Delete configs

configs/alpaca.json

@@ -1,6 +0,0 @@
-{
-    "description": "Template used by Alpaca-LoRA.",
-    "prompt_input": "Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.\n\n### Instruction:\n{instruction}\n\n### Input:\n{input}\n\n### Response:\n",
-    "prompt_no_input": "Below is an instruction that describes a task. Write a response that appropriately completes the request.\n\n### Instruction:\n{instruction}\n\n### Response:\n",
-    "response_split": "### Response:"    
-}

configs/cleaning_config.yaml

@@ -1,203 +0,0 @@
-patterns_and_insertions:
-    [
-        {
-            "pattern" : '(?:\\figcomp{\s*)(?P<first>.*?)\s*}\s*{\s*(?P<second>.*?)\s*}\s*{\s*(?P<third>.*?)\s*}',
-            "insertion" : '\parbox[c]{{ {second} \linewidth}} {{ \includegraphics[width= {third} \linewidth]{{figures/{first} }} }}',
-            "description" : "Replace figcomp",
-        },
-    ]
-
-verbose: False
-
-commands_to_delete: [
-    'footnote',
-    'footnote ',
-    'crdata',
-    'appendixhead',
-    'selectlanguage',
-    'name',
-    'expandafter',
-    'copyrightyear',
-    'acmYear',
-    'acmBooktitle',
-    'acmPrice',
-    'authorcontributions',
-    'thanksref',
-    'funding',
-    'conflictsofinterest',
-    'externalbibliography',
-    'acmDOI',
-    'acmISBN',
-    'acmConference',
-    'titleheader',
-    'affil',
-    'authorrunning',
-    'pagenumbering',
-    'enlargethispage',
-    'author',
-    'AuthorNames',
-    'author\n',
-    'Author\n',
-    'Author',
-    'fntext',
-    'icmlauthor',
-    'icmlauthor\n',
-    'icmladdress',
-    'icmladdress\n',
-    'received',
-    'runninghead',
-    'bstctlcite',
-    'slugger',
-    'tocauthor',
-    'author\*',
-    'vspace\*',
-    '\write18',
-    'hspace\*',
-    'vspace',
-    'hspace',
-    'maketitle',
-    'institute',
-    'label',
-    'urlstyle',
-    'acks',
-    'tnoteref',
-    'Appendix',
-    'urlstyle',
-    'url',
-    'editor',
-    'ccsdesc',
-    'cortext',
-    'bibliography',
-    'permission',
-    'usetikzlibrary',
-    'thanks',
-    'thispagestyle',
-    'abovedisplayskip',
-    'belowdisplayskip',
-    'bibliographystyle',
-    'IfSubStringInString',
-    'hyphenation',
-    'theoremstyle',
-    'colorbox',
-    'textcolor',
-    'color',
-    'caption',
-    'thlabel',
-    'fancyfoot',
-    'captionof',
-    'settopmatter',
-    'IEEEtriggeratref',
-    'IEEEauthorblockN',
-    'IEEEauthorblockA',
-    'IEEEauthorblockN\n',
-    'IEEEauthorblockA\n',
-    'IEEEauthorrefmark',
-    'orcid',
-    'typeout',
-    'fancyhead',
-    'pagestyle',
-    'biboptions',
-    'affiliation',
-    'address',
-    'institution',
-    'printalgoIEEE',
-    'date',
-    'authornote',
-    'numberofauthors',
-    'footnotetext',
-    'email',
-    'reftitle',
-    'setcopyright',
-    'ead',
-    'deleted',
-    'includegraphics',
-    'comment',
-    'abstract',
-    'replaced',
-    'xspace',
-]
-
-commands_only_to_delete: [
-    'titlerunning',
-    'runningtitle',
-    'title',
-    'title\*',
-    'accept',
-    'added',
-    'icmltitle',
-    'textsuperscript',
-    'texttt',
-    'textsc',
-    'textit',
-    'mathit',
-    'makebox',
-    'mbox',
-    'textbf',
-    'acl',
-    'textnormal',
-    'texttt ',
-    'textsc ',
-    'textit ',
-    'mathit ',
-    'textbf ',
-    'gls',
-    'Gls',
-    'glspl',
-    'textnormal ',
-    'inlinetitle',
-    'mbox',
-    'hl',
-    'highlight',
-    'IEEEraisesectionheading',
-    'IEEEtitleabstractindextext',
-    'IEEEPARstart',
-    'it',
-    'added',
-    'paragraph',
-    'paragraph\*',
-    'MakeLowercase',
-    'emph',
-    'emph ',
-    'text',
-    'acp',
-    'ac'
-]
-
-environments_to_delete: [
-    'icmlauthorlist',
-    'tikzpicture',
-    'groupplot',
-    'biography',
-    'IEEEbiographynophoto',
-    'acronym',
-    'MSCcodes',
-    'IEEEbiography',
-    'figure',
-    'AMS',
-    'acknowledgement',
-    'acknowledgments',
-    'acknowledgements',
-    'figure\*',
-    'minipage',
-    'table',
-    'table\*',
-    'glist',
-    'tabular',
-    'tabular\*',
-    'center',
-    'remark',
-    'algorithm',
-    'algorithmic',
-    'CCSXML',
-    'acks',
-    'lstlisting',
-    'tabu',
-    'algorithm\*',
-    'algorithmic\*',
-    'longtable',
-    'sidewaystable\*',
-    'sidewaystable',
-    'appendices',
-    'wrapfigure',
-    'appendix'
-]

configs/config.yaml

@@ -1,55 +0,0 @@
-data_downloading:
-  download_directory: "quant_bio_retrieval/"    # directory where the papers will be downloaded and the graph will be saved
-  gexf_file: "test_graph.gexf"                  # name of the graph file that will be created only if downloading option is true
-  processing:
-    random_seed: 10
-    keep_unstructured_content: false            # keep unstructured content of the papers as graph node attribute if true
-    arxiv_rate_limit: 3                         # time in seconds to wait between each arxiv api call to avoid ban
-
-retriever:
-  embedder: BAAI/bge-large-en-v1.5
-  num_retrievals: 30000
-  load_arxiv_embeds: True                       # load arxiv embeddings from huggingface if true else generate them
-
-inference:
-  base_model: meta-llama/Meta-Llama-3-8B
-  pretrained_model: "models/Robotics/Meta-LLama-3-8B-Quantative-Robotics" # used only if training option is false
-  generation_args:
-    max_new_tokens: 1000
-    do_sample: True
-    top_p: 0.9
-    top_k: 50
-    temperature: 0.7
-    no_repeat_ngram_size: 2
-    num_beams: 1
-  gen_related_work_instruct_model: meta-llama/Llama-3.1-8B-Instruct      # Model assisting at the generation of related work instructions
-
-training:
-  predefined_graph_path: "robotics.gexf"                                 # path to the graph dataset used for fine-tuning only if downloading option is false
-  trainer_args:
-    per_device_train_batch_size: 4
-    warmup_steps: 100
-    num_train_epochs: 1
-    learning_rate: 0.0002
-    lr_scheduler_type: 'cosine'
-    fp16: true
-    logging_steps: 1
-    save_steps: 50
-    trainer_output_dir: trainer_outputs/
-  tokenizer:
-    max_length: 1024
-  qlora:
-    rank: 8
-    lora_alpha: 32
-    lora_dropout: 0.05
-    target_modules:          # modules for which to train lora adapters
-    - q_proj
-    - k_proj
-    - v_proj
-    - o_proj
-
-# Used only if training option is true to save and load the fine-tuned model
-model_saving:
-  model_name: llama_1b_qlora_uncensored
-  model_output_dir: models   # model saved in {model_output_dir}/{model_name}_{index} # model saved in {model_output_dir}/{model_name}_{index} after fine-tuning completion
-  index: 1

configs/config_noUI.yaml

@@ -1,37 +0,0 @@
-# Note: In order to train the model and then evaluate the same model, you need to
-# make sure that the base_model to be the same in both the train and eval sections
-
-# Evaluation configuration
-eval:
-  base_model: meta-llama/Llama-3.2-1B
-  graph_path: datasets/quantum_graph.gexf
-  model_name: llama_1b_qlora_uncensored
-
-# Training configuration
-training:
-  graph_path: datasets/quantum_graph.gexf # path to the graph file to train on
-  base_model: meta-llama/Llama-3.2-1B
-  trainer_args:
-    per_device_train_batch_size: 4
-    warmup_steps: 100
-    num_train_epochs: 1
-    learning_rate: 0.0002
-    lr_scheduler_type: 'cosine'
-    fp16: true
-    logging_steps: 1
-    save_steps: 50
-    trainer_output_dir: trainer_outputs/
-  tokenizer:
-    max_length: 1024
-  qlora:
-    rank: 8
-    lora_alpha: 32
-    lora_dropout: 0.05
-    target_modules:  # modules for which to train lora adapters
-    - q_proj
-    - k_proj
-    - v_proj
-    - o_proj
-  model_saving:
-    model_output_dir: models  # model saved in {model_output_dir}/{model_name} after fine-tuning completion
-    model_name: llama_1b_qlora_uncensored

configs/latex_commands.yaml

@@ -1,162 +0,0 @@
-verbatim_to_delete: [
-    '\IEEEpeerreviewmaketitle',
-    '\normalcolor',
-    '\ifCLASSOPTIONcaptionsoff',
-    '\pagebreak',
-    '\makeatletter',
-    '\makeatother',
-    '\maketitle',
-    '\preface',
-    '\eShell',
-    '\medskip',
-    '\tableofcontents',
-    '\begin{@twocolumnfalse}',
-    '\end{@twocolumnfalse}',
-    '\bgroup',
-    '\egroup',
-    '\ifnalpaper1',
-    '\let\thefootnote',
-    '\begin{spacing}{2.0}',
-    '\end{landscape}',
-    '\begin{landscape}',
-    '\begin{doublespacing}',
-    '\end{doublespacing}',
-    '\begin{spacing}',
-    '\end{spacing}',
-    '\printbibliography',
-    '\begin{sloppypar}',
-    '\end{sloppypar}',
-    '\ifbd',
-    '\iftr',
-    '\fussy', 
-    '\sloppy', 
-    '\emergencystretch', 
-    '\hideLIPIcs',
-    '\tolerance', 
-    '\hbadness',
-    '\bShell',
-    '\glsresetall',
-    '\copyrightnotice',
-    '\copyright',
-    '\centering',
-    '\immediate',
-    '\doublespacing',
-    '\flushbottom',
-    '\printAffiliationsAndNotice',
-    '\IEEEpubid',
-    '\twocolumn',
-    '\noindent',
-    '\indent',
-    '\onecolumn',
-    '\ignore',
-    '\selectfont',
-    '\raggedbottom',
-    '\IEEEoverridecommandlockouts',
-    '\newline',
-    '\tiny',
-    '\break',
-    '\mainmatter',
-    '\let\rc\rangle',
-    '\let\lc\langle',
-    '\acresetall',
-    '\acknowledgments',
-    '\begin{section}',
-    '\begin{small}',
-    '\end{small}',
-    '\relax',
-    '\ninept',
-    '\FloatBarrier',
-    '\boldmath',
-    '\end{section}',
-    '\Huge',
-    '\fancyhf',
-    '\fancyhead',
-    '\begin{frontmatter}',
-    '\end{frontmatter}',
-    '\clearpage',
-    '\huge',
-    '\newpage',
-    '\IEEEdisplaynontitleabstractindextext',
-    '\DontPrintSemicolon',
-    '\Large',
-    '\LARGE',
-    '\ifCLASSOPTIONcompsoc',
-    '\ifCLASSOPTIONonecolumn',
-    '\xspace',
-    '\large',
-    '\acmcopyr',
-    '\flushleft',
-    '\newpage',
-    '\protect',
-    '\begingroup',
-    '\endgroup',
-    '\bigskip',
-    '\smallskip',
-    '\small',
-    '\left',
-    '\right',
-    '\vfill',
-    '\hfill',
-    '\begin{appendices}',
-    '\IEEEQED',
-    '\leavevmode',
-    '\footnotesize',
-    '\nonumber',
-    '\scriptsize',
-    '\IEEEpubidadjcol',
-    '\balance',
-    '\normalsize',
-    '\ifloguseIEEEConf',
-    '\else',
-    '\fi', 
-    '\bf ',
-    '\it ',
-    '\verb ',
-    '\tt ',
-    '\em ',
-    '\par '
-]
-
-two_arguments: [
-    'IEEEPARstart',
-    'pgfdeclareplotmark',
-    'setcounter',
-    'texorpdfstring',
-    'fontsize',
-    'addtocounter',
-    'addtolength'
-]
-
-three_arguments: [
-    'definecolor'
-]
-
-two_arguments_elaborate: [
-    'markboth',
-    'setlength',
-    'pgfdeclareplotmark',
-    'icmlsetsymbol',
-    'texorpdfstring',
-    'conferenceinfo',
-    'acrodef',
-    'icmlcorrespondingauthor',
-    'pdfbookmark',
-    'icmlaffiliation',
-    'icmlcorrespondingauthor'
-]
-
-three_arguments_elaborate: [
-    'ifthenelse',
-    'addcontentsline'
-]
-
-replace_comments: [
-    'def\\',
-    'def ',
-    'newglossaryentry',
-    'newtheorem',
-    'newcommand',
-    'renewcommand',
-    'newenvironment',
-    'renewenvironment'
-]