Update README.md

README.md

@@ -23,50 +23,28 @@ tags:
 - BERT
 - Transformers
 - Digital Currency
+library_name: transformers
 ---
 
 # CBDC-Type-BERT: Classifying Retail vs Wholesale vs General CBDC Sentences
 
 **A domain-specialized BERT classifier that labels central-bank text about CBDCs into three categories:**
-
 * **Retail CBDC** – statements about a **general-purpose** CBDC for the public (households, merchants, wallets, offline use, legal-tender for everyday payments, holding limits, tiered remuneration, “digital euro/pound/rupee” for citizens, etc.).
 * **Wholesale CBDC** – statements about a **financial-institution** CBDC (RTGS/settlement, DLT platforms, PvP/DvP, tokenised assets/markets, interbank use, central-bank reserves on ledger, etc.).
 * **General/Unspecified** – CBDC mentions that **don’t clearly indicate retail or wholesale** scope, or discuss CBDCs at a conceptual/policy level without specifying the type.
 
-
 **Training data:** 1,417 manually annotated CBDC sentences from BIS central-bank speeches — **Retail CBDC** (543), **Wholesale CBDC** (329), and **General/Unspecified** (545) — split **80/10/10** (train/validation/test) with stratification. 
-
 **Base model:** [`bilalzafar/CentralBank-BERT`](https://huggingface.co/bilalzafar/CentralBank-BERT) - **CentralBank-BERT** is a domain-adapted BERT trained on \~2M sentences (66M tokens) of **central bank speeches** (BIS, 1996–2024). It captures monetary-policy and payments vocabulary far better than generic BERT, which materially helps downstream CBDC classification.
 
 ---
 
 ## Preprocessing, Class Weights & Training
-
-* Text cleaning: minimal.
-* Tokenizer: CentralBank-BERT WordPiece (max length **192**).
-* **Class imbalance:** fewer **Wholesale** examples, hence used **inverse-frequency class weights** in `CrossEntropyLoss` to balance learning:
-  * General ≈ 0.866, Retail ≈ 0.870, Wholesale ≈ 1.436 (computed from train split).
-
-* Optimizer: AdamW; **lr=2e-5**, **weight\_decay=0.01**, **warmup\_ratio=0.1**
-* Batch sizes: train **8**, eval **16**; epochs: **5**; **fp16**
-* Early stopping on validation **macro-F1** (patience=2), best model loaded at end.
-* Hardware: single GPU (Colab).
+Performed light **manual cleaning** (trimming whitespace, normalizing quotes/dashes, de-duplication, dropping nulls) and tokenized with [`bilalzafar/CentralBank-BERT`](https://huggingface.co/bilalzafar/CentralBank-BERT)’s WordPiece (max length **192**). Because **Wholesale** had fewer examples, we applied **inverse-frequency class weights** in `CrossEntropyLoss` to balance learning (train-split weights ≈ General **0.866**, Retail **0.870**, Wholesale **1.436**). The model was fine-tuned with AdamW (lr **2e-5**, weight decay **0.01**, warmup ratio **0.1**), batch sizes **8/16** (train/eval), for **5 epochs** with **fp16** mixed precision. Early stopping monitored validation **macro-F1** (patience = 2), and the best checkpoint was restored at the end. Training ran on a single Colab GPU.
 
 ---
 
 ## Performance & Evaluation
-
-**Held-out test (10%)**
-
-* **Accuracy:** **0.887**
-* **F1 (macro):** **0.898**
-* **F1 (weighted):** **0.887**
-
-Class-wise F1 (test):
-
-* **Retail:** \~0.86
-* **Wholesale:** \~0.97
-* **General:** \~0.86
+On a 10% held-out test set, the model achieved **88.7% accuracy**, **0.898 macro-F1**, and **0.887 weighted-F1**. Class-wise, performance was strong across categories, with **Retail ≈ 0.86 F1**, **Wholesale ≈ 0.97 F1**, and **General ≈ 0.86 F1**, indicating particularly high precision/recall on Wholesale, and balanced, reliable performance on Retail and General.
 
 ---