""" Emotion detector using the j-hartmann/emotion-english-distilroberta-base model. The pipeline is loaded once at module import time and reused across all calls. Valid emotion labels: joy, sadness, anger, fear, disgust, surprise, neutral """ from transformers import pipeline # Load the model pipeline once at module import time (Requirement 1.5) _pipeline = pipeline( "text-classification", model="j-hartmann/emotion-english-distilroberta-base", top_k=1, ) def detect(text: str) -> tuple[str, float]: """ Classify the dominant emotion in `text`. Returns a tuple of (emotion_label, intensity) where: - emotion_label is one of: joy, sadness, anger, fear, disgust, surprise, neutral - intensity is a float in [0.0, 1.0] representing the model's confidence score Raises: ValueError: if `text` is empty or whitespace-only """ if not text or not text.strip(): raise ValueError( "text must be non-empty: received an empty or whitespace-only string" ) # pipeline returns a list of lists when top_k=1: [[{'label': ..., 'score': ...}]] results = _pipeline(text) top = results[0][0] label: str = top["label"].lower() score: float = float(top["score"]) return (label, score)