|
|
|
|
|
import torch |
|
|
from transformers import AutoTokenizer, AutoModelForSequenceClassification |
|
|
|
|
|
class MetaphorScorer: |
|
|
def __init__(self, model_path='.'): |
|
|
""" |
|
|
Initialize the metaphor scorer. |
|
|
|
|
|
Args: |
|
|
model_path: Path or Hugging Face repo ID. |
|
|
Default '.' uses current directory (where model files are) |
|
|
Or use 'your-username/Metaphor_Scoring_Model' to load from Hub |
|
|
""" |
|
|
|
|
|
self.device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') |
|
|
|
|
|
print(f"Loading model from: {model_path}") |
|
|
self.tokenizer = AutoTokenizer.from_pretrained(model_path) |
|
|
self.model = AutoModelForSequenceClassification.from_pretrained(model_path) |
|
|
self.model.to(self.device) |
|
|
self.model.eval() |
|
|
print(f"Model loaded on {self.device}") |
|
|
|
|
|
def score_sentence(self, sentence): |
|
|
""" |
|
|
Score a sentence for metaphorical novelty. |
|
|
|
|
|
Args: |
|
|
sentence: Input sentence to score |
|
|
|
|
|
Returns: |
|
|
score: Novelty score (1-4) |
|
|
confidence: Model confidence (0-1) |
|
|
""" |
|
|
inputs = self.tokenizer( |
|
|
sentence, |
|
|
return_tensors='pt', |
|
|
max_length=256, |
|
|
truncation=True, |
|
|
padding='max_length' |
|
|
) |
|
|
|
|
|
inputs = {k: v.to(self.device) for k, v in inputs.items()} |
|
|
|
|
|
with torch.no_grad(): |
|
|
outputs = self.model(**inputs) |
|
|
logits = outputs.logits |
|
|
predicted_class = torch.argmax(logits, dim=-1).item() |
|
|
score = predicted_class + 1 |
|
|
probabilities = torch.softmax(logits, dim=-1) |
|
|
confidence = probabilities[0][predicted_class].item() |
|
|
|
|
|
return score, confidence |
|
|
|
|
|
def main(): |
|
|
|
|
|
scorer = MetaphorScorer() |
|
|
|
|
|
print("\n=== Metaphorical Sentence Scorer ===") |
|
|
print("Enter metaphorical sentences to get novelty scores (1-4)") |
|
|
print("Higher scores = Higher metaphorical novelty") |
|
|
print("Type 'quit' to exit\n") |
|
|
|
|
|
while True: |
|
|
sentence = input("Enter sentence: ").strip() |
|
|
|
|
|
if sentence.lower() in ['quit', 'exit', 'q']: |
|
|
print("Goodbye!") |
|
|
break |
|
|
|
|
|
if not sentence: |
|
|
continue |
|
|
|
|
|
score, confidence = scorer.score_sentence(sentence) |
|
|
print(f"Score: {score}/4 (confidence: {confidence:.3f})\n") |
|
|
|
|
|
if __name__ == "__main__": |
|
|
main() |
|
|
|
