Haystack docs home page

Module base

BaseReader

class BaseReader(BaseComponent)

BaseReader.timing

def timing(fn, attr_name)

Wrapper method used to time functions.

Module farm

FARMReader

class FARMReader(BaseReader)

Transformer based model for extractive Question Answering using the FARM framework (https://github.com/deepset-ai/FARM). While the underlying model can vary (BERT, Roberta, DistilBERT, ...), the interface remains the same.

| With a FARMReader, you can:

  • directly get predictions via predict()
  • fine-tune the model on QA data via train()

FARMReader.__init__

def __init__(model_name_or_path: str, model_version: Optional[str] = None, context_window_size: int = 150, batch_size: int = 50, use_gpu: bool = True, devices: List[torch.device] = [], no_ans_boost: float = 0.0, return_no_answer: bool = False, top_k: int = 10, top_k_per_candidate: int = 3, top_k_per_sample: int = 1, num_processes: Optional[int] = None, max_seq_len: int = 256, doc_stride: int = 128, progress_bar: bool = True, duplicate_filtering: int = 0, use_confidence_scores: bool = True, confidence_threshold: Optional[float] = None, proxies: Optional[Dict[str, str]] = None, local_files_only=False, force_download=False, use_auth_token: Optional[Union[str, bool]] = None)

Arguments:

  • model_name_or_path: Directory of a saved model or the name of a public model e.g. 'bert-base-cased', 'deepset/bert-base-cased-squad2', 'deepset/bert-base-cased-squad2', 'distilbert-base-uncased-distilled-squad'. See https://huggingface.co/models for full list of available models.
  • model_version: The version of model to use from the HuggingFace model hub. Can be tag name, branch name, or commit hash.
  • context_window_size: The size, in characters, of the window around the answer span that is used when displaying the context around the answer.
  • batch_size: Number of samples the model receives in one batch for inference. Memory consumption is much lower in inference mode. Recommendation: Increase the batch size to a value so only a single batch is used.
  • use_gpu: Whether to use GPUs or the CPU. Falls back on CPU if no GPU is available.
  • devices: List of GPU devices to limit inference to certain GPUs and not use all available ones (e.g. [torch.device('cuda:0')]). Unused if use_gpu is False.
  • no_ans_boost: How much the no_answer logit is boosted/increased. If set to 0 (default), the no_answer logit is not changed. If a negative number, there is a lower chance of "no_answer" being predicted. If a positive number, there is an increased chance of "no_answer"
  • return_no_answer: Whether to include no_answer predictions in the results.
  • top_k: The maximum number of answers to return
  • top_k_per_candidate: How many answers to extract for each candidate doc that is coming from the retriever (might be a long text). Note that this is not the number of "final answers" you will receive (see top_k in FARMReader.predict() or Finder.get_answers() for that) and that FARM includes no_answer in the sorted list of predictions.
  • top_k_per_sample: How many answers to extract from each small text passage that the model can process at once (one "candidate doc" is usually split into many smaller "passages"). You usually want a very small value here, as it slows down inference and you don't gain much of quality by having multiple answers from one passage. Note that this is not the number of "final answers" you will receive (see top_k in FARMReader.predict() or Finder.get_answers() for that) and that FARM includes no_answer in the sorted list of predictions.
  • num_processes: The number of processes for multiprocessing.Pool. Set to value of 0 to disable multiprocessing. Set to None to let Inferencer determine optimum number. If you want to debug the Language Model, you might need to disable multiprocessing!
  • max_seq_len: Max sequence length of one input text for the model
  • doc_stride: Length of striding window for splitting long texts (used if len(text) > max_seq_len)
  • progress_bar: Whether to show a tqdm progress bar or not. Can be helpful to disable in production deployments to keep the logs clean.
  • duplicate_filtering: Answers are filtered based on their position. Both start and end position of the answers are considered. The higher the value, answers that are more apart are filtered out. 0 corresponds to exact duplicates. -1 turns off duplicate removal.
  • use_confidence_scores: Determines the type of score that is used for ranking a predicted answer. True => a scaled confidence / relevance score between [0, 1]. This score can also be further calibrated on your dataset via self.eval() (see https://haystack.deepset.ai/components/reader#confidence-scores). False => an unscaled, raw score [-inf, +inf] which is the sum of start and end logit from the model for the predicted span. Using confidence scores can change the ranking of no_answer compared to using the unscaled raw scores.
  • confidence_threshold: Filters out predictions below confidence_threshold. Value should be between 0 and 1. Disabled by default.
  • proxies: Dict of proxy servers to use for downloading external models. Example: {'http': 'some.proxy:1234', 'http://hostname': 'my.proxy:3111'}
  • local_files_only: Whether to force checking for local files only (and forbid downloads)
  • force_download: Whether fo force a (re-)download even if the model exists locally in the cache.
  • use_auth_token: API token used to download private models from Huggingface. If this parameter is set to True, the local token will be used, which must be previously created via transformer-cli login. Additional information can be found here https://huggingface.co/transformers/main_classes/model.html#transformers.PreTrainedModel.from_pretrained

FARMReader.train

def train(data_dir: str, train_filename: str, dev_filename: Optional[str] = None, test_filename: Optional[str] = None, use_gpu: Optional[bool] = None, devices: List[torch.device] = [], batch_size: int = 10, n_epochs: int = 2, learning_rate: float = 1e-5, max_seq_len: Optional[int] = None, warmup_proportion: float = 0.2, dev_split: float = 0, evaluate_every: int = 300, save_dir: Optional[str] = None, num_processes: Optional[int] = None, use_amp: str = None, checkpoint_root_dir: Path = Path("model_checkpoints"), checkpoint_every: Optional[int] = None, checkpoints_to_keep: int = 3, caching: bool = False, cache_path: Path = Path("cache/data_silo"), grad_acc_steps: int = 1)

Fine-tune a model on a QA dataset. Options:

  • Take a plain language model (e.g. bert-base-cased) and train it for QA (e.g. on SQuAD data)
  • Take a QA model (e.g. deepset/bert-base-cased-squad2) and fine-tune it for your domain (e.g. using your labels collected via the haystack annotation tool)

Checkpoints can be stored via setting checkpoint_every to a custom number of steps. If any checkpoints are stored, a subsequent run of train() will resume training from the latest available checkpoint.

Arguments:

  • data_dir: Path to directory containing your training data in SQuAD style
  • train_filename: Filename of training data
  • dev_filename: Filename of dev / eval data
  • test_filename: Filename of test data
  • dev_split: Instead of specifying a dev_filename, you can also specify a ratio (e.g. 0.1) here that gets split off from training data for eval.
  • use_gpu: Whether to use GPU (if available)
  • devices: List of GPU devices to limit inference to certain GPUs and not use all available ones (e.g. [torch.device('cuda:0')]). Unused if use_gpu is False.
  • batch_size: Number of samples the model receives in one batch for training
  • n_epochs: Number of iterations on the whole training data set
  • learning_rate: Learning rate of the optimizer
  • max_seq_len: Maximum text length (in tokens). Everything longer gets cut down.
  • warmup_proportion: Proportion of training steps until maximum learning rate is reached. Until that point LR is increasing linearly. After that it's decreasing again linearly. Options for different schedules are available in FARM.
  • evaluate_every: Evaluate the model every X steps on the hold-out eval dataset
  • save_dir: Path to store the final model
  • num_processes: The number of processes for multiprocessing.Pool during preprocessing. Set to value of 1 to disable multiprocessing. When set to 1, you cannot split away a dev set from train set. Set to None to use all CPU cores minus one.
  • use_amp: Optimization level of NVIDIA's automatic mixed precision (AMP). The higher the level, the faster the model. Available options: None (Don't use AMP) "O0" (Normal FP32 training) "O1" (Mixed Precision => Recommended) "O2" (Almost FP16) "O3" (Pure FP16). See details on: https://nvidia.github.io/apex/amp.html
  • checkpoint_root_dir: the Path of directory where all train checkpoints are saved. For each individual checkpoint, a subdirectory with the name epoch_{epoch_num}step{step_num} is created.
  • checkpoint_every: save a train checkpoint after this many steps of training.
  • checkpoints_to_keep: maximum number of train checkpoints to save.
  • caching: whether or not to use caching for preprocessed dataset
  • cache_path: Path to cache the preprocessed dataset
  • processor: The processor to use for preprocessing. If None, the default SquadProcessor is used.
  • grad_acc_steps: The number of steps to accumulate gradients for before performing a backward pass.

Returns:

None

FARMReader.distil_prediction_layer_from

def distil_prediction_layer_from(teacher_model: "FARMReader", data_dir: str, train_filename: str, dev_filename: Optional[str] = None, test_filename: Optional[str] = None, use_gpu: Optional[bool] = None, devices: List[torch.device] = [], student_batch_size: int = 10, teacher_batch_size: Optional[int] = None, n_epochs: int = 2, learning_rate: float = 3e-5, max_seq_len: Optional[int] = None, warmup_proportion: float = 0.2, dev_split: float = 0, evaluate_every: int = 300, save_dir: Optional[str] = None, num_processes: Optional[int] = None, use_amp: str = None, checkpoint_root_dir: Path = Path("model_checkpoints"), checkpoint_every: Optional[int] = None, checkpoints_to_keep: int = 3, caching: bool = False, cache_path: Path = Path("cache/data_silo"), distillation_loss_weight: float = 0.5, distillation_loss: Union[str, Callable[[torch.Tensor, torch.Tensor], torch.Tensor]] = "kl_div", temperature: float = 1.0, grad_acc_steps: int = 1)

Fine-tune a model on a QA dataset using logit-based distillation. You need to provide a teacher model that is already finetuned on the dataset

and a student model that will be trained using the teacher's logits. The idea of this is to increase the accuracy of a lightweight student model. using a more complex teacher. Originally proposed in: https://arxiv.org/pdf/1503.02531.pdf This can also be considered as the second stage of distillation finetuning as described in the TinyBERT paper: https://arxiv.org/pdf/1909.10351.pdf Example

student = FARMReader(model_name_or_path="prajjwal1/bert-medium")
teacher = FARMReader(model_name_or_path="deepset/bert-large-uncased-whole-word-masking-squad2")
student.distil_prediction_layer_from(teacher, data_dir="squad2", train_filename="train.json", test_filename="dev.json",
                    learning_rate=3e-5, distillation_loss_weight=1.0, temperature=5)

Checkpoints can be stored via setting checkpoint_every to a custom number of steps. If any checkpoints are stored, a subsequent run of train() will resume training from the latest available checkpoint.

Arguments:

  • teacher_model: Model whose logits will be used to improve accuracy
  • data_dir: Path to directory containing your training data in SQuAD style
  • train_filename: Filename of training data
  • dev_filename: Filename of dev / eval data
  • test_filename: Filename of test data
  • dev_split: Instead of specifying a dev_filename, you can also specify a ratio (e.g. 0.1) here that gets split off from training data for eval.
  • use_gpu: Whether to use GPU (if available)
  • devices: List of GPU devices to limit inference to certain GPUs and not use all available ones (e.g. [torch.device('cuda:0')]). Unused if use_gpu is False.
  • student_batch_size: Number of samples the student model receives in one batch for training
  • student_batch_size: Number of samples the teacher model receives in one batch for distillation
  • n_epochs: Number of iterations on the whole training data set
  • learning_rate: Learning rate of the optimizer
  • max_seq_len: Maximum text length (in tokens). Everything longer gets cut down.
  • warmup_proportion: Proportion of training steps until maximum learning rate is reached. Until that point LR is increasing linearly. After that it's decreasing again linearly. Options for different schedules are available in FARM.
  • evaluate_every: Evaluate the model every X steps on the hold-out eval dataset
  • save_dir: Path to store the final model
  • num_processes: The number of processes for multiprocessing.Pool during preprocessing. Set to value of 1 to disable multiprocessing. When set to 1, you cannot split away a dev set from train set. Set to None to use all CPU cores minus one.
  • use_amp: Optimization level of NVIDIA's automatic mixed precision (AMP). The higher the level, the faster the model. Available options: None (Don't use AMP) "O0" (Normal FP32 training) "O1" (Mixed Precision => Recommended) "O2" (Almost FP16) "O3" (Pure FP16). See details on: https://nvidia.github.io/apex/amp.html
  • checkpoint_root_dir: the Path of directory where all train checkpoints are saved. For each individual checkpoint, a subdirectory with the name epoch_{epoch_num}step{step_num} is created.
  • checkpoint_every: save a train checkpoint after this many steps of training.
  • checkpoints_to_keep: maximum number of train checkpoints to save.
  • caching: whether or not to use caching for preprocessed dataset and teacher logits
  • cache_path: Path to cache the preprocessed dataset and teacher logits
  • distillation_loss_weight: The weight of the distillation loss. A higher weight means the teacher outputs are more important.
  • distillation_loss: Specifies how teacher and model logits should be compared. Can either be a string ("mse" for mean squared error or "kl_div" for kl divergence loss) or a callable loss function (needs to have named parameters student_logits and teacher_logits)
  • temperature: The temperature for distillation. A higher temperature will result in less certainty of teacher outputs. A lower temperature means more certainty. A temperature of 1.0 does not change the certainty of the model.
  • tinybert_loss: Whether to use the TinyBERT loss function for distillation. This requires the student to be a TinyBERT model and the teacher to be a finetuned version of bert-base-uncased.
  • tinybert_epochs: Number of epochs to train the student model with the TinyBERT loss function. After this many epochs, the student model is trained with the regular distillation loss function.
  • tinybert_learning_rate: Learning rate to use when training the student model with the TinyBERT loss function.
  • tinybert_train_filename: Filename of training data to use when training the student model with the TinyBERT loss function. To best follow the original paper, this should be an augmented version of the training data created using the augment_squad.py script. If not specified, the training data from the original training is used.
  • processor: The processor to use for preprocessing. If None, the default SquadProcessor is used.
  • grad_acc_steps: The number of steps to accumulate gradients for before performing a backward pass.

Returns:

None

FARMReader.distil_intermediate_layers_from

def distil_intermediate_layers_from(teacher_model: "FARMReader", data_dir: str, train_filename: str, dev_filename: Optional[str] = None, test_filename: Optional[str] = None, use_gpu: Optional[bool] = None, devices: List[torch.device] = [], batch_size: int = 10, n_epochs: int = 5, learning_rate: float = 5e-5, max_seq_len: Optional[int] = None, warmup_proportion: float = 0.2, dev_split: float = 0, evaluate_every: int = 300, save_dir: Optional[str] = None, num_processes: Optional[int] = None, use_amp: str = None, checkpoint_root_dir: Path = Path("model_checkpoints"), checkpoint_every: Optional[int] = None, checkpoints_to_keep: int = 3, caching: bool = False, cache_path: Path = Path("cache/data_silo"), distillation_loss: Union[str, Callable[[torch.Tensor, torch.Tensor], torch.Tensor]] = "mse", temperature: float = 1.0, processor: Optional[Processor] = None, grad_acc_steps: int = 1)

The first stage of distillation finetuning as described in the TinyBERT paper:

https://arxiv.org/pdf/1909.10351.pdf Example

student = FARMReader(model_name_or_path="prajjwal1/bert-medium")
teacher = FARMReader(model_name_or_path="huawei-noah/TinyBERT_General_6L_768D")
student.distil_intermediate_layers_from(teacher, data_dir="squad2", train_filename="train.json", test_filename="dev.json",
                    learning_rate=3e-5, distillation_loss_weight=1.0, temperature=5)

Checkpoints can be stored via setting checkpoint_every to a custom number of steps. If any checkpoints are stored, a subsequent run of train() will resume training from the latest available checkpoint.

Arguments:

  • teacher_model: Model whose logits will be used to improve accuracy
  • data_dir: Path to directory containing your training data in SQuAD style
  • train_filename: Filename of training data. To best follow the original paper, this should be an augmented version of the training data created using the augment_squad.py script
  • dev_filename: Filename of dev / eval data
  • test_filename: Filename of test data
  • dev_split: Instead of specifying a dev_filename, you can also specify a ratio (e.g. 0.1) here that gets split off from training data for eval.
  • use_gpu: Whether to use GPU (if available)
  • devices: List of GPU devices to limit inference to certain GPUs and not use all available ones (e.g. [torch.device('cuda:0')]). Unused if use_gpu is False.
  • student_batch_size: Number of samples the student model receives in one batch for training
  • student_batch_size: Number of samples the teacher model receives in one batch for distillation
  • n_epochs: Number of iterations on the whole training data set
  • learning_rate: Learning rate of the optimizer
  • max_seq_len: Maximum text length (in tokens). Everything longer gets cut down.
  • warmup_proportion: Proportion of training steps until maximum learning rate is reached. Until that point LR is increasing linearly. After that it's decreasing again linearly. Options for different schedules are available in FARM.
  • evaluate_every: Evaluate the model every X steps on the hold-out eval dataset
  • save_dir: Path to store the final model
  • num_processes: The number of processes for multiprocessing.Pool during preprocessing. Set to value of 1 to disable multiprocessing. When set to 1, you cannot split away a dev set from train set. Set to None to use all CPU cores minus one.
  • use_amp: Optimization level of NVIDIA's automatic mixed precision (AMP). The higher the level, the faster the model. Available options: None (Don't use AMP) "O0" (Normal FP32 training) "O1" (Mixed Precision => Recommended) "O2" (Almost FP16) "O3" (Pure FP16). See details on: https://nvidia.github.io/apex/amp.html
  • checkpoint_root_dir: the Path of directory where all train checkpoints are saved. For each individual checkpoint, a subdirectory with the name epoch_{epoch_num}step{step_num} is created.
  • checkpoint_every: save a train checkpoint after this many steps of training.
  • checkpoints_to_keep: maximum number of train checkpoints to save.
  • caching: whether or not to use caching for preprocessed dataset and teacher logits
  • cache_path: Path to cache the preprocessed dataset and teacher logits
  • distillation_loss_weight: The weight of the distillation loss. A higher weight means the teacher outputs are more important.
  • distillation_loss: Specifies how teacher and model logits should be compared. Can either be a string ("mse" for mean squared error or "kl_div" for kl divergence loss) or a callable loss function (needs to have named parameters student_logits and teacher_logits)
  • temperature: The temperature for distillation. A higher temperature will result in less certainty of teacher outputs. A lower temperature means more certainty. A temperature of 1.0 does not change the certainty of the model.
  • processor: The processor to use for preprocessing. If None, the default SquadProcessor is used.
  • grad_acc_steps: The number of steps to accumulate gradients for before performing a backward pass.

Returns:

None

FARMReader.update_parameters

def update_parameters(context_window_size: Optional[int] = None, no_ans_boost: Optional[float] = None, return_no_answer: Optional[bool] = None, max_seq_len: Optional[int] = None, doc_stride: Optional[int] = None)

Hot update parameters of a loaded Reader. It may not to be safe when processing concurrent requests.

FARMReader.save

def save(directory: Path)

Saves the Reader model so that it can be reused at a later point in time.

Arguments:

  • directory: Directory where the Reader model should be saved

FARMReader.save_to_remote

def save_to_remote(repo_id: str, private: Optional[bool] = None, commit_message: str = "Add new model to Hugging Face.")

Saves the Reader model to Hugging Face Model Hub with the given model_name. For this to work:

  • Be logged in to Hugging Face on your machine via transformers-cli
  • Have git lfs installed (https://packagecloud.io/github/git-lfs/install), you can test it by git lfs --version

Arguments:

  • repo_id: A namespace (user or an organization) and a repo name separated by a '/' of the model you want to save to Hugging Face
  • private: Set to true to make the model repository private
  • commit_message: Commit message while saving to Hugging Face

FARMReader.predict_batch

def predict_batch(queries: List[str], documents: Union[List[Document], List[List[Document]]], top_k: Optional[int] = None, batch_size: Optional[int] = None)

Use loaded QA model to find answers for the queries in the Documents.

  • If you provide a list containing a single query...

    • ... and a single list of Documents, the query will be applied to each Document individually.
    • ... and a list of lists of Documents, the query will be applied to each list of Documents and the Answers will be aggregated per Document list.

  • If you provide a list of multiple queries...

    • ... and a single list of Documents, each query will be applied to each Document individually.
    • ... and a list of lists of Documents, each query will be applied to its corresponding list of Documents and the Answers will be aggregated per query-Document pair.

Arguments:

  • queries: Single query or list of queries.
  • documents: Related documents (e.g. coming from a retriever) that the answer shall be conditioned on. Can be a single list of Documents or a list of lists of Documents.
  • top_k: Number of returned answers per query.
  • batch_size: Number of query-document pairs to be processed at a time.

FARMReader.predict

def predict(query: str, documents: List[Document], top_k: Optional[int] = None)

Use loaded QA model to find answers for a query in the supplied list of Document.

Returns dictionaries containing answers sorted by (desc.) score. Example:

|{
   |    'query': 'Who is the father of Arya Stark?',
   |    'answers':[Answer(
   |                 'answer': 'Eddard,',
   |                 'context': "She travels with her father, Eddard, to King's Landing when he is",
   |                 'score': 0.9787139466668613,
   |                 'offsets_in_context': [Span(start=29, end=35],
   |                 'offsets_in_context': [Span(start=347, end=353],
   |                 'document_id': '88d1ed769d003939d3a0d28034464ab2'
   |                 ),...
   |              ]
   |}

Arguments:

  • query: Query string
  • documents: List of Document in which to search for the answer
  • top_k: The maximum number of answers to return

Returns:

Dict containing query and answers

FARMReader.eval_on_file

def eval_on_file(data_dir: Union[Path, str], test_filename: str, device: Optional[Union[str, torch.device]] = None, calibrate_conf_scores: bool = False)

Performs evaluation on a SQuAD-formatted file.

Returns a dict containing the following metrics: - "EM": exact match score - "f1": F1-Score - "top_n_accuracy": Proportion of predicted answers that overlap with correct answer

Arguments:

  • data_dir: The directory in which the test set can be found
  • test_filename: The name of the file containing the test data in SQuAD format.
  • device: The device on which the tensors should be processed. Choose from torch.device("cpu") and torch.device("cuda") (or simply "cpu" or "cuda") or use the Reader's device by default.
  • calibrate_conf_scores: Whether to calibrate the temperature for scaling of the confidence scores.

FARMReader.eval

def eval(document_store: BaseDocumentStore, device: Optional[Union[str, torch.device]] = None, label_index: str = "label", doc_index: str = "eval_document", label_origin: str = "gold-label", calibrate_conf_scores: bool = False)

Performs evaluation on evaluation documents in the DocumentStore.

Returns a dict containing the following metrics: - "EM": Proportion of exact matches of predicted answers with their corresponding correct answers - "f1": Average overlap between predicted answers and their corresponding correct answers - "top_n_accuracy": Proportion of predicted answers that overlap with correct answer

Arguments:

  • document_store: DocumentStore containing the evaluation documents
  • device: The device on which the tensors should be processed. Choose from torch.device("cpu") and torch.device("cuda") (or simply "cpu" or "cuda") or use the Reader's device by default.
  • label_index: Index/Table name where labeled questions are stored
  • doc_index: Index/Table name where documents that are used for evaluation are stored
  • label_origin: Field name where the gold labels are stored
  • calibrate_conf_scores: Whether to calibrate the temperature for scaling of the confidence scores.

FARMReader.calibrate_confidence_scores

def calibrate_confidence_scores(document_store: BaseDocumentStore, device: Optional[Union[str, torch.device]] = None, label_index: str = "label", doc_index: str = "eval_document", label_origin: str = "gold_label")

Calibrates confidence scores on evaluation documents in the DocumentStore.

Arguments:

  • document_store: DocumentStore containing the evaluation documents
  • device: The device on which the tensors should be processed. Choose from torch.device("cpu") and torch.device("cuda") (or simply "cpu" or "cuda") or use the Reader's device by default.
  • label_index: Index/Table name where labeled questions are stored
  • doc_index: Index/Table name where documents that are used for evaluation are stored
  • label_origin: Field name where the gold labels are stored

FARMReader.predict_on_texts

def predict_on_texts(question: str, texts: List[str], top_k: Optional[int] = None)

Use loaded QA model to find answers for a question in the supplied list of Document.

Returns dictionaries containing answers sorted by (desc.) score. Example:

|{
   |    'question': 'Who is the father of Arya Stark?',
   |    'answers':[
   |                 {'answer': 'Eddard,',
   |                 'context': " She travels with her father, Eddard, to King's Landing when he is ",
   |                 'offset_answer_start': 147,
   |                 'offset_answer_end': 154,
   |                 'score': 0.9787139466668613,
   |                 'document_id': '1337'
   |                 },...
   |              ]
   |}

Arguments:

  • question: Question string
  • documents: List of documents as string type
  • top_k: The maximum number of answers to return

Returns:

Dict containing question and answers

FARMReader.convert_to_onnx

@classmethod
def convert_to_onnx(cls, model_name: str, output_path: Path, convert_to_float16: bool = False, quantize: bool = False, task_type: str = "question_answering", opset_version: int = 11)

Convert a PyTorch BERT model to ONNX format and write to ./onnx-export dir. The converted ONNX model

can be loaded with in the FARMReader using the export path as model_name_or_path param.

Usage:

`from haystack.reader.farm import FARMReader
from pathlib import Path
onnx_model_path = Path("roberta-onnx-model")
FARMReader.convert_to_onnx(model_name="deepset/bert-base-cased-squad2", output_path=onnx_model_path)
reader = FARMReader(onnx_model_path)`

Arguments:

  • model_name: transformers model name
  • output_path: Path to output the converted model
  • convert_to_float16: Many models use float32 precision by default. With the half precision of float16, inference is faster on Nvidia GPUs with Tensor core like T4 or V100. On older GPUs, float32 could still be be more performant.
  • quantize: convert floating point number to integers
  • task_type: Type of task for the model. Available options: "question_answering" or "embeddings".
  • opset_version: ONNX opset version

Module transformers

TransformersReader

class TransformersReader(BaseReader)

Transformer based model for extractive Question Answering using the HuggingFace's transformers framework (https://github.com/huggingface/transformers). While the underlying model can vary (BERT, Roberta, DistilBERT ...), the interface remains the same. With this reader, you can directly get predictions via predict()

TransformersReader.__init__

def __init__(model_name_or_path: str = "distilbert-base-uncased-distilled-squad", model_version: Optional[str] = None, tokenizer: Optional[str] = None, context_window_size: int = 70, use_gpu: bool = True, top_k: int = 10, top_k_per_candidate: int = 3, return_no_answers: bool = False, max_seq_len: int = 256, doc_stride: int = 128, batch_size: int = 16)

Load a QA model from Transformers.

Available models include:

  • 'distilbert-base-uncased-distilled-squad`'
  • 'bert-large-cased-whole-word-masking-finetuned-squad'
  • 'bert-large-uncased-whole-word-masking-finetuned-squad'

See https://huggingface.co/models for full list of available QA models

Arguments:

  • model_name_or_path: Directory of a saved model or the name of a public model e.g. 'bert-base-cased', 'deepset/bert-base-cased-squad2', 'deepset/bert-base-cased-squad2', 'distilbert-base-uncased-distilled-squad'. See https://huggingface.co/models for full list of available models.
  • model_version: The version of model to use from the HuggingFace model hub. Can be tag name, branch name, or commit hash.
  • tokenizer: Name of the tokenizer (usually the same as model)
  • context_window_size: Num of chars (before and after the answer) to return as "context" for each answer. The context usually helps users to understand if the answer really makes sense.
  • use_gpu: Whether to use GPU (if available).
  • top_k: The maximum number of answers to return
  • top_k_per_candidate: How many answers to extract for each candidate doc that is coming from the retriever (might be a long text). Note that this is not the number of "final answers" you will receive (see top_k in TransformersReader.predict() or Finder.get_answers() for that) and that no_answer can be included in the sorted list of predictions.
  • return_no_answers: If True, the HuggingFace Transformers model could return a "no_answer" (i.e. when there is an unanswerable question) If False, it cannot return a "no_answer". Note that no_answer_boost is unfortunately not available with TransformersReader. If you would like to set no_answer_boost, use a FARMReader.
  • max_seq_len: max sequence length of one input text for the model
  • doc_stride: length of striding window for splitting long texts (used if len(text) > max_seq_len)
  • batch_size: Number of documents to process at a time.

TransformersReader.predict

def predict(query: str, documents: List[Document], top_k: Optional[int] = None)

Use loaded QA model to find answers for a query in the supplied list of Document.

Returns dictionaries containing answers sorted by (desc.) score. Example:

|{
   |    'query': 'Who is the father of Arya Stark?',
   |    'answers':[
   |                 {'answer': 'Eddard,',
   |                 'context': " She travels with her father, Eddard, to King's Landing when he is ",
   |                 'offset_answer_start': 147,
   |                 'offset_answer_end': 154,
   |                 'score': 0.9787139466668613,
   |                 'document_id': '1337'
   |                 },...
   |              ]
   |}

Arguments:

  • query: Query string
  • documents: List of Document in which to search for the answer
  • top_k: The maximum number of answers to return

Returns:

Dict containing query and answers

TransformersReader.predict_batch

def predict_batch(queries: List[str], documents: Union[List[Document], List[List[Document]]], top_k: Optional[int] = None, batch_size: Optional[int] = None)

Use loaded QA model to find answers for the queries in the Documents.

  • If you provide a list containing a single query...

    • ... and a single list of Documents, the query will be applied to each Document individually.
    • ... and a list of lists of Documents, the query will be applied to each list of Documents and the Answers will be aggregated per Document list.

  • If you provide a list of multiple queries...

    • ... and a single list of Documents, each query will be applied to each Document individually.
    • ... and a list of lists of Documents, each query will be applied to its corresponding list of Documents and the Answers will be aggregated per query-Document pair.

Arguments:

  • queries: Single query or list of queries.
  • documents: Related documents (e.g. coming from a retriever) that the answer shall be conditioned on. Can be a single list of Documents or a list of lists of Documents.
  • top_k: Number of returned answers per query.
  • batch_size: Number of query-document pairs to be processed at a time.

Module table

TableReader

class TableReader(BaseReader)

Transformer-based model for extractive Question Answering on Tables with TaPas using the HuggingFace's transformers framework (https://github.com/huggingface/transformers). With this reader, you can directly get predictions via predict()

Example:

from haystack import Document
from haystack.reader import TableReader
import pandas as pd

table_reader = TableReader(model_name_or_path="google/tapas-base-finetuned-wtq")
data = {
    "actors": ["brad pitt", "leonardo di caprio", "george clooney"],
    "age": ["57", "46", "60"],
    "number of movies": ["87", "53", "69"],
    "date of birth": ["7 february 1967", "10 june 1996", "28 november 1967"],
}
table = pd.DataFrame(data)
document = Document(content=table, content_type="table")
query = "When was DiCaprio born?"
prediction = table_reader.predict(query=query, documents=[document])
answer = prediction["answers"][0].answer  # "10 june 1996"

TableReader.__init__

def __init__(model_name_or_path: str = "google/tapas-base-finetuned-wtq", model_version: Optional[str] = None, tokenizer: Optional[str] = None, use_gpu: bool = True, top_k: int = 10, top_k_per_candidate: int = 3, return_no_answer: bool = False, max_seq_len: int = 256)

Load a TableQA model from Transformers.

Available models include:

  • 'google/tapas-base-finetuned-wtq`'
  • 'google/tapas-base-finetuned-wikisql-supervised'
  • 'deepset/tapas-large-nq-hn-reader'
  • 'deepset/tapas-large-nq-reader'

See https://huggingface.co/models?pipeline_tag=table-question-answering for full list of available TableQA models.

The nq-reader models are able to provide confidence scores, but cannot handle questions that need aggregation over multiple cells. The returned answers are sorted first by a general table score and then by answer span scores. All the other models can handle aggregation questions, but don't provide reasonable confidence scores.

Arguments:

  • model_name_or_path: Directory of a saved model or the name of a public model e.g. See https://huggingface.co/models?pipeline_tag=table-question-answering for full list of available models.
  • model_version: The version of model to use from the HuggingFace model hub. Can be tag name, branch name, or commit hash.
  • tokenizer: Name of the tokenizer (usually the same as model)
  • use_gpu: Whether to use GPU or CPU. Falls back on CPU if no GPU is available.
  • top_k: The maximum number of answers to return
  • top_k_per_candidate: How many answers to extract for each candidate table that is coming from the retriever.
  • return_no_answer: Whether to include no_answer predictions in the results. (Only applicable with nq-reader models.)
  • max_seq_len: Max sequence length of one input table for the model. If the number of tokens of query + table exceed max_seq_len, the table will be truncated by removing rows until the input size fits the model.

TableReader.predict

def predict(query: str, documents: List[Document], top_k: Optional[int] = None) -> Dict

Use loaded TableQA model to find answers for a query in the supplied list of Documents

of content_type 'table'.

Returns dictionary containing query and list of Answer objects sorted by (desc.) score. WARNING: The answer scores are not reliable, as they are always extremely high, even if a question cannot be answered by a given table.

Arguments:

  • query: Query string
  • documents: List of Document in which to search for the answer. Documents should be of content_type 'table'.
  • top_k: The maximum number of answers to return

Returns:

Dict containing query and answers

TableReader.predict_batch

def predict_batch(queries: List[str], documents: Union[List[Document], List[List[Document]]], top_k: Optional[int] = None, batch_size: Optional[int] = None)

Use loaded TableQA model to find answers for the supplied queries in the supplied Documents

of content_type 'table'.

Returns dictionary containing query and list of Answer objects sorted by (desc.) score.

WARNING: The answer scores are not reliable, as they are always extremely high, even if a question cannot be answered by a given table.

  • If you provide a list containing a single query...

    • ... and a single list of Documents, the query will be applied to each Document individually.
    • ... and a list of lists of Documents, the query will be applied to each list of Documents and the Answers will be aggregated per Document list.

  • If you provide a list of multiple queries...

    • ... and a single list of Documents, each query will be applied to each Document individually.
    • ... and a list of lists of Documents, each query will be applied to its corresponding list of Documents and the Answers will be aggregated per query-Document pair.

Arguments:

  • queries: Single query string or list of queries.
  • documents: Single list of Documents or list of lists of Documents in which to search for the answers. Documents should be of content_type 'table'.
  • top_k: The maximum number of answers to return per query.
  • batch_size: Not applicable.

RCIReader

class RCIReader(BaseReader)

Table Reader model based on Glass et al. (2021)'s Row-Column-Intersection model. See the original paper for more details: Glass, Michael, et al. (2021): "Capturing Row and Column Semantics in Transformer Based Question Answering over Tables" (https://aclanthology.org/2021.naacl-main.96/)

Each row and each column is given a score with regard to the query by two separate models. The score of each cell is then calculated as the sum of the corresponding row score and column score. Accordingly, the predicted answer is the cell with the highest score.

Pros and Cons of RCIReader compared to TableReader:

  • Provides meaningful confidence scores
  • Allows larger tables as input
  • Does not support aggregation over table cells
  • Slower

RCIReader.__init__

def __init__(row_model_name_or_path: str = "michaelrglass/albert-base-rci-wikisql-row", column_model_name_or_path: str = "michaelrglass/albert-base-rci-wikisql-col", row_model_version: Optional[str] = None, column_model_version: Optional[str] = None, row_tokenizer: Optional[str] = None, column_tokenizer: Optional[str] = None, use_gpu: bool = True, top_k: int = 10, max_seq_len: int = 256)

Load an RCI model from Transformers.

Available models include:

  • 'michaelrglass/albert-base-rci-wikisql-row' + 'michaelrglass/albert-base-rci-wikisql-col'
  • 'michaelrglass/albert-base-rci-wtq-row' + 'michaelrglass/albert-base-rci-wtq-col'

Arguments:

  • row_model_name_or_path: Directory of a saved row scoring model or the name of a public model
  • column_model_name_or_path: Directory of a saved column scoring model or the name of a public model
  • row_model_version: The version of row model to use from the HuggingFace model hub. Can be tag name, branch name, or commit hash.
  • column_model_version: The version of column model to use from the HuggingFace model hub. Can be tag name, branch name, or commit hash.
  • row_tokenizer: Name of the tokenizer for the row model (usually the same as model)
  • column_tokenizer: Name of the tokenizer for the column model (usually the same as model)
  • use_gpu: Whether to use GPU or CPU. Falls back on CPU if no GPU is available.
  • top_k: The maximum number of answers to return
  • max_seq_len: Max sequence length of one input table for the model. If the number of tokens of query + table exceed max_seq_len, the table will be truncated by removing rows until the input size fits the model.

RCIReader.predict

def predict(query: str, documents: List[Document], top_k: Optional[int] = None) -> Dict

Use loaded RCI models to find answers for a query in the supplied list of Documents

of content_type 'table'.

Returns dictionary containing query and list of Answer objects sorted by (desc.) score. The existing RCI models on the HF model hub don"t allow aggregation, therefore, the answer will always be composed of a single cell.

Arguments:

  • query: Query string
  • documents: List of Document in which to search for the answer. Documents should be of content_type 'table'.
  • top_k: The maximum number of answers to return

Returns:

Dict containing query and answers