Runner

class FlagEmbedding.finetune.reranker.decoder_only.layerwise.DecoderOnlyRerankerRunner(model_args: RerankerModelArguments, data_args: AbsRerankerDataArguments, training_args: AbsRerankerTrainingArguments)

Decoder only layerwise reranker runner for finetuning.

Parameters:

• model_args (RerankerModelArguments) – Model arguments instance.

• data_args (AbsRerankerDataArguments) – Data arguments instance.

• training_args (AbsRerankerTrainingArguments) – Trainer arguments.

load_tokenizer_and_model() → Tuple[PreTrainedTokenizer, AbsRerankerModel]

Load the tokenizer and model.

Returns:

Tokenizer and model instances.

Return type:

Tuple[PreTrainedTokenizer, AbsEmbedderModel]

load_trainer() → DecoderOnlyRerankerTrainer

Load the trainer.

Returns:

Loaded trainer instance.

Return type:

DecoderOnlyRerankerTrainer

run()

Run the finetuning.