Parameters

• output_dir (str, defaults to "checkpoints") — The output directory where the model predictions and checkpoints will be written.
• batch_size (Union[int, Tuple[int, int]], defaults to (16, 2)) — Set the batch sizes for the embedding and classifier training phases respectively, or set both if an integer is provided. Note that the batch size for the classifier is only used with a differentiable PyTorch head.
• num_epochs (Union[int, Tuple[int, int]], defaults to (1, 16)) — Set the number of epochs the embedding and classifier training phases respectively, or set both if an integer is provided. Note that the number of epochs for the classifier is only used with a differentiable PyTorch head.
• max_steps (int, defaults to -1) — If set to a positive number, the total number of training steps to perform. Overrides num_epochs. The training may stop before reaching the set number of steps when all data is exhausted.
• sampling_strategy (str, defaults to "oversampling") — The sampling strategy of how to draw pairs in training. Possible values are:
  • "oversampling": Draws even number of positive/ negative sentence pairs until every sentence pair has been drawn.
  • "undersampling": Draws the minimum number of positive/ negative sentence pairs until every sentence pair in the minority class has been drawn.
  • "unique": Draws every sentence pair combination (likely resulting in unbalanced number of positive/ negative sentence pairs).

  The default is set to "oversampling", ensuring all sentence pairs are drawn at least once. Alternatively, setting num_iterations will override this argument and determine the number of generated sentence pairs.

• num_iterations (int, optional) — If not set the sampling_strategy will determine the number of sentence pairs to generate. This argument sets the number of iterations to generate sentence pairs for and provides compatability with Setfit CosineSimilarityLoss.
• body_learning_rate (Union[float, Tuple[float, float]], defaults to (2e-5, 1e-5)) — Set the learning rate for the SentenceTransformer body for the embedding and classifier training phases respectively, or set both if a float is provided. Note that the body learning rate for the classifier is only used with a differentiable PyTorch head and if end_to_end=True.
• head_learning_rate (float, defaults to 1e-2) — Set the learning rate for the head for the classifier training phase. Only used with a differentiable PyTorch head.
• loss (nn.Module, defaults to CosineSimilarityLoss) — The loss function to use for contrastive training of the embedding training phase.
• distance_metric (Callable, defaults to BatchHardTripletLossDistanceFunction.cosine_distance) — Function that returns a distance between two embeddings. It is set for the triplet loss and ignored for CosineSimilarityLoss and SupConLoss.
• margin (float, defaults to 0.25) — Margin for the triplet loss. Negative samples should be at least margin further apart from the anchor than the positive. It is ignored for CosineSimilarityLoss, BatchHardSoftMarginTripletLoss and SupConLoss.
• end_to_end (bool, defaults to False) — If True, train the entire model end-to-end during the classifier training phase. Otherwise, freeze the SentenceTransformer body and only train the head. Only used with a differentiable PyTorch head.
• use_amp (bool, defaults to False) — Whether to use Automatic Mixed Precision (AMP) during the embedding training phase. Only for Pytorch >= 1.6.0
• warmup_proportion (float, defaults to 0.1) — Proportion of the warmup in the total training steps. Must be greater than or equal to 0.0 and less than or equal to 1.0.
• l2_weight (float, optional) — Optional l2 weight for both the model body and head, passed to the AdamW optimizer in the classifier training phase if a differentiable PyTorch head is used.
• max_length (int, optional) — The maximum token length a tokenizer can generate. If not provided, the maximum length for the SentenceTransformer body is used.
• samples_per_label (int, defaults to 2) — Number of consecutive, random and unique samples drawn per label. This is only relevant for triplet loss and ignored for CosineSimilarityLoss. Batch size should be a multiple of samples_per_label.
• show_progress_bar (bool, defaults to True) — Whether to display a progress bar for the training epochs and iterations.
• seed (int, defaults to 42) — Random seed that will be set at the beginning of training. To ensure reproducibility across runs, use the model_init argument to Trainer to instantiate the model if it has some randomly initialized parameters.
• report_to (str or List[str], optional, defaults to "all") — The list of integrations to report the results and logs to. Supported platforms are "azure_ml", "comet_ml", "mlflow", "neptune", "tensorboard","clearml" and "wandb". Use "all" to report to all integrations installed, "none" for no integrations.
• run_name (str, optional) — A descriptor for the run. Typically used for wandb and mlflow logging.
• logging_dir (str, optional) — TensorBoard log directory. Will default to *runs/CURRENT_DATETIME_HOSTNAME*.
• logging_strategy (str or IntervalStrategy, optional, defaults to "steps") — The logging strategy to adopt during training. Possible values are:
  • "no": No logging is done during training.
  • "epoch": Logging is done at the end of each epoch.
  • "steps": Logging is done every logging_steps.
• logging_first_step (bool, optional, defaults to False) — Whether to log and evaluate the first global_step or not.
• logging_steps (int, defaults to 50) — Number of update steps between two logs if logging_strategy="steps".
• evaluation_strategy (str or IntervalStrategy, optional, defaults to "no") — The evaluation strategy to adopt during training. Possible values are:
  • "no": No evaluation is done during training.
  • "steps": Evaluation is done (and logged) every eval_steps.
  • "epoch": Evaluation is done at the end of each epoch.
• eval_steps (int, optional) — Number of update steps between two evaluations if evaluation_strategy="steps". Will default to the same value as logging_steps if not set.
• eval_delay (float, optional) — Number of epochs or steps to wait for before the first evaluation can be performed, depending on the evaluation_strategy.
• eval_max_steps (int, defaults to -1) — If set to a positive number, the total number of evaluation steps to perform. The evaluation may stop before reaching the set number of steps when all data is exhausted.
• save_strategy (str or IntervalStrategy, optional, defaults to "steps") — The checkpoint save strategy to adopt during training. Possible values are:
  • "no": No save is done during training.
  • "epoch": Save is done at the end of each epoch.
  • "steps": Save is done every save_steps.
• save_steps (int, optional, defaults to 500) — Number of updates steps before two checkpoint saves if save_strategy="steps".
• save_total_limit (int, optional, defaults to 1) — If a value is passed, will limit the total amount of checkpoints. Deletes the older checkpoints in output_dir. Note, the best model is always preserved if the evaluation_strategy is not "no".
• load_best_model_at_end (bool, optional, defaults to False) — Whether or not to load the best model found during training at the end of training.

  When set to True, the parameters save_strategy needs to be the same as evaluation_strategy, and in the case it is “steps”, save_steps must be a round multiple of eval_steps.

Parameters

• model (SetFitModel, optional) — The model to train. If not provided, a model_init must be passed.
• args (TrainingArguments, optional) — The training arguments to use.
• train_dataset (Dataset) — The training dataset.
• eval_dataset (Dataset, optional) — The evaluation dataset.
• model_init (Callable[[], SetFitModel], optional) — A function that instantiates the model to be used. If provided, each call to Trainer.train() will start from a new instance of the model as given by this function when a trial is passed.
• metric (str or Callable, optional, defaults to "accuracy") — The metric to use for evaluation. If a string is provided, we treat it as the metric name and load it with default settings. If a callable is provided, it must take two arguments (y_pred, y_test) and return a dictionary with metric keys to values.
• metric_kwargs (Dict[str, Any], optional) — Keyword arguments passed to the evaluation function if metric is an evaluation string like “f1”. For example useful for providing an averaging strategy for computing f1 in a multi-label setting.
• callbacks (List[TrainerCallback], optional) — A list of callbacks to customize the training loop. Will add those to the list of default callbacks detailed in here. If you want to remove one of the default callbacks used, use the Trainer.remove_callback() method.
• column_mapping (Dict[str, str], optional) — A mapping from the column names in the dataset to the column names expected by the model. The expected format is a dictionary with the following format: {"text_column_name": "text", "label_column_name: "label"}.

Parameters

• teacher_model (SetFitModel) — The teacher model to mimic.
• student_model (SetFitModel, optional) — The model to train. If not provided, a model_init must be passed.
• args (TrainingArguments, optional) — The training arguments to use.
• train_dataset (Dataset) — The training dataset.
• eval_dataset (Dataset, optional) — The evaluation dataset.
• model_init (Callable[[], SetFitModel], optional) — A function that instantiates the model to be used. If provided, each call to train() will start from a new instance of the model as given by this function when a trial is passed.
• metric (str or Callable, optional, defaults to "accuracy") — The metric to use for evaluation. If a string is provided, we treat it as the metric name and load it with default settings. If a callable is provided, it must take two arguments (y_pred, y_test).
• column_mapping (Dict[str, str], optional) — A mapping from the column names in the dataset to the column names expected by the model. The expected format is a dictionary with the following format: {"text_column_name": "text", "label_column_name: "label"}.

Parameters

• model (AbsaModel) — The AbsaModel model to train.
• args (TrainingArguments, optional) — The training arguments to use. If polarity_args is not defined, then args is used for both the aspect and the polarity model.
• polarity_args (TrainingArguments, optional) — The training arguments to use for the polarity model. If not defined, args is used for both the aspect and the polarity model.
• train_dataset (Dataset) — The training dataset. The dataset must have “text”, “span”, “label” and “ordinal” columns.
• eval_dataset (Dataset, optional) — The evaluation dataset. The dataset must have “text”, “span”, “label” and “ordinal” columns.
• metric (str or Callable, optional, defaults to "accuracy") — The metric to use for evaluation. If a string is provided, we treat it as the metric name and load it with default settings. If a callable is provided, it must take two arguments (y_pred, y_test).
• metric_kwargs (Dict[str, Any], optional) — Keyword arguments passed to the evaluation function if metric is an evaluation string like “f1”. For example useful for providing an averaging strategy for computing f1 in a multi-label setting.
• callbacks (List[TrainerCallback], optional) — A list of callbacks to customize the training loop. Will add those to the list of default callbacks detailed in here. If you want to remove one of the default callbacks used, use the Trainer.remove_callback() method.
• column_mapping (Dict[str, str], optional) — A mapping from the column names in the dataset to the column names expected by the model. The expected format is a dictionary with the following format: {"text_column_name": "text", "span_column_name": "span", "label_column_name: "label", "ordinal_column_name": "ordinal"}.