RAGchain.reranker.time package

Submodules

RAGchain.reranker.time.simple_time module

class RAGchain.reranker.time.simple_time.SimpleTimeReranker

Bases: BaseReranker

Rerank passages by their content_datetime only. It is simple reranker for time-aware RAG.

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(passages: List[Passage]) → List[Passage]

Rerank passages by their content_datetime only. :param passages: list of passages to be reranked.

RAGchain.reranker.time.weighted_time module

class RAGchain.reranker.time.weighted_time.WeightedTimeReranker(decay_rate: float = 0.01)

Bases: BaseReranker

Rerank passages by their content_datetime and relevance score. First, relevance score must be normalized to [0, 1] range. And calculate the combined score by the following formula:

score = (1 - decay_rate) ** hours_passed + relevance_score

The larger the decay_rate, the score from the past will be lowed.

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(passages: List[Passage], scores: List[float]) → List[Passage]
Parameters:

  • passages – list of passages to be reranked.

  • scores – list of relevance scores of passages.

Module contents