset.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The results,  reported in Table 3, show that BPE outperforms other base-  lines.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Here, the model seems to benefit from larger vocabu-  lary sizes.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We also remark that the model size plays in its  capacity to handle large vocabularies.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' While the results of  BPE100 for the small model indicate it was unable to learn  anything, the larger one performed almost as good as the top  baseline.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' A second observation is the good performances  of embedding pooling strategies (CPWord and OCtuple).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  While they performed poorly for generative tasks, they are  among the best for this classification task.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' They seem to be  better for MIR tasks than generation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' As stated in Section 1,  generation implies sampling, and sampling from several  distributions is delicate, as for training a model with an  autoregressive objective on several output distributions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Table 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' IsoScore results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Generator  POP909 TSD  POP909 Remi  GiantMIDI TSD  GiantMIDI Remi  No BPE  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='14  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  BPE×4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07  BPE×20  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='13  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×50  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  BPE×100  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  PVm  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  PVDm  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  CPWord 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  Octuple 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  Classifier  TSD (↑)  Remi (↑)  TSD Large (↑)  Remi Large (↑)  No BPE  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='74  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='71  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='80  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='77  BPE×4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='35  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='33  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='54  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='37  BPE×10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='36  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='31  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='48  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='50  BPE×20  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='54  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='57  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='64  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='53  BPE×50  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='77  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='80  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='75  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='82  BPE×100  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='82  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='90  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='87  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='89  PVm  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='27  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='27  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='32  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='32  PVDm  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='69  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='88  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='88  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='88  CPWord 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05  Octuple 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Learned embedding spaces  Results presented in this section rely on Table 4, and Fig-  ures 5 and 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Isotropy is a measure of the uniformity of the  space occupied by a distribution, across all dimensions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In  our case, the distribution is a manifold X ∈ RN×d where  N = |V | and d is the model/embedding dimension.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' It  has been associated with improved performances with lan-  guage models (Bi´s et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Liang et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021), mostly  because embeddings are more discriminative and enable  models to capture and distinguish more easily subtle seman-  tic information.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' It has been observed that representations  from Transformers often exhibit anisotropy, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', they tend  to occupy only a small subspace of the embedding space,  and often not uniformly (Gao et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2019;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Ethayarajh, 2019;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Wang et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2020a;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Gong et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2018;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Reif et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2019),  especially causal generative models (Ethayarajh, 2019).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Isotropy is often estimated by different ways: singular value  decomposition (Bi´s et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Gao et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2019;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Liang  et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Wang et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2020a), intrinsic dimension (Cai  et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021), partition function (Arora et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2016;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Mu &  Viswanath, 2018), average cosine similarity (Ethayarajh,  2019).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Although these methods are correlated with isotropy,  recent research shed light on some of their limits (Rudman  et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2022).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We choose to estimate it with intrinsic value,  IsoScore (Rudman et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2022), singular value and cosine  similarity, to have results that corroborate and complement  themselves.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The results of the two latter can be found in  Appendix D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For tokenizations with embedding pooling, we  used 50k randomly sampled embeddings of combinations  of tokens representing notes, as using all the embedding  combinations would be intractable and would not reflect the  ones actually learned by the models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Results for tokeniza-  tions where N ≲ d (no BPE) have to be interpreted loosely.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Isotropy cannot be reliably measured with less samples than  the number of dimensions they occupy.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The estimations are  more accurate when N ≫ d, as more samples populate all  Byte Pair Encoding for Symbolic Music  Gen.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' POP909 TSD  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  0  20  40  60  80  Dimension  lPCA  MLE  MOM  TLE  TwoNN  FisherS  Gen.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' POP909 Remi  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  CPWord  Octuple  0  20  40  60  lPCA  MLE  MOM  TLE  TwoNN  FisherS  Gen.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' GiantMIDI TSD  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  0  5  10  15  20  25  lPCA  MLE  MOM  TLE  TwoNN  FisherS  Gen.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' GiantMIDI Remi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='noBPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe10  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe50  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe100  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVDm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='CPWord  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Octuple  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='40  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='60  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='80  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='lPCA  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='MLE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='MOM  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='TLE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='TwoNN  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='FisherS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Clasmall TSD  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='noBPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe10  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe50  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe100  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVDm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='40  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='60  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='80  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Dimension  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='200  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='300  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='400  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='500  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='600  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='700  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='lPCA  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='MLE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='MOM  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='TLE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='TwoNN  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='FisherS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Clasmall Remi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='noBPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe10  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe50  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe100  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVDm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='CPWord  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Octuple  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='40  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='60  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='80  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='100  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='200  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='300  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='400  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='500  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='600  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='700  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='800  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='lPCA  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='MLE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='MOM  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='TLE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='TwoNN  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='FisherS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Clalarge TSD  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='noBPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe10  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe50  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe100  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVDm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='40  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='60  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='80  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='100  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='200  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='400  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='600  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='800  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1000  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='lPCA  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='MLE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='MOM  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='TLE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='TwoNN  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='FisherS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Clalarge Remi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='noBPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe10  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe50  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='bpe100  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='PVDm  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='CPWord  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Octuple  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='40  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='60  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='80  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='100  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='200  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='400  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='600  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='800  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1000  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='lPCA  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='MLE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='MOM  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='TLE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='TwoNN  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='FisherS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Figure 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Intrinsic dimension estimations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' A second x axis has been added on the right for lPCA on classifier plots for better readability.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Gen.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' POP909 no BPE  Gen.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' POP909 BPE×20  Clasmall GiantMIDI no BPE  Clasmall GiantMIDI BPE×20  Figure 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 3d UMAP representations of learning embedding spaces,  with TSD tokenization.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Abbreviations in legend stand for: Pi:  Pitch;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' V: Velocity;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' D: Duration;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Po: Position: TS: TimeShift.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  dimensions of Rd.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  The IsoScore results (See Table 4), show that BPE does  not increase the score for generative models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' It seems that  big vocabularies with BPE yield lower IsoScore results,  that corroborate with the intrinsic dimension results (Fig-  ure 5).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Causal generative models have been shown to learn  anisotropic embedding representations (Cai et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Ethayarajh, 2019).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Embeddings form cones and clusters,  that can be observed in Figure 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' As we estimated isotropy  on all embeddings altogether, the presence of clusters nat-  urally correlate with anisotropy, as the variance is mostly  pronounced on their distances.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The cluster themselves might  be more isotropic (Cai et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  On the other hand, BPE can help bi-directional models  to learn more isotropic embedding representations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The  IsoScore grows with the vocabulary size, as do the intrinsic  dimension.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Figure 6 we observe that the embeddings  have no preferred direction in space, forming a sphere (See  more figures in Appendix D).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Conclusion  We showed that BPE can increase the quality of results  for symbolic music generation, and composer classifica-  tion, while improving the performances, with a well chosen  vocabulary size.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' BPE can be applied on top of any tokeniza-  tion, and we advice the reader to do so for projects involving  symbolic music.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The drawbacks are a time-consuming vo-  cabulary learning, and a slower tokenization of data.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' BPE  can also helps models to learn more isotropic embedding  representations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Future work will explore more in depth  the isotropy of clusters of embeddings of generative models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  We also plan to experiment with larger model, dataset and  vocabulary sizes, hoping to find guidelines for choosing an  optimum vocabulary size.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Special  Pitch  Velocity  Duration  TimeShift  8  7  6  5  4  3  8  7  6  3  5  2  1  4  0Special  Pitch  Velocity  Duration  Time-Shift  Pi-V-D  7  Pi-V-D-TS  V-D-TS  6  V-D  TS-Pi  5  Other BPE  4  5  4  3  5  6  7  2  8  9  10  1  11  12Special  Pitch  Velocity  Duration  TimeShift  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5Special  Pitch  Velocity  Duration  Time-shift  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Pi-V-D  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  V-D-Pi  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  V-D-TS  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Pi-V-D-TS  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  V-D  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Other BPE  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0Byte Pair Encoding for Symbolic Music  References  Arora, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Li, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Liang, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Ma, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Risteski, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' A Latent  Variable Model Approach to PMI-based Word Embed-  dings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Transactions of the Association for Computational  Linguistics, 4:385–399, 07 2016.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ISSN 2307-387X.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' doi:  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1162/tacl a 00106.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  1162/tacl_a_00106.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Bi´s, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Podkorytov, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Liu, X.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Too much in com-  mon: Shifting of embeddings in transformer language  models and its implications.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Proceedings of the  2021 Conference of the North American Chapter of  the Association for Computational Linguistics: Human  Language Technologies, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 5117–5130, Online, June  2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Association for Computational Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' doi:  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18653/v1/2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='naacl-main.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='403.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://  aclanthology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='naacl-main.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='403.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Cai, X.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Huang, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Bian, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Church, K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Isotropy in the  contextual embedding space: Clusters and manifolds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In  International Conference on Learning Representations,  2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://openreview.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='net/forum?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  id=xYGNO86OWDH.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Choi, K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Hawthorne, C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Simon, I.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Dinculescu, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and  Engel, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Encoding musical style with transformer au-  toencoders.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In III, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' and Singh, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' (eds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ), Pro-  ceedings of the 37th International Conference on Ma-  chine Learning, volume 119 of Proceedings of Ma-  chine Learning Research, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 1899–1908.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' PMLR, Jul  2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://proceedings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='mlr.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='press/  v119/choi20b.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='html.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Devlin, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Chang, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-W.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Lee, K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Toutanova, K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' BERT:  Pre-training of deep bidirectional transformers for lan-  guage understanding.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Proceedings of the 2019 Confer-  ence of the North American Chapter of the Association for  Computational Linguistics: Human Language Technolo-  gies, Volume 1 (Long and Short Papers), pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 4171–4186,  Minneapolis, Minnesota, June 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Association for  Computational Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18653/v1/N19-1423.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://aclanthology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/N19-1423.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Donahue, C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Mao, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Li, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' E.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Cottrell, G.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' W.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and  McAuley, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Lakhnes: Improving multi-instrumental  music generation with cross-domain pre-training.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In  Proceedings of the 20th International Society for Mu-  sic Information Retrieval Conference, ISMIR 2019,  Delft, The Netherlands, November 4-8, 2019, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 685–  692, 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL http://archives.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='ismir.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='net/  ismir2019/paper/000083.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='pdf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Dong, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-W.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Hsiao, W.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Yang, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Yang,  Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Musegan:  Multi-track sequential genera-  tive adversarial networks for symbolic music gen-  eration and accompaniment.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Proceedings of the  AAAI Conference on Artificial Intelligence, 32(1), Apr.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://ojs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='aaai.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/index.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  php/AAAI/article/view/11312.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Ethayarajh, K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' How contextual are contextualized word rep-  resentations?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Comparing the geometry of BERT, ELMo,  and GPT-2 embeddings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Proceedings of the 2019  Conference on Empirical Methods in Natural Language  Processing and the 9th International Joint Conference on  Natural Language Processing (EMNLP-IJCNLP), pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 55–  65, Hong Kong, China, November 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Association for  Computational Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18653/v1/D19-1006.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://aclanthology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/D19-1006.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Fradet, N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Briot, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-P.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Chhel, F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', El Fallah Seghrouchni,  A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Gutowski, N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Miditok: A python package  for midi file tokenization.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Extended Abstracts for  the Late-Breaking Demo Session of the 22nd Interna-  tional Society for Music Information Retrieval Con-  ference, 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://archives.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='ismir.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  net/ismir2021/latebreaking/000005.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='pdf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Gage, P.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  A new algorithm for data compression.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  C Users J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 12(2):23–38, feb 1994.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  ISSN 0898-  9788.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL  https://dl.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='acm.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/doi/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  5555/177910.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='177914.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Gao, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', He, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Tan, X.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Qin, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Wang, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Liu, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Representation degeneration problem in training natural  language generation models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In International Conference  on Learning Representations, 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://  openreview.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='net/forum?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='id=SkEYojRqtm.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Gong, C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', He, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Tan, X.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Qin, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Wang, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Liu, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Frage: Frequency-agnostic word representation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Ben-  gio, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Wallach, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Larochelle, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Grauman, K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Cesa-  Bianchi, N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Garnett, R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' (eds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ), Advances in Neural  Information Processing Systems, volume 31.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Curran As-  sociates, Inc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://proceedings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  neurips.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='cc/paper/2018/file/  e555ebe0ce426f7f9b2bef0706315e0c-Paper.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  pdf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Hadjeres, G.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' and Crestel, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The piano inpainting applica-  tion, 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://arxiv.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/abs/2107.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  05944.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Hadjeres, G.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Pachet, F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Nielsen, F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  DeepBach:  a steerable model for Bach chorales generation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In  Precup, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' and Teh, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' W.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' (eds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ), Proceedings of  the 34th International Conference on Machine Learn-  ing, volume 70 of Proceedings of Machine Learning  Research, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 1362–1371.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' PMLR, 06–11 Aug 2017.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://proceedings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='mlr.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='press/v70/  hadjeres17a.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='html.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Byte Pair Encoding for Symbolic Music  Heusel, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Ramsauer, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Unterthiner, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Nessler, B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and  Hochreiter, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Gans trained by a two time-scale update  rule converge to a local nash equilibrium.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Guyon, I.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  Luxburg, U.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' V.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Bengio, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Wallach, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Fergus, R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Vish-  wanathan, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Garnett, R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' (eds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ), Advances in Neural  Information Processing Systems, volume 30.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Curran As-  sociates, Inc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2017.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://proceedings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  neurips.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='cc/paper/2017/file/  8a1d694707eb0fefe65871369074926d-Paper.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  pdf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Holtzman, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Buys, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Du, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Forbes, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Choi,  Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  The curious case of neural text degeneration.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In  International Conference on Learning Representations,  2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://openreview.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='net/forum?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  id=rygGQyrFvH.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Hsiao, W.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Liu, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Yeh, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Yang, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Com-  pound word transformer: Learning to compose full-  song music over dynamic directed hypergraphs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Pro-  ceedings of the AAAI Conference on Artificial Intel-  ligence, 35(1):178–186, May 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  doi:  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1609/  aaai.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='v35i1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='16091.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://ojs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='aaai.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/  index.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='php/AAAI/article/view/16091.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Huang, C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Z.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Vaswani, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Uszkoreit, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Shazeer, N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  Simon, I.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Hawthorne, C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Dai, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Hoffman, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  Dinculescu, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Eck, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Music transformer, 2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://arxiv.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/abs/1809.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04281.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Huang, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' and Yang, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Pop music transformer:  Beat-based modeling and generation of expressive pop  piano compositions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Proceedings of the 28th ACM  International Conference on Multimedia, MM ’20, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  1180–1188, New York, NY, USA, 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Association for  Computing Machinery.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ISBN 9781450379885.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  1145/3394171.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3413671.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1145/3394171.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3413671.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Kermarec, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Bigo, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Keller, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Improving tokeniza-  tion expressiveness with pitch intervals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Extended Ab-  stracts for the Late-Breaking Demo Session of the 23nd In-  ternational Society for Music Information Retrieval Con-  ference, 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://ismir2022program.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  ismir.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='net/lbd_369.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='html.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Kilgour,  K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  Zuluaga,  M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  Roblek,  D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  and Shar-  ifi,  M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Fr´echet audio distance:  A reference-  free  metric  for  evaluating  music  enhancement  algorithms.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Interspeech,  pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 2350–2354,  09  2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='21437/Interspeech.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2019-2219.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL  https://www.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='isca-speech.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/archive_  v0/Interspeech_2019/pdfs/2219.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='pdf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Kingma, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' P.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' and Ba, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Adam: A method for stochastic op-  timization.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In 3rd International Conference on Learning  Representations, ICLR 2015, San Diego, CA, USA, May  7-9, 2015, Conference Track Proceedings, 2015.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL  http://arxiv.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/abs/1412.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6980.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Kong, Q.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Li, B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Chen, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Wang, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Giantmidi-  piano:  A large-scale midi dataset for classical  piano  music.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In  Transactions  of  the  Interna-  tional Society for Music Information Retrieval, vol-  ume 5, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 87–98, 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  doi:  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5334/tismir:  80.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://transactions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='ismir.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='net/  articles/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5334/tismir.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='80/#.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Liang, F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Gotham, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Johnson, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Shotton,  J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Automatic stylistic composition of bach chorales  with deep LSTM.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Proceedings of the 18th In-  ternational Society for Music Information Retrieval  Conference, ISMIR 2017, Suzhou, China, October  23-27, 2017, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 449–456, 2017.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https:  //ismir2017.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='smcnus.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/wp-content/  uploads/2017/10/156_Paper.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='pdf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Liang, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Cao, R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Zheng, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Ren, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Gao, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Learning  to remove: Towards isotropic pre-trained bert embedding.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Artificial Neural Networks and Machine Learning –  ICANN 2021: 30th International Conference on Artificial  Neural Networks, Bratislava, Slovakia, September 14–17,  2021, Proceedings, Part V, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 448–459, Berlin, Heidel-  berg, 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Springer-Verlag.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ISBN 978-3-030-86382-1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1007/978-3-030-86383-8 36.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://  doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1007/978-3-030-86383-8_36.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Lin, C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  ROUGE: A package for automatic evalua-  tion of summaries.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Text Summarization Branches  Out, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 74–81, Barcelona, Spain, July 2004.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Asso-  ciation for Computational Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https:  //aclanthology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/W04-1013.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Liu, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Dong, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Cheng, Z.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Zhang, X.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Li, X.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Yu, F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  and Sun, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Symphony generation with permutation  invariant language model.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Proceedings of the 21rd  International Society for Music Information Retrieval  Conference, Bengalore, India, December 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ISMIR.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://arxiv.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/abs/2205.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05448.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  McInnes, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Healy, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Saul, N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Grossberger, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Umap: Uniform manifold approximation and projec-  tion.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The Journal of Open Source Software, 3(29):861,  2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://joss.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='theoj.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/papers/  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='21105/joss.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00861.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Micikevicius, P.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Narang, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Alben, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Diamos, G.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Elsen,  E.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Garcia, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Ginsburg, B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Houston, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Kuchaiev, O.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  Venkatesh, G.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Wu, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Mixed precision training.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In  International Conference on Learning Representations,  2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://openreview.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='net/forum?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  id=r1gs9JgRZ.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Byte Pair Encoding for Symbolic Music  Mittal, G.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Engel, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Hawthorne, C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Simon,  I.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Symbolic music generation with diffusion mod-  els.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Proceedings of the 22nd International Soci-  ety for Music Information Retrieval Conference, IS-  MIR 2021, Online, November 7-12, 2021, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 468–475,  2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://archives.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='ismir.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='net/  ismir2021/paper/000058.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='pdf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Mu, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' and Viswanath, P.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  All-but-the-top: Simple and  effective postprocessing for word representations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In  International Conference on Learning Representations,  2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://openreview.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='net/forum?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  id=HkuGJ3kCb.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Oore, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Simon, I.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Dieleman, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Eck, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Si-  monyan, K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  This time with feeling:  Learning  expressive musical performance.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Neural Comput-  ing and Applications,  32:955–967,  2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL  https://link.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='springer.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='com/article/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  1007/s00521-018-3758-9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Papineni, K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Roukos, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Ward, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Zhu, W.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Bleu:  a method for automatic evaluation of machine transla-  tion.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Proceedings of the 40th Annual Meeting of  the Association for Computational Linguistics, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 311–  318, Philadelphia, Pennsylvania, USA, July 2002.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' As-  sociation for Computational Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3115/  1073083.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1073135.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://aclanthology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  org/P02-1040.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Payne, C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Musenet, 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://openai.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  com/blog/musenet.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Press, O.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' and Wolf, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Using the output embedding to  improve language models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Proceedings of the 15th  Conference of the European Chapter of the Associa-  tion for Computational Linguistics: Volume 2, Short  Papers, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 157–163, Valencia, Spain, April 2017.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' As-  sociation for Computational Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https:  //aclanthology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/E17-2025.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Radford, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Wu, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Child, R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Luan, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Amodei, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and  Sutskever, I.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Language models are unsupervised multitask  learners, 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://openai.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='com/blog/  better-language-models/.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Reif, E.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Yuan, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Wattenberg, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Viegas, F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Co-  enen, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Pearce, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Kim, B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Visualizing and  measuring the geometry of bert.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Wallach, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  Larochelle, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Beygelzimer, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=", d'Alch´e-Buc, F." metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  Fox, E.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Garnett, R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' (eds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ), Advances in Neural  Information Processing Systems, volume 32.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Curran As-  sociates, Inc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://proceedings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  neurips.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='cc/paper/2019/file/  159c1ffe5b61b41b3c4d8f4c2150f6c4-Paper.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  pdf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Ren, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', He, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Tan, X.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Qin, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Zhao, Z.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Liu, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-  Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Popmag: Pop music accompaniment generation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Proceedings of the 28th ACM International Confer-  ence on Multimedia, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 1198–1206.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Association for  Computing Machinery, 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ISBN 9781450379885.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1145/3394171.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3413721.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1145/3394171.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3413721.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Rudman, W.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Gillman, N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Rayne, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Eickhoff,  C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' IsoScore: Measuring the uniformity of embedding  space utilization.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  In Findings of the Association for  Computational Linguistics: ACL 2022, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 3325–3339,  Dublin, Ireland, May 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Association for Computa-  tional Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18653/v1/2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='findings-acl.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  262.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://aclanthology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  findings-acl.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='262.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Sennrich, R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Haddow, B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Birch, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Neural machine  translation of rare words with subword units.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Pro-  ceedings of the 54th Annual Meeting of the Association  for Computational Linguistics (Volume 1: Long Papers),  pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' 1715–1725, Berlin, Germany, August 2016.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Associ-  ation for Computational Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18653/v1/  P16-1162.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://aclanthology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/  P16-1162.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Sturm, B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Santos, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Korshunova, I.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Folk mu-  sic style modelling by recurrent neural networks with  long short-term memory units.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Extended abstracts  for the Late-Breaking Demo Session of the 16th Inter-  national Society for Music Information Retrieval Con-  ference, 2015.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://ismir2015.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='ismir.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  net/LBD/LBD13.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='pdf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Tay, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Dehghani, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Abnar, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Shen, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Bahri, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Pham,  P.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Rao, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Yang, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Ruder, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Metzler, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Long  range arena : A benchmark for efficient transformers.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In  International Conference on Learning Representations,  2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://openreview.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='net/forum?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  id=qVyeW-grC2k.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Vaswani, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Shazeer, N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Parmar, N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Uszkoreit, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  Jones, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Gomez, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' N.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Kaiser, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' u.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Polosukhin,  I.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Attention is all you need.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Guyon, I.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Luxburg,  U.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' V.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Bengio, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Wallach, H.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Fergus, R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Vish-  wanathan, S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Garnett, R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' (eds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ), Advances in Neural  Information Processing Systems, volume 30.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Curran As-  sociates, Inc.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2017.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://proceedings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  neurips.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='cc/paper/2017/file/  3f5ee243547dee91fbd053c1c4a845aa-Paper.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  pdf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  von R¨utte, D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Biggio, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Kilcher, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Hoffman, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Figaro: Generating symbolic music with fine-grained  artistic control, 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://arxiv.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/  abs/2201.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10936.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Byte Pair Encoding for Symbolic Music  Wang, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Huang, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Huang, K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Hu, Z.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Wang, G.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Gu,  Q.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Improving neural language generation with spectrum  control.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In International Conference on Learning Rep-  resentations, 2020a.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://openreview.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  net/forum?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='id=ByxY8CNtvr.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Wang, Z.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Chen, K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Jiang, J.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Zhang, Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Xu, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Dai,  S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Bin, G.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Xia, G.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Pop909: A pop-song dataset  for music arrangement generation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Proceedings of  21st International Conference on Music Information Re-  trieval, ISMIR, 2020b.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://arxiv.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/  abs/2008.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07142.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Yang, L.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' and Lerch, A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' On the evaluation of gener-  ative models in music.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Neural Comput.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Appl.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 32(9):  4773–4784, may 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' ISSN 0941-0643.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1007/  s00521-018-3849-7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' URL https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  1007/s00521-018-3849-7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Zeng, M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Tan, X.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Wang, R.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Ju, Z.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Qin, T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', and Liu,  T.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='-Y.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' MusicBERT: Symbolic music understanding with  large-scale pre-training.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' In Findings of the Association  for Computational Linguistics: ACL-IJCNLP 2021, pp.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  791–800, Online, August 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Association for Compu-  tational Linguistics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' doi: 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18653/v1/2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='findings-acl.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  70.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  URL https://aclanthology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='org/2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  findings-acl.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='70.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Byte Pair Encoding for Symbolic Music  A.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Model and training  Table 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Model configurations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The number of parameters is based on the baseline with no BPE, and may vary depending on the baseline  with the size of the first and last layers.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Gen stands for generator and Cla for classifier.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Gen  Clasmall  Clalarge  Dimension  512  768  1024  Nb attention heads  8  12  16  Nb layers  10  10  18  Feedforward size  2048  2048  3078  Parameters  32.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6M  58.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0M  193.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3M  The sizes of the models are reported in Table 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The generator is trained with a teacher forcing objective on 100k steps.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The  classifier pre-trained on 60k steps to retrieve the value of randomized positions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Between 1 to 15% of each input sequences  is randomized during pre-training.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' It is then fine-tuned on 100k steps to predict the composer of the input sequence, from  the first output hidden state, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', the BOS position, which is projected through an output classification layer.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The input  embedding and output pre-training module weights are tied to improve the performances (Press & Wolf, 2017).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  The batch size is set to 16 for the generator, and 24 for the classifier.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' All trainings are done with automatic mixed-precision  (Micikevicius et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2018), the Adam optimizer (Kingma & Ba, 2015) with β1 = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9, β2 = 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='999 and ϵ = 10−8, and  dropout, weight decay and a gradient clip norm of respectively 10−1, 10−2 and 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We use a one cycle learning rate scheduler:  the initial learning rate is close to 0 and gradually grows for the 30% first steps to 5e−6, 1e−6 and 5e−7 for the generators,  classifier pre-training and classifier fine-tuning respectively, then slowly decreases down to 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We perform 5 validations steps  every 30 training steps, and compute their average accuracy and loss.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The model parameters are saved when the validation  loss is the lowest ever observed, and after training the last version saved is used for testing.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The training is stopped early if  the validation losses did not decrease for 15k steps and 25k steps for respectively the generator and classifier.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Data downsampling  0  1  2  3  4  5  6  7  duration  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  density  Dataset  POP909  GiantMIDI  0  20  40  60  80  100  120  velocity  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='000  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='005  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='010  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='015  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='020  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='025  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='030  density  Dataset  POP909  GiantMIDI  Figure 7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Distributions of the note durations and velocities of the POP909 and GiantMIDI datasets.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The duration axis is limited to 7 beats.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Figure 7 shows the distributions of velocity and duration values of the notes from the two datasets we use.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' As there is a  larger proportion of low note durations (below two beats), we decided to downsample the Duration and TimeShift  tokens with different resolutions: those up to one beat are downsampled to 8 samples per beat (spb), those from one to  two beats to 4 spb, those from two to four beats to 2 spb, and those from four to eight beats to 1 spb.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' This way, short  notes are represented more precisely than longer ones, reducing the vocabulary size.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' For Remi, Position tokens are  downsampled to 8 spb, resulting in 32 different tokens as we only consider the 4/* time signature.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' This allows to represent  the 16th note.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We only consider pitches within the recommended range for piano (program 0) specified in the General MIDI  2 specifications2: 21 to 108.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We then deduplicate all duplicated notes.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Velocities are downsampled to 8 distinct values.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' No  additional token (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', Chord, Tempo) is used.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  2Available on the MIDI Manufacturers Association website  Byte Pair Encoding for Symbolic Music  C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' BPE Learning  Table 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Vocabulary sizes, mean tokens per beat (tpb), and variation of tpb from without BPE, average tokenizing time and detokenizing  time.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' A maximum of 1000k randomly sampled MIDI files were used for each row.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Vocabulary sizes for CPWord and Octuple are the  product of the sizes of their ”sub-vocabularies”, or in other words the number of possible token combinations, and are rounded for better  readability.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Tokenizing and detokenizing times were run on an Intel Xeon Gold 5128 CPU.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Data  Vocab.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' size  tpb  tpb variation (%)  Tok.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' time (sec)  Detok.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' time (sec)  POP909 TSD  No BPE  139  17.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='81 ± 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×4  556  9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='71 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  45.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='50  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×10  1390  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='75  54.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='80  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='44 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×20  2780  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='95 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='53  60.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='99  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='77 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×50  6950  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='84 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='28  67.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='59 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='37  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×100  13.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9k  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='33 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='16  70.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='72 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='63  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  PVm  747  12.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='72 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='92  28.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='59  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  PVDm  14.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1k  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='63 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='73  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='17  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  POP909 Remi  No BPE  152  18.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06 ± 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  BPE×4  608  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='55 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='26  41.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='61  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='21 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×10  1520  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='85 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='90  51.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='00  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='47 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×20  3040  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='74  55.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='64  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='86 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='19  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×50  7600  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='32 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='58  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='46  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='97 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='43  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  BPE×100  15.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2k  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='70 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='43  62.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='92  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='64 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='79  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  PVm  760  12.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='97 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='92  28.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='19  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  PVDm  14k  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='88 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='73  56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='38  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  CPWord  49k  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='88 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='73  56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='38  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  Octuple  161k  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='21  71.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='81  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='01  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02  GiantMIDI TSD  No BPE  139  15.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='64 ± 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='29 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05  BPE×4  556  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='87 ± 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='30  43.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='26  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='45 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='57  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='16  BPE×10  1390  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='88 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='86  49.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='64  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='29  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='16  BPE×20  2780  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='40  54.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='98  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='90 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='34  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  BPE×50  6950  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='94 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  62.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='11 ± 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='14  BPE×100  13.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9k  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='45 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04  65.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='49 ± 9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='16  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='14  PVm  747  11.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='26 ± 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='46  28.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04  PVDm  14.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1k  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='57 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='59  57.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='98  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03  GiantMIDI Remi  no BPE  152  15.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='89 ± 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='42 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05  BPE×4  608  9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='58 ± 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='39  39.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='70  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='53 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='67  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18  BPE×10  1520  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='96  48.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='51  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='22 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='51  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='17  BPE×20  3040  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='22 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='78  54.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='56  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='18 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='70  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='17  BPE×50  7600  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='41 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='42  59.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='67  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='87 ± 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='98  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='16  BPE×100  15.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2k  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='96 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='20  62.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='51  9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08 ± 11.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='12  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='15  PVm  760  11.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='30 ± 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='66  28.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='90  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='08  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='04  PVDm  14k  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='94 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='63  56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='29  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='02 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='03  CPWord  49k  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='90 ± 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='55  56.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='60  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='09  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10  Octuple  161k  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='37 ± 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='88  72.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='51  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='05 ± 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='07  4  10  20  50  100  BPE Factor  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='1  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Proportion  Vel-Dur-TimeShift  Pch-Vel-Dur  Pch-Vel-Dur-TimeShift  Vel-Dur-Pch  Vel-Dur  Pch-Vel-Dur-Pch  Other  (a) TSD  4  10  20  50  100  BPE Factor  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  Proportion  Pch-Vel-Dur  Pch-Vel-Dur-Pos  Vel-Dur  Pch-Vel-Dur-Pch-Vel-Dur  Pos-Pch-Vel-Dur  Pos-Pch  Other  (b) Remi  Figure 8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Normalized distributions of token types per BPE factor for the GiantMIDI dataset.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Table 6 shows the vocabulary sizes, sequence length variation and tokenization times of all baselines.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' When learning BPE,  the average number of tokens per beat (tpb) quickly decreases, so the sequence length.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' As the vocabulary grows, the tpb  decreases more slowly, as the most recurrent token successions have already be learned and replaced.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' A lower tpb allows to  generate faster.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Byte Pair Encoding for Symbolic Music  The tradeoff of BPE, besides the vocabulary learning time, is the tokenization time, as a MIDI file is first tokenized without  BPE, then BPE is applied by finding the token subsequences to be replaced by the BPE tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The decoding step time, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=',  the time of the conversion of tokens to a MIDI file, is almost not impacted by BPE.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The tokenization and detokenization  times have been gotten with MidiTok (Fradet et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2021) which is implemented in Python.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The tokenization time could be  decreased if performed by a faster compiled language such as Rust or C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The Figure 8 complements the Figure 3, with the  GiantMIDI dataset.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  D.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Learned embedding space  Figure 9 shows the singular values for the generative and classification models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' As the different tokenizations features  vocabularies with very different sizes, the values are normalized for better readability.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Note that the NoBPE tokenizations  feature vocabularies with a size inferior to the embedding dimension.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' NoBPE adj.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' corresponds to the NoBPE results adjusted  to cover the x-axis on the whole embedding size.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Figure 10 shows the pairwise cosine similarity of the learned embedding vectors, for the TSD and Remi representation on  the POP909 dataset.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The first tokens up to 90 are Pitches, followed by Velocities up to 125, Durations up to 160  and then Time-Shift or Position.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Without BPE, we can clearly distinguish patterns in the cosine similarity matrices.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  These high similarities shows that embeddings are close to each other.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' With BPE and larger vocabulary sizes, the average  cosine similarity tend to decrease, especially between BPE tokens.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Embeddings are less similar and more discriminative.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  UMAP (McInnes et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=', 2018) representations shown in Figure 6, Figure 12 and Figure 11 have been calculated with the  default parameters of the official Python package.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' We clearly see that generative models learn clusters of embeddings of the  same type, distant from each other.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The embeddings do not occupy the space uniformly.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' On the other hand, pre-trained  bi-directional models learn more isotropic embedding representations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' The embeddings are spread uniformly across all  directions, for all token types.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Byte Pair Encoding for Symbolic Music  Gen POP909  100  101  102  Dimension  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Singular value  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  noBPE adj.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  100  101  102  Dimension  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Singular value  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  CPWord  Octuple  noBPE adj.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Gen GiantMIDI  100  101  102  Dimension  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Singular value  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  noBPE adj.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  100  101  102  Dimension  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Singular value  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  CPWord  Octuple  noBPE adj.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Clasmall  100  101  102  Dimension  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Singular value  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  noBPE adj.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  100  101  102  Dimension  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Singular value  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  CPWord  Octuple  noBPE adj.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Clalarge  100  101  102  103  Dimension  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Singular value  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  noBPE adj.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  TSD  100  101  102  103  Dimension  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Singular value  noBPE  bpe4  bpe10  bpe20  bpe50  bpe100  PVm  PVDm  CPWord  Octuple  noBPE adj.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Remi  Figure 9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Normalized singular values of embedding matrices of classifier models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Byte Pair Encoding for Symbolic Music  TSD  0  20  40  60  80  100  120  0  20  40  60  80  100  120  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0  100  200  300  400  500  0  100  200  300  400  500  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0  200  400  600  800  1000  1200  0  200  400  600  800  1000  1200  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0  500  1000  1500  2000  2500  0  500  1000  1500  2000  2500  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Remi  0  20  40  60  80  100  120  140  0  20  40  60  80  100  120  140  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  No BPE  0  100  200  300  400  500  600  0  100  200  300  400  500  600  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  BPE x4  0  200  400  600  800  1000  1200  1400  0  200  400  600  800  1000  1200  1400  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  BPE x10  0  500  1000  1500  2000  2500  3000  0  500  1000  1500  2000  2500  3000  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  BPE x20  Figure 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Pairwise cosine similarity matrix of learned embedding of the generative models, on the POP909 dataset.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  No BPE  BPE x4  BPE x10  BPE x20  BPE x50  BPE x100  PVm  PVDm  Figure 11.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' UMAP 2d representations of the embeddings of classifier models pre-trained with the GiantMIDI dataset and TSD tokenization.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Abbreviations in legend stand for: Pi: Pitch;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' V: Velocity;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' D: Duration;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Po: Position;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' TS: TimeShift.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Special  Pitch  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Velocity  Duration  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  TimeShift  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  9.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='06  5  Special  Pitch  Velocity  ！' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='C  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Duration  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Time-Shift  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D-TS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D-Pi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Other BPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='10Special  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pitch  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Velocity  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Duration  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Time-Shift  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D-TS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D-Pi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D-TS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Other BPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7Special  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pitch  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Velocity  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Duration  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Time-Shift  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D-Pi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D-TS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D-TS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Other BPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='2  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6Special  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pitch  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Velocity  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='9  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Duration  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Time-Shift  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D-TS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D-Pi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D-TS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D-Pi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Other BPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='97  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Special  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pitch  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Velocity  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Duration  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Time-Shift  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D-TS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D-Pi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D-Pi  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='3  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='V-D-TS  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Other BPE  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='4  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='6  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='7  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='8  ' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='97.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Special  PitchVel  3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Duration  Time-Shift  3  4  5  63  2  0  1  Special  PitchVelDur  2  Time-Shift  6  7  8  9  10  11  1284009 0400344Byte Pair Encoding for Symbolic Music  TSD No BPE  TSD BPE×4  TSD BPE×10  TSD BPE×20  TSD BPE×50  TSD BPE×100  TSD PVm  TSD PVDm  Remi No BPE  Remi BPE×4  Remi BPE×10  Remi BPE×20  Remi BPE×50  Remi BPE×100  Remi PVm  Remi PVDm  Figure 12.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' UMAP 3d representations of the embeddings of generative models with the POP909 dataset.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Abbreviations in legend stand for:  Pi: Pitch;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' V: Velocity;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' D: Duration;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content=' Po: Position: TS: TimeShift.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='  Special  Pitch  Velocity  Duration  TimeShift  8  7  6  5  4  3  8  7  6  3  5  2  1  4  0Special  Pitch  Velocity  Duration  Time-Shift  Pi-V-D  7  Pi-V-D-TS  V-D-TS  6  V-D  TS-Pi  5  Other BPE  4  5  4  3  5  6  7  2  8  9  10  1  11  128  7  6  5  Special  4  Pitch  3  Velocity  2  Duration  Time-Shift  Pi-V-D  6  V-D-TS  4  V-D  O Pi-V-D-TS  2  5  D-TS:  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Other BPE  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  12.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  2Special  Pitch  Velocity  Duration  Time-Shift  14  Pi-V-D  13  Pi-V-D-TS  12  V-D-TS  11  V-D  10  D-TS  9  Other BPE  8  7  1  0  1  9  2  8  3  7  6  4  58  6  4  Special  2  Pitch  Velocity  0  Duration  Time-Shift  Pi-V-D-TS  12.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Pi-V-D  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  V-D-TS  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  5 Pi-V-D-Pi-V-D  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  V-D  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Other BPE  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='58  6  4  2  Special  Pitch  0  Velocity  2  Duration  4  Time-Shift  Pi-V-D-TS  12.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Pi-V-D  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Pi-V-D-Pi-V-D  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  V-D-TS  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Pi-V-D-TS-Pi-V-D-TS  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  8  Other BPE  10  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  12Special  PitchVel  Duration  Time-Shift  3  2  1  0  1  2  8  6  7  8  9  4  10  11  2Special  PitchVeiDur  Time-Shift  8  7  6  5  4  3  2  10  8  6  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  4  2  0  2Special  Bar  Pitch  Velocity  Duration  2  Position  1  0  1  2  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  1  6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  3  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='010  8  6  Special  Bar  4  Pitch  Velocity  Duration  Position  Pi-V-D  V-D  0  5  Bar-Po  10  Other BPE  15Special  Bar  Pitch  Velocity  Duration  Position  3  Pi-V-D  4  V-D  V-D-Po  5  Bar-Po  6  V-D-Bar-Po  Other BPE  7  7  8  7  6  9  5  4  3  10  28  6  Special  4  Bar  Pitch  2  Velocity  Duration  0  Position  19  Pi-V-D  V-D  18  Po-Pi  4  V-D-Po  6  Pi-V-D-Pi-V-D  8  Other BPE  10  15  1212  10  8  Special  6  Bar  4  Pitch  2  Velocity  0  Duration  Position  Pi-V-D  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Pi-V-D-Po  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Po-Pi-V-D  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='Pi-V-D-Pi-V-D  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  Po-Pi  5  0  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  Other BPE  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  12.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5Special  Bar  Pitch  Velocity  Duration  10  Position  Pi-V-D-Po  8  Pi-V-D  6  Po-Pi-V-D  4  Pi-V-D-Pi-V-D  Po-Pi  2  Other BPE  0  10  8  6  4  0  2  2  4  6  0  8  2  10  12Special  Bar  PitchVel  Duration  Position  11  10  9  8  7  6  5  14  13  12  14  12  16  18  11  20  22Special  Bar  PitchVelDur  Position  8  6  4  2  0  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0  7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='5  10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}
+page_content='0' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/99FLT4oBgHgl3EQfCS7y/content/2301.11975v1.pdf'}