AbsArguments#
- class FlagEmbedding.abc.finetune.embedder.AbsEmbedderModelArguments(model_name_or_path: str, config_name: str | None = None, tokenizer_name: str | None = None, cache_dir: str | None = None, trust_remote_code: bool = False, token: str = <factory>)[source]#
Abstract class for model arguments.
- class FlagEmbedding.abc.finetune.embedder.AbsEmbedderDataArguments(train_data: str | None = None, cache_path: str | None = None, train_group_size: int = 8, query_max_len: int = 32, passage_max_len: int = 128, pad_to_multiple_of: int | None = None, max_example_num_per_dataset: int = 100000000, query_instruction_for_retrieval: str | None = None, query_instruction_format: str = '{}{}', knowledge_distillation: bool = False, passage_instruction_for_retrieval: str | None = None, passage_instruction_format: str | None = '{}{}', shuffle_ratio: float = 0.0, same_dataset_within_batch: bool = False, small_threshold: int = 0, drop_threshold: int = 0)[source]#
Abstract class for data arguments.