diff --git a/src/intelligence_layer/evaluation/evaluation/graders.py b/src/intelligence_layer/evaluation/evaluation/graders.py index 8eb6aaace..31cbaf510 100644 --- a/src/intelligence_layer/evaluation/evaluation/graders.py +++ b/src/intelligence_layer/evaluation/evaluation/graders.py @@ -1,6 +1,5 @@ import math from dataclasses import dataclass -from multiprocessing import Lock from typing import List, Sequence, Tuple from langdetect import LangDetectException, detect_langs # type: ignore @@ -9,8 +8,6 @@ from sacrebleu import BLEU from semantic_text_splitter import TextSplitter -_evaluate_lock = Lock() - class BleuGrader: def __init__(self) -> None: diff --git a/src/intelligence_layer/examples/summarize/summarize.py b/src/intelligence_layer/examples/summarize/summarize.py index dc1e0d50c..b9b0a4fea 100644 --- a/src/intelligence_layer/examples/summarize/summarize.py +++ b/src/intelligence_layer/examples/summarize/summarize.py @@ -78,7 +78,7 @@ class SummarizeEvaluation(BaseModel): Attributes: bleu: roughly corresponds to precision - rouge: roughly corresponds to the recall + rouge: roughly corresponds to recall output: The actual output from the task run """