update readme

README.md

@@ -81,9 +81,9 @@ print(tokenizer.decode(outputs[0]))
 ---
 ## Training Details
 ### Training Data
-- Dataset: Claudette To
+- Dataset: Claudette ToS
 - Balanced: 1000 anomalous, 1000 normal clauses
-- Splits: 70% train (1400), 20% val (400), 10% test (200)
+- Splits: 70% train (1400), 20% validation (400), 10% test (200)
 
 ### Training Procedure
 - Quantization: 4-bit (NF4, bitsandbytes)
@@ -122,25 +122,13 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
 ### Model Architecture and Objective
 
 - Base: Saul-7B (LLaMA-style causal LM)
-- LoRA params: ~13M trainable (~0.18% of total)
+- LoRA params: around 13M trainable (approx. 0.18% of total)
 
 ### Compute Infrastructure
 
 - Hardware: 1x NVIDIA Titan X
 - Software: PyTorch 2.2, Transformers 4.51, PEFT 0.15.2, bitsandbytes
 
-
-## Citation 
-
-**APA:**
-@misc{juttu2025saullm7b,
-  author = {Juttu, Noshitha},
-  title = {SaulLM-7B-AnomalyDetector: LoRA Fine-Tuned Model for ToS Anomaly Detection},
-  year = {2025},
-  publisher = {Hugging Face},
-  howpublished = {\url{https://huggingface.co/Noshitha98/SaulLM-7B-AnomalyDetector}}
-}
-
 ## Glossary
 
 - **LoRA (Low-Rank Adaptation):** A parameter-efficient fine-tuning method where only small adapter matrices are trained, while the large base model remains frozen. This drastically reduces compute and storage costs.\
@@ -152,8 +140,8 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
 
 ## Model Card Authors
 
-- **Noshitha Juttu** – M.S. in Computer Science, University of Massachusetts Amherst  
-  Research focus: NLP, Legal AI, and Parameter-Efficient Fine-Tuning (PEFT).  
+- **Noshitha Juttu** – M.S. in Computer Science, University of Massachusetts Amherst
+- Research focus: NLP, model compression, On device NLP and Parameter-Efficient Fine-Tuning (PEFT).  
 
 ## Model Card Contact