Update README.md

README.md

@@ -1,143 +1,35 @@
 ---
+language: en
 license: mit
 base_model: FacebookAI/roberta-base
 tags:
+- sequence-classification
+- metaphor-scoring
 - roberta
-- metaphor
+- nlp
+task_categories:
 - text-classification
-language:
-- en
 ---
 
-# Metaphor Scoring Model
+# How to Use
 
-RoBERTa-base fine-tuned for metaphorical novelty scoring (1-4 scale).
+1. Install Python
+2. Open Command Prompt and install required packages: Type "pip install torch transformers"
+3. In Command Prompt, go to the folder where you saved "Metaphor_Scoring_Model"
+   Example: cd C:\Downloads\Metaphor_Scoring_Model (You have to add the letters "cd")
+4. Run: Type "python Interactive.py"
 
-## 🚀 Quick Start
+## Model Details
 
-### Installation
-```bash
-pip install transformers torch
-```
-
-### Download and Run
-```bash
-git clone https://huggingface.co/pa90/Metaphor_Scoring_Model
-cd Metaphor_Scoring_Model
-python Interactive.py
-```
-
-### Usage Example
-```
-Enter sentence: Time is money
-Score: 3/4 (confidence: 0.892)
-
-Enter sentence: Life is a journey
-Score: 4/4 (confidence: 0.945)
-
-Enter sentence: quit
-Goodbye!
-```
-
-## 💻 Programmatic Usage
-
-```python
-from transformers import AutoTokenizer, AutoModelForSequenceClassification
-import torch
-
-# Load model and tokenizer
-tokenizer = AutoTokenizer.from_pretrained("pa90/Metaphor_Scoring_Model")
-model = AutoModelForSequenceClassification.from_pretrained("pa90/Metaphor_Scoring_Model")
-
-# Score a sentence
-sentence = "Time is money"
-inputs = tokenizer(
-    sentence, 
-    max_length=256, 
-    truncation=True, 
-    padding='max_length', 
-    return_tensors='pt'
-)
-
-with torch.no_grad():
-    outputs = model(**inputs)
-    predicted_class = torch.argmax(outputs.logits, dim=-1).item()
-    score = predicted_class + 1  # Convert to 1-4 scale
-
-print(f"Metaphor Novelty Score: {score}/4")
-```
-
-## 📊 Model Details
-
-- **Base Model**: [FacebookAI/roberta-base](https://huggingface.co/FacebookAI/roberta-base)
-- **Architecture**: RoBERTa (Robustly Optimized BERT Pretraining Approach)
-- **Parameters**: 125M
-- **Task**: 4-class classification for metaphorical novelty
-- **Input**: Single sentence (max 256 tokens)
-- **Output Scores**: 
+- **Task**: Metaphor novelty scoring (1-4 scale)
+- **Output Scores**:
   - 1: Conventional/literal expression
   - 2: Slightly metaphorical
   - 3: Moderately metaphorical
   - 4: Highly novel metaphor
 
-## 🎯 Use Cases
-
-- Literary analysis
-- Creative writing assistance
-- Figurative language detection
-- Linguistic research
-- Educational tools for teaching metaphors
-
-## 📄 License
-
-This model is released under the **MIT License**.
-
-### Base Model
-RoBERTa-base by Facebook AI is licensed under the MIT License.
-
-### Permissions
-- ✅ Commercial use
-- ✅ Modification
-- ✅ Distribution
-- ✅ Private use
-
-### Conditions
-- Attribution required (see citation below)
-
-## 📚 Citation
-
-If you use this model in your research or application, please cite:
-
-```bibtex
-@misc{metaphor-scoring-model-2024,
-  author = {Your Name},
-  title = {Metaphor Scoring Model: RoBERTa-based Metaphorical Novelty Classifier},
-  year = {2024},
-  publisher = {Hugging Face},
-  howpublished = {\url{https://huggingface.co/pa90/Metaphor_Scoring_Model}}
-}
-```
-
-Please also cite the original RoBERTa paper:
-
-```bibtex
-@article{liu2019roberta,
-  title={RoBERTa: A Robustly Optimized BERT Pretraining Approach},
-  author={Liu, Yinhan and Ott, Myle and Goyal, Naman and Du, Jingfei and Joshi, Mandar and Chen, Danqi and Levy, Omer and Lewis, Mike and Zettlemoyer, Luke and Stoyanov, Veselin},
-  journal={arXiv preprint arXiv:1907.11692},
-  year={2019}
-}
-```
+## Attribution
 
-## 🔧 Technical Details
+This model is fine-tuned from [FacebookAI/roberta-base](https://huggingface.co/FacebookAI/roberta-base), which is licensed under MIT.
 
-### Training
-- Fine-tuned from RoBERTa-base checkpoint
-- Task: 4-class sequence classification
-- Max sequence length: 256 tokens
 
-### Requirements
-```
-transformers>=4.30.0
-torch>=2.0.0
-```