scores.py

import torch
import numpy as np
from nltk.translate.bleu_score import sentence_bleu
from transformers import BertTokenizer, BertModel

# Function to Calculate the BLEU score
def calculate_bleu(reference, candidate):
    return sentence_bleu([reference], candidate)

# Function to calculate BERT score
def calculate_bert(reference, candidate):
    tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
    model = BertModel.from_pretrained('bert-base-uncased')

    reference_tokens = tokenizer.tokenize(reference)
    candidate_tokens = tokenizer.tokenize(candidate)

    reference_ids = tokenizer.encode(reference, add_special_tokens=True, max_length=512, truncation=True, return_tensors="pt")
    candidate_ids = tokenizer.encode(candidate, add_special_tokens=True, max_length=512, truncation=True, return_tensors="pt")

    with torch.no_grad():
        reference_outputs = model(reference_ids)
        candidate_outputs = model(candidate_ids)

    reference_embeddings = reference_outputs[0][:, 0, :].numpy()
    candidate_embeddings = candidate_outputs[0][:, 0, :].numpy()

    cosine_similarity = np.dot(reference_embeddings, candidate_embeddings.T) / (np.linalg.norm(reference_embeddings) * np.linalg.norm(candidate_embeddings))
    return np.mean(cosine_similarity)

# Function to calculate minimum edit distance
def min_edit_distance(reference, candidate):
    m = len(reference)
    n = len(candidate)

    dp = [[0] * (n + 1) for _ in range(m + 1)]

    for i in range(m + 1):
        for j in range(n + 1):
            if i == 0:
                dp[i][j] = j
            elif j == 0:
                dp[i][j] = i
            elif reference[i - 1] == candidate[j - 1]:
                dp[i][j] = dp[i - 1][j - 1]
            else:
                dp[i][j] = 1 + min(dp[i][j - 1],         # Insert
                                   dp[i - 1][j],         # Remove
                                   dp[i - 1][j - 1])    # Replace

    return dp[m][n]