!pip install rouge-score !pip install accelerate>=0.21.0 !pip install datasets import pandas as pd from sklearn.model_selection import train_test_split from transformers import pipeline from nltk.translate.bleu_score import corpus_bleu from rouge_score import rouge_scorer
Load the dataset
dataset_path = "text_summarization_dataset.csv" dataset = pd.read_csv(dataset_path)
Split the dataset into train, validation, and test sets
train_data, test_data = train_test_split(dataset, test_size=0.2, random_state=42) train_data, val_data = train_test_split(train_data, test_size=0.1, random_state=42)
Fine-tune the pre-trained model on the training set
text_summarization_pipeline = pipeline("summarization", model="Paddy007/TextSummaryOp", device="cpu")
Prepare the training data
train_texts = train_data["article"].tolist() train_summaries = train_data["summary"].tolist()
Fine-tune the model
text_summarization_pipeline(train_texts, train_summaries, max_length=150, min_length=30)
Evaluate the fine-tuned model on the validation set
val_texts = val_data["article"].tolist() val_summaries = val_data["summary"].tolist() results = text_summarization_pipeline(val_texts, val_summaries, max_length=150, min_length=30)
Generate text summaries for the test set
test_texts = test_data["article"].tolist() generated_summaries = text_summarization_pipeline(test_texts, max_length=150, min_length=30)
Extract the summary texts from the dictionaries
generated_summaries = [summary['summary_text'] for summary in generated_summaries]
Calculate BLEU scores
reference_texts = test_data["summary"].tolist() bleu_scores = corpus_bleu([[ref] for ref in reference_texts], generated_summaries)
Calculate ROUGE scores
scorer = rouge_scorer.RougeScorer(['rouge1', 'rougeL'], use_stemmer=True) rouge_scores = scorer.score(reference_texts, generated_summaries)
print("BLEU Scores:", bleu_scores) print("ROUGE Scores:", rouge_scores)