yujunjun/haystack
1
0
Fork 0
mirror of https://github.com/deepset-ai/haystack.git synced 2026-01-06 12:07:04 +00:00
Code Issues Packages Projects Releases Wiki Activity

Clean OpenAIAnswerGenerator docstrings (#2797)

Browse Source 
* Clean OpenAIAnswerGenerator docstrings

* Incorporate reviewer feedback

Co-authored-by: Sara Zan <sara.zanzottera@deepset.ai>
This commit is contained in:
Branden Chan 2022-07-14 09:35:30 +02:00 committed by GitHub
parent 2a7e333d9a
commit 0388284d71
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 18 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

37
haystack/nodes/answer_generator/openai.py
View File

@ -14,11 +14,11 @@ logger = logging.getLogger(__name__)
class OpenAIAnswerGenerator(BaseGenerator):
"""
Uses the GPT-3 models from the OpenAI API to generate answers based on supplied documents (e.g. from any retriever
in Haystack).
Uses the GPT-3 models from the OpenAI API to generate Answers based on supplied Documents.
These can come from a Retriever or be manually supplied.
To be able to use this node, you need an API key from an active OpenAI account (you can sign-up for an account
[here](https://openai.com/api/)).
To use this Node, you need an API key from an active OpenAI account. You can sign-up for an account
on the [OpenAI API website](https://openai.com/api/)).
"""
def __init__(
@ -39,19 +39,18 @@ class OpenAIAnswerGenerator(BaseGenerator):
:param api_key: Your API key from OpenAI. It is required for this node to work.
:param model: ID of the engine to use for generating the answer. You can select one of `"text-ada-001"`,
`"text-babbage-001"`, `"text-curie-001"`, or `"text-davinci-002"`
(from worst to best + cheapest to most expensive). Please refer to the
[OpenAI Documentation](https://beta.openai.com/docs/models/gpt-3) for more information about the
models.
:param max_tokens: The maximum number of tokens allowed for the generated answer.
:param top_k: Number of generated answers.
(from worst to best and from cheapest to most expensive). For more information about the models,
refer to the [OpenAI Documentation](https://beta.openai.com/docs/models/gpt-3).
:param max_tokens: The maximum number of tokens allowed for the generated Answer.
:param top_k: Number of generated Answers.
:param temperature: What sampling temperature to use. Higher values mean the model will take more risks and
value 0 (argmax sampling) works better for scenarios with a well-defined answer.
:param presence_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear
in the text so far, increasing the model's likelihood to talk about new topics.
:param presence_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they have already appeared
in the text. This increases the model's likelihood to talk about new topics.
:param frequency_penalty: Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing
frequency in the text so far, decreasing the model's likelihood to repeat the same line
verbatim.
:param examples_context: A text snippet containing the contextual information used to generate the answers for
:param examples_context: A text snippet containing the contextual information used to generate the Answers for
the examples you provide.
If not supplied, the default from OpenAPI docs is used:
"In 2017, U.S. life expectancy was 78.6 years."
@ -93,10 +92,10 @@ class OpenAIAnswerGenerator(BaseGenerator):
def predict(self, query: str, documents: List[Document], top_k: Optional[int] = None):
"""
Use loaded QA model to generate answers for a query based on the supplied list of Documents.
Use loaded QA model to generate Answers for a query based on the supplied list of Documents.
Returns dictionaries containing answers.
Be aware that OpenAI doesn't return scores for those answers.
Returns dictionaries containing Answers.
Note that OpenAI doesn't return scores for those Answers.
Example:
```python
@ -111,9 +110,9 @@ class OpenAIAnswerGenerator(BaseGenerator):
```
:param query: Query string
:param documents: List of Document in which to search for the answer
:param top_k: The maximum number of answers to return
:return: Dict containing query and answers
:param documents: List of Documents in which to search for the answer
:param top_k: The maximum number of Answers to return
:return: Dictionary containing query and Answers
"""
if top_k is None:
top_k = self.top_k
@ -147,7 +146,7 @@ class OpenAIAnswerGenerator(BaseGenerator):
def _build_prompt(self, query: str, documents: List[Document]) -> Tuple[str, List[Document]]:
"""
Builds the prompt for the GPT-3 model in order for it to generate an answer.
Builds the prompt for the GPT-3 model so that it can generate an Answer.
"""
example_context = f"===\nContext: {self.examples_context}\n===\n"
example_prompts = "\n---\n".join([f"Q: {question}\nA: {answer}" for question, answer in self.examples])