haystack/docs/_src/api/api/generator.md

Module base

BaseGenerator

class BaseGenerator(BaseComponent)

Abstract class for Generators

predict

@abstractmethod
def predict(query: str, documents: List[Document], top_k: Optional[int]) -> Dict

Abstract method to generate answers.

Arguments:

• query: Query
• documents: Related documents (e.g. coming from a retriever) that the answer shall be conditioned on.
• top_k: Number of returned answers

Returns:

Generated answers plus additional infos in a dict

Module transformers

RAGenerator

class RAGenerator(BaseGenerator)

Implementation of Facebook's Retrieval-Augmented Generator (https://arxiv.org/abs/2005.11401) based on HuggingFace's transformers (https://huggingface.co/transformers/model_doc/rag.html).

Instead of "finding" the answer within a document, these models generate the answer. In that sense, RAG follows a similar approach as GPT-3 but it comes with two huge advantages for real-world applications: a) it has a manageable model size b) the answer generation is conditioned on retrieved documents, i.e. the model can easily adjust to domain documents even after training has finished (in contrast: GPT-3 relies on the web data seen during training)

Example

|     query = "who got the first nobel prize in physics?"
|
|     # Retrieve related documents from retriever
|     retrieved_docs = retriever.retrieve(query=query)
|
|     # Now generate answer from query and retrieved documents
|     generator.predict(
|        query=query,
|        documents=retrieved_docs,
|        top_k=1
|     )
|
|     # Answer
|
|     {'query': 'who got the first nobel prize in physics',
|      'answers':
|          [{'query': 'who got the first nobel prize in physics',
|            'answer': ' albert einstein',
|            'meta': { 'doc_ids': [...],
|                      'doc_scores': [80.42758 ...],
|                      'doc_probabilities': [40.71379089355469, ...
|                      'content': ['Albert Einstein was a ...]
|                      'titles': ['"Albert Einstein"', ...]
|      }}]}

__init__

def __init__(model_name_or_path: str = "facebook/rag-token-nq", model_version: Optional[str] = None, retriever: Optional[DensePassageRetriever] = None, generator_type: str = "token", top_k: int = 2, max_length: int = 200, min_length: int = 2, num_beams: int = 2, embed_title: bool = True, prefix: Optional[str] = None, use_gpu: bool = True)

Load a RAG model from Transformers along with passage_embedding_model.

See https://huggingface.co/transformers/model_doc/rag.html for more details

Arguments:

• model_name_or_path: Directory of a saved model or the name of a public model e.g. 'facebook/rag-token-nq', 'facebook/rag-sequence-nq'. 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.
• retriever: DensePassageRetriever used to embedded passages for the docs passed to predict(). This is optional and is only needed if the docs you pass don't already contain embeddings in Document.embedding.
• generator_type: Which RAG generator implementation to use ("token" or "sequence")
• top_k: Number of independently generated text to return
• max_length: Maximum length of generated text
• min_length: Minimum length of generated text
• num_beams: Number of beams for beam search. 1 means no beam search.
• embed_title: Embedded the title of passage while generating embedding
• prefix: The prefix used by the generator's tokenizer.
• use_gpu: Whether to use GPU. Falls back on CPU if no GPU is available.

predict

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

Generate the answer to the input query. The generation will be conditioned on the supplied documents.

These document can for example be retrieved via the Retriever.

Arguments:

• query: Query
• documents: Related documents (e.g. coming from a retriever) that the answer shall be conditioned on.
• top_k: Number of returned answers

Returns:

Generated answers plus additional infos in a dict like this:

|     {'query': 'who got the first nobel prize in physics',
|      'answers':
|          [{'query': 'who got the first nobel prize in physics',
|            'answer': ' albert einstein',
|            'meta': { 'doc_ids': [...],
|                      'doc_scores': [80.42758 ...],
|                      'doc_probabilities': [40.71379089355469, ...
|                      'content': ['Albert Einstein was a ...]
|                      'titles': ['"Albert Einstein"', ...]
|      }}]}

Seq2SeqGenerator

class Seq2SeqGenerator(BaseGenerator)

A generic sequence-to-sequence generator based on HuggingFace's transformers.

Text generation is supported by so called auto-regressive language models like GPT2, XLNet, XLM, Bart, T5 and others. In fact, any HuggingFace language model that extends GenerationMixin can be used by Seq2SeqGenerator.

Moreover, as language models prepare model input in their specific encoding, each model specified with model_name_or_path parameter in this Seq2SeqGenerator should have an accompanying model input converter that takes care of prefixes, separator tokens etc. By default, we provide model input converters for a few well-known seq2seq language models (e.g. ELI5). It is the responsibility of Seq2SeqGenerator user to ensure an appropriate model input converter is either already registered or specified on a per-model basis in the Seq2SeqGenerator constructor.

For mode details on custom model input converters refer to _BartEli5Converter

See https://huggingface.co/transformers/main_classes/model.html?transformers.generation_utils.GenerationMixin#transformers.generation_utils.GenerationMixin as well as https://huggingface.co/blog/how-to-generate

For a list of all text-generation models see https://huggingface.co/models?pipeline_tag=text-generation

Example

|     query = "Why is Dothraki language important?"
|
|     # Retrieve related documents from retriever
|     retrieved_docs = retriever.retrieve(query=query)
|
|     # Now generate answer from query and retrieved documents
|     generator.predict(
|        query=query,
|        documents=retrieved_docs,
|        top_k=1
|     )
|
|     # Answer
|
|     {'query': 'who got the first nobel prize in physics',
|      'answers':
|          [{'query': 'who got the first nobel prize in physics',
|            'answer': ' albert einstein',
|            'meta': { 'doc_ids': [...],
|                      'doc_scores': [80.42758 ...],
|                      'doc_probabilities': [40.71379089355469, ...
|                      'content': ['Albert Einstein was a ...]
|                      'titles': ['"Albert Einstein"', ...]
|      }}]}

__init__

def __init__(model_name_or_path: str, input_converter: Optional[Callable] = None, top_k: int = 1, max_length: int = 200, min_length: int = 2, num_beams: int = 8, use_gpu: bool = True)

Arguments:

• model_name_or_path: a HF model name for auto-regressive language model like GPT2, XLNet, XLM, Bart, T5 etc
• input_converter: an optional Callable to prepare model input for the underlying language model specified in model_name_or_path parameter. The required call method signature for the Callable is: call(tokenizer: PreTrainedTokenizer, query: str, documents: List[Document], top_k: Optional[int] = None) -> BatchEncoding:
• top_k: Number of independently generated text to return
• max_length: Maximum length of generated text
• min_length: Minimum length of generated text
• num_beams: Number of beams for beam search. 1 means no beam search.
• use_gpu: Whether to use GPU or the CPU. Falls back on CPU if no GPU is available.

predict

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

Generate the answer to the input query. The generation will be conditioned on the supplied documents.

These document can be retrieved via the Retriever or supplied directly via predict method.

Arguments:

• query: Query
• documents: Related documents (e.g. coming from a retriever) that the answer shall be conditioned on.
• top_k: Number of returned answers

Returns:

Generated answers