Update README.md

README.md

@@ -1,12 +1,12 @@
-# BERT-Base-Uncased Quantized Model for Text Summarization for Educational Books
+# Text-to-Text Transfer Transformer Quantized Model for Text Summarization for Educational Books
 
-This repository hosts a quantized version of the BERT model, fine-tuned for stock-market-analysis-sentiment-classification tasks. The model has been optimized for efficient deployment while maintaining high accuracy, making it suitable for resource-constrained environments.
+This repository hosts a quantized version of the T5 model, fine-tuned for text summarization tasks. The model has been optimized for efficient deployment while maintaining high accuracy, making it suitable for resource-constrained environments.
 
 ## Model Details
 
-- **Model Architecture:** BERT Base Uncased  
+- **Model Architecture:** T5  
 - **Task:** Text Summarization for Educational Books
-- **Dataset:** Stanford Sentiment Treebank v2 (SST2)  
+- **Dataset:** Hugging Face's `cnn_dailymail'  
 - **Quantization:** Float16  
 - **Fine-tuning Framework:** Hugging Face Transformers  
 
@@ -18,65 +18,62 @@ This repository hosts a quantized version of the BERT model, fine-tuned for stoc
 pip install transformers torch
 ```
 
-
 ### Loading the Model
 
 ```python
-
-from transformers import BertForSequenceClassification, BertTokenizer
+from transformers import T5Tokenizer, T5ForConditionalGeneration
 import torch
 
-# Load quantized model
-quantized_model_path = "AventIQ-AI/text-summarization-for-educational-books"
-quantized_model = BertForSequenceClassification.from_pretrained(quantized_model_path)
-quantized_model.eval()  # Set to evaluation mode
-quantized_model.half()  # Convert model to FP16
-
-# Load tokenizer
-tokenizer = BertTokenizer.from_pretrained("bert-base-uncased")
-
-# Define a test sentence
-test_sentence = "Photosynthesis is the process by which green plants and some other organisms use sunlight to synthesize foods with the help of chlorophyll pigments. The process primarily occurs in the chloroplasts of plant cells. During photosynthesis, plants take in carbon dioxide (CO₂) from the atmosphere and water (H₂O) from the soil. These are converted into glucose (C₆H₁₂O₆) and oxygen (O₂) under the influence of sunlight. The overall chemical reaction can be summarized as: 6CO₂ + 6H₂O + light energy → C₆H₁₂O₆ + 6O₂. This process is crucial not only because it provides food for the plant itself but also because it produces oxygen, which is essential for the survival of most living organisms on Earth. Additionally, it forms the basis of the food chain in almost all ecosystems."
-
-# Tokenize input
-inputs = tokenizer(test_sentence, return_tensors="pt", padding=True, truncation=True, max_length=128)
-
-# Ensure input tensors are in correct dtype
-inputs["input_ids"] = inputs["input_ids"].long()  # Convert to long type
-inputs["attention_mask"] = inputs["attention_mask"].long()  # Convert to long type
-
-# Make prediction
-with torch.no_grad():
-    outputs = quantized_model(**inputs)
+device = "cuda" if torch.cuda.is_available() else "cpu"
 
-# Get predicted class
-predicted_class = torch.argmax(outputs.logits, dim=1).item()
-print(f"Predicted Class: {predicted_class}")
+model_name = "AventIQ-AI/text-summarization-for-educational-books"
+tokenizer = T5Tokenizer.from_pretrained(model_name)
+model = T5ForConditionalGeneration.from_pretrained(model_name).to(device)
 
+def test_summarization(model, tokenizer):
+    user_text = input("\nEnter your text for summarization:\n")
+    input_text = "summarize: " + user_text
+    inputs = tokenizer(input_text, return_tensors="pt", truncation=True, max_length=512).to(device)
 
-label_mapping = {0: "very_negative", 1: "nagative", 2: "neutral", 3: "Positive", 4: "very_positive"}  # Example
+    output = model.generate(
+        **inputs,
+        max_new_tokens=100,
+        num_beams=5,
+        length_penalty=0.8,
+        early_stopping=True
+    )
 
-predicted_label = label_mapping[predicted_class]
-print(f"Predicted Label: {predicted_label}")
+    summary = tokenizer.decode(output[0], skip_special_tokens=True)
+    return summary
 
+print("\n📝 **Model Summary:**")
+print(test_summarization(model, tokenizer))
 ```
 
-## Performance Metrics
+# 📊 ROUGE Evaluation Results
+ 
+After fine-tuning the **T5-Small** model for text summarization, we obtained the following **ROUGE** scores:
 
-- **Accuracy:** 0.82 
+| **Metric**  | **Score**  | **Meaning** |
+|-------------|-----------|-------------|
+| **ROUGE-1** | **0.3061** (~30%) | Measures overlap of **unigrams (single words)** between the reference and generated summary. |
+| **ROUGE-2** | **0.1241** (~12%) | Measures overlap of **bigrams (two-word phrases)**, indicating coherence and fluency. |
+| **ROUGE-L** | **0.2233** (~22%) | Measures **longest matching word sequences**, testing sentence structure preservation. |
+| **ROUGE-Lsum** | **0.2620** (~26%) | Similar to ROUGE-L but optimized for summarization tasks. |
+ 
 
 ## Fine-Tuning Details
 
 ### Dataset
 
-The dataset is taken from Kaggle Stanford Sentiment Treebank v2 (SST2).
+The Hugging Face's `cnn_dailymail` dataset was used, containing the text and their summarization examples.
 
 ### Training
 
-- Number of epochs: 3  
-- Batch size: 8  
+- Number of epochs: 3
+- Batch size: 4  
 - Evaluation strategy: epoch  
-- Learning rate: 2e-5  
+- Learning rate: 3e-5  
 
 ### Quantization
 
@@ -88,7 +85,7 @@ Post-training quantization was applied using PyTorch's built-in quantization fra
 .
 ├── model/               # Contains the quantized model files
 ├── tokenizer_config/    # Tokenizer configuration and vocabulary files
-├── model.safensors/     # Fine Tuned Model
+├── model.safetensors/   # Quantized Model
 ├── README.md            # Model documentation
 ```
 
@@ -99,4 +96,4 @@ Post-training quantization was applied using PyTorch's built-in quantization fra
 
 ## Contributing
 
-Contributions are welcome! Feel free to open an issue or submit a pull request if you have suggestions or improvements.
+Contributions are welcome! Feel free to open an issue or submit a pull request if you have suggestions or improvements.