Functions — github.com/IlyaRice/RAG-Challenge-2

hub / github.com/IlyaRice/RAG-Challenge-2 / functions

Functions158 in github.com/IlyaRice/RAG-Challenge-2

⨍ Functions 158◇ Types & classes 45

↓ 1 callersMethodretrieve_all

(self, company_name: str)

src/retrieval.py:210

↓ 1 callersMethodretrieve_by_company_name

(self, company_name: str, query: str, top_n: int = 3, return_parent_pages: bool = False)

src/retrieval.py:21

↓ 1 callersMethodretrieve_by_company_name

(self, company_name: str, query: str, llm_reranking_sample_size: int = None, top_n: int = 3, return_parent_pag

src/retrieval.py:153

↓ 1 callersMethodsend_message

( self, # model='meta-llama/llama-3-1-8b-instruct', model=None, temperature=0.

src/api_requests.py:138

↓ 1 callersMethodsend_message

( self, model=None, temperature: float = 0.5, seed=12345, # For back compatib

src/api_requests.py:330

↓ 1 callersMethodserialize_tables

Process tables in files using parallel threading

src/pipeline.py:145

↓ 1 callersMethodset_up_llm

()

src/retrieval.py:97

↓ 1 callersMethodset_up_llm

(self)

src/reranking.py:41

↓ 1 callersMethodset_up_llm

(self)

src/api_requests.py:26

↓ 1 callersMethodsplit_all_reports

(self, all_report_dir: Path, output_dir: Path, serialized_tables_dir: Optional[Path] = None)

src/text_splitter.py:87

↓ 1 callersFunctiontask_id_generator_function

Generate integers 0, 1, 2, and so on.

src/api_request_parallel_processor.py:428

Method__init__

(self, bm25_db_dir: Path, documents_dir: Path)

src/retrieval.py:17

Method__init__

(self, vector_db_dir: Path, documents_dir: Path)

src/retrieval.py:81

Method__init__

(self, vector_db_dir: Path, documents_dir: Path)

(self, preserve_temp_files: bool = True)

src/tables_serialization.py:34

Method__init__

(self, root_path: Path, subset_name: str = "subset.csv", questions_file_name: str = "questions.json", pdf_repo

src/pipeline.py:19

Method__init__

(self, root_path: Path, subset_name: str = "subset.csv", questions_file_name: str = "questions.json", pdf_repo

src/pipeline.py:65

Method__init__

( self, pdf_backend=DoclingParseV2DocumentBackend, output_dir: Path = Path("./parsed_p

src/pdf_parsing.py:33

Method__init__

(self, metadata_lookup: dict = None, debug_data_path: Path = None)

src/pdf_parsing.py:251

Method__init__

( self, vector_db_dir: Union[str, Path] = './vector_dbs', documents_dir: Union[str, Pa

src/questions_processing.py:14

Method__init__

Initialize with option to add serialized tables to markdown ones.

src/parsed_reports_merging.py:12

src/api_requests.py:21

Method__init__

(self)

src/api_requests.py:88

Method__init__

(self)

src/api_requests.py:240

Method__init__

(self, provider: Literal["openai", "ibm", "gemini"] ="openai")

src/api_requests.py:371

Method_log_retry_attempt

Print information about the retry attempt

src/api_requests.py:263

Function_process_chunk

Helper function to process a chunk of PDFs in a separate process.

src/pdf_parsing.py:18

Methodcheck_balance

Check the current balance for the provided token.

src/api_requests.py:93

Methodcount_tokens

(string, encoding_name="o200k_base")

src/api_requests.py:77

Methodcreate_bm25_db

Create BM25 database from chunked reports.

src/pipeline.py:195

Functiondownload_models

Download required docling models.

main.py:11

Methodemit

(self, record)

src/tables_serialization.py:21

Methodget_available_models

Get a list of available foundation models.

src/api_requests.py:106

Methodget_embeddings

Get vector embeddings for the provided text inputs.

src/api_requests.py:118

Methodget_strings_cosine_similarity

(str1, str2)

src/retrieval.py:144

Methodlist_available_models

Prints available Gemini models that support text generation.

src/api_requests.py:251

Methodmonitor_progress

()

src/api_requests.py:535

Functionparse_pdfs

Parse PDF reports with optional parallel processing.

Methodprocess_company_question

Helper function to process one company's question and return (company, answer)

src/questions_processing.py:470

Methodprocess_file

(self, json_path: Path)

src/tables_serialization.py:228

Functionprocess_questions

Process questions using the pipeline.

main.py:51

Functionprocess_reports

Process parsed reports through the pipeline stages.

main.py:40

Methodprocess_single_doc

src/pdf_parsing.py:284

Methodprocess_with_progress

()

src/api_requests.py:552

Methodreplace_cap

(match)

src/parsed_reports_merging.py:169

Methodreplace_command

(match)

src/parsed_reports_merging.py:158

Methodreplace_glyph

(match)

src/parsed_reports_merging.py:165

Methodrerank

(self, query, documents, top_n = 10)

src/reranking.py:21

Methodretrieve_by_company_name

Retrieve and rerank documents using hybrid approach. Args: company_name: Name of the company to search documents

src/retrieval.py:245

Methodsend_message

( self, model=None, temperature=0.5, seed=None, # For deterministic ouptputs

src/api_requests.py:35

Functionserialize_tables

Serialize tables in parsed reports using parallel threading.

main.py:30

Methodserialize_tables

Process all tables in the report and add serialization results to each table's info

src/tables_serialization.py:154

← previous101–158 of 158, ranked by callers