RAGchain.reranker package

Subpackages

• RAGchain.reranker.importance package
  • Submodules
  • RAGchain.reranker.importance.simple_importance module
    • SimpleImportanceReranker
      • SimpleImportanceReranker.invoke()
      • SimpleImportanceReranker.rerank()
  • RAGchain.reranker.importance.weighted_importance module
    • WeightedImportanceReranker
      • WeightedImportanceReranker.invoke()
      • WeightedImportanceReranker.rerank()
  • Module contents
• RAGchain.reranker.llm package
  • Submodules
  • RAGchain.reranker.llm.llm module
    • LLMReranker
      • LLMReranker.find_passages()
      • LLMReranker.invoke()
      • LLMReranker.make_item()
      • LLMReranker.make_passages()
      • LLMReranker.rerank()
      • LLMReranker.rerank_sliding_window()
  • RAGchain.reranker.llm.rank_gpt module
    • SafeOpenai
      • SafeOpenai.chat()
      • SafeOpenai.text()
    • clean_response()
    • create_permutation_instruction()
    • get_post_prompt()
    • get_prefix_prompt()
    • max_tokens()
    • num_tokens_from_messages()
    • permutation_pipeline()
    • receive_permutation()
    • remove_duplicate()
    • run_llm()
    • sliding_windows()
  • Module contents
• RAGchain.reranker.pygaggle package
  • Subpackages
    • RAGchain.reranker.pygaggle.model package
      • Submodules
      • RAGchain.reranker.pygaggle.model.decode module
      • RAGchain.reranker.pygaggle.model.tokenize module
      • Module contents
  • Submodules
  • RAGchain.reranker.pygaggle.base module
    • Query
    • Reranker
      • Reranker.rerank()
      • Reranker.rescore()
    • Text
  • RAGchain.reranker.pygaggle.monoT5 module
    • MonoT5Reranker
      • MonoT5Reranker.invoke()
      • MonoT5Reranker.rerank()
  • RAGchain.reranker.pygaggle.transformer module
    • MonoT5
      • MonoT5.get_model()
      • MonoT5.get_prediction_tokens()
      • MonoT5.get_tokenizer()
      • MonoT5.rescore()
  • Module contents
• RAGchain.reranker.tart package
  • Submodules
  • RAGchain.reranker.tart.modeling_enc_t5 module
    • EncT5ForSequenceClassification
      • EncT5ForSequenceClassification.deparallelize()
      • EncT5ForSequenceClassification.forward()
      • EncT5ForSequenceClassification.get_encoder()
      • EncT5ForSequenceClassification.get_input_embeddings()
      • EncT5ForSequenceClassification.parallelize()
      • EncT5ForSequenceClassification.set_input_embeddings()
      • EncT5ForSequenceClassification.training
  • RAGchain.reranker.tart.tart module
    • TARTReranker
      • TARTReranker.invoke()
      • TARTReranker.rerank()
  • RAGchain.reranker.tart.tokenization_enc_t5 module
    • EncT5Tokenizer
      • EncT5Tokenizer.build_inputs_with_special_tokens()
      • EncT5Tokenizer.create_token_type_ids_from_sequences()
      • EncT5Tokenizer.get_special_tokens_mask()
  • Module contents
• RAGchain.reranker.time package
  • Submodules
  • RAGchain.reranker.time.simple_time module
    • SimpleTimeReranker
      • SimpleTimeReranker.invoke()
      • SimpleTimeReranker.rerank()
  • RAGchain.reranker.time.weighted_time module
    • WeightedTimeReranker
      • WeightedTimeReranker.invoke()
      • WeightedTimeReranker.rerank()
  • Module contents
• RAGchain.reranker.upr package
  • Submodules
  • RAGchain.reranker.upr.upr module
    • UPRReranker
      • UPRReranker.calculate_likelihood()
      • UPRReranker.invoke()
      • UPRReranker.rerank()
  • Module contents

Submodules

RAGchain.reranker.base module

class RAGchain.reranker.base.BaseReranker

Bases: Runnable[RetrievalResult, RetrievalResult], ABC

property InputType: Type[Input]

The type of input this runnable accepts specified as a type annotation.

property OutputType: Type[RetrievalResult]

The type of output this runnable produces specified as a type annotation.

abstract rerank(query: str, passages: List[Passage]) → List[Passage]

Reranks a list of passages based on a specific ranking algorithm.

Parameters:

• passages (List[Passage]) – A list of Passage objects representing the passages to be reranked.

• query – str: The query that was used for retrieving the passages.

Returns:

The reranked list of passages.

Return type:

List[Passage]

RAGchain.reranker.bm25 module

class RAGchain.reranker.bm25.BM25Reranker(tokenizer_name: str = 'gpt2', *args, **kwargs)

Bases: BaseReranker

BM25Reranker class for reranker based on BM25. You can rerank the passages with BM25 scores .

invoke(input: Input, config: RunnableConfig | None = None) → Output

Transform a single input into an output. Override to implement.

Args:

input: The input to the runnable. config: A config to use when invoking the runnable.

The config supports standard keys like ‘tags’, ‘metadata’ for tracing purposes, ‘max_concurrency’ for controlling how much work to do in parallel, and other keys. Please refer to the RunnableConfig for more details.

Returns:

The output of the runnable.

rerank(query: str, passages: List[Passage]) → List[Passage]

Reranks a list of passages based on a specific ranking algorithm.

Parameters:

• passages (List[Passage]) – A list of Passage objects representing the passages to be reranked.

• query – str: The query that was used for retrieving the passages.

Returns:

The reranked list of passages.

Return type:

List[Passage]

Module contents