RAGchain.reranker.pygaggle.model package

Submodules

RAGchain.reranker.pygaggle.model.decode module

RAGchain.reranker.pygaggle.model.decode.greedy_decode(model: PreTrainedModel, input_ids: Tensor, length: int, attention_mask: Tensor = None, return_last_logits: bool = True) → Tensor | Tuple[Tensor, Tensor]

RAGchain.reranker.pygaggle.model.tokenize module

class RAGchain.reranker.pygaggle.model.tokenize.DuoQueryDocumentBatch(query: RAGchain.reranker.pygaggle.base.Query, doc_pairs: List[Tuple[RAGchain.reranker.pygaggle.base.Text, RAGchain.reranker.pygaggle.base.Text]], output: Mapping[str, torch.Tensor | List[int] | List[List[int]] | List[List[str]]] | None = None)

Bases: object

doc_pairs: List[Tuple[Text, Text]]

output: Mapping[str, Tensor | List[int] | List[List[int]] | List[List[str]]] | None = None

query: Query

class RAGchain.reranker.pygaggle.model.tokenize.QueryDocumentBatch(query: RAGchain.reranker.pygaggle.base.Query, documents: List[RAGchain.reranker.pygaggle.base.Text], output: Mapping[str, torch.Tensor | List[int] | List[List[int]] | List[List[str]]] | None = None)

Bases: object

documents: List[Text]

output: Mapping[str, Tensor | List[int] | List[List[int]] | List[List[str]]] | None = None

query: Query

class RAGchain.reranker.pygaggle.model.tokenize.QueryDocumentBatchTokenizer(tokenizer: PreTrainedTokenizer, batch_size: int, pattern: str = '{query} {document}', **tokenizer_kwargs)

Bases: TokenizerEncodeMixin

traverse_duo_query_document(batch_input: DuoQueryDocumentBatch) → Iterable[DuoQueryDocumentBatch]

traverse_query_document(batch_input: QueryDocumentBatch) → Iterable[QueryDocumentBatch]

class RAGchain.reranker.pygaggle.model.tokenize.T5BatchTokenizer(*args, **kwargs)

Bases: QueryDocumentBatchTokenizer

class RAGchain.reranker.pygaggle.model.tokenize.TokenizerEncodeMixin

Bases: object

encode(strings: List[str]) → Mapping[str, Tensor | List[int] | List[List[int]] | List[List[str]]]

tokenizer: PreTrainedTokenizer = None

tokenizer_kwargs = None

Module contents