Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

refactor: In PromptNode reuse tokenizer instead of loading new one for stop words #4016

Merged
merged 2 commits into from
Feb 1, 2023
Merged

refactor: In PromptNode reuse tokenizer instead of loading new one for stop words #4016

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
123a39f
Reuse tokenizer instead of loading new one.
sjrl Jan 31, 2023
5a3b38a
Merge branch 'main' into prompt_node/stop_word_update
vblagoje Jan 31, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 10 additions & 5 deletions haystack/nodes/prompt/prompt_node.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,14 @@

import requests
import torch
from transformers import pipeline, AutoModelForSeq2SeqLM, StoppingCriteria, StoppingCriteriaList, AutoTokenizer
from transformers import (
pipeline,
AutoModelForSeq2SeqLM,
StoppingCriteria,
StoppingCriteriaList,
PreTrainedTokenizerFast,
PreTrainedTokenizer,
)

from haystack import MultiLabel
from haystack.environment import HAYSTACK_REMOTE_API_BACKOFF_SEC, HAYSTACK_REMOTE_API_MAX_RETRIES
Expand Down Expand Up @@ -215,9 +222,8 @@ class StopWordsCriteria(StoppingCriteria):
Stops text generation if any one of the stop words is generated.
"""

def __init__(self, model_name_or_path: str, stop_words: List[str]):
def __init__(self, tokenizer: Union[PreTrainedTokenizer, PreTrainedTokenizerFast], stop_words: List[str]):
super().__init__()
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
self.stop_words = tokenizer.encode(stop_words, add_special_tokens=False, return_tensors="pt")

def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs) -> bool:
Expand Down Expand Up @@ -245,7 +251,6 @@ def __init__(
"""
Creates an instance of HFLocalInvocationLayer used to invoke local Hugging Face models.


:param model_name_or_path: The name or path of the underlying model.
:param max_length: The maximum length of the output text.
:param use_auth_token: The token to use as HTTP bearer authorization for remote files.
Expand Down Expand Up @@ -342,7 +347,7 @@ def invoke(self, *args, **kwargs):
if key in kwargs
}
if stop_words:
sw = StopWordsCriteria(model_name_or_path=self.model_name_or_path, stop_words=stop_words)
sw = StopWordsCriteria(tokenizer=self.pipe.tokenizer, stop_words=stop_words)
model_input_kwargs["stopping_criteria"] = StoppingCriteriaList([sw])
output = self.pipe(prompt, max_length=self.max_length, **model_input_kwargs)
generated_texts = [o["generated_text"] for o in output if "generated_text" in o]
Expand Down