Update README.md

README.md

@@ -64,31 +64,26 @@ Together, these datasets amount to approximately **4.5 million Kannada sentences
 
 These metrics evaluate tokenizer quality independent of any NLP model.
 
-## **1. Compression Ratio (CR)**  
-Lower = better (fewer bytes used)
-
-## **2. Fertility Score (FS)**  
-Lower = better (#tokens produced per input unit)
-
-## **3. Tokenization Speed**  
-Measured in chars/sec
+## Compression Ratio (CR)**  
+Higher = better  
 
+ 
 ### **Results Across Vocab Sizes**
 
-| Tokenizer | Vocab | CR ↓ | FS ↓ | Speed (chars/s) ↑ |
+| Tokenizer | Vocab | CR | FS  |
 |-----------|-------|------|------|--------------------|
-| **GAT (ours)** | 8k | 2.188 | 2.168 | 14,970 |
-| SentencePiece | 8k | 3.500 | 2.445 | 12,300 |
-| BPE | 8k | 3.300 | 2.311 | 16,081 |
-| WordPiece | 8k | 4.743 | 3.486 | 17,647 |
-| **GAT (ours)** | 16k | 2.400 | 1.986 | 764,647 |
-| SentencePiece | 16k | 3.978 | 1.917 | 970,981 |
-| BPE | 16k | 3.840 | 2.340 | 347,656 |
-| WordPiece | 16k | 4.743 | 2.676 | 401,430 |
-| **GAT (ours)** | 32k | 2.606 | 1.827 | 1,145,656 |
-| SentencePiece | 32k | 4.555 | 1.675 | 620,047 |
-| BPE | 32k | 4.312 | 1.769 | 994,434 |
-| WordPiece | 32k | 4.743 | 1.708 | 1,025,616 |
+| **GAT (ours)** | 8k | 3.588 | 2.168 |  
+| SentencePiece | 8k | 3.100 | 2.445 |  
+| BPE | 8k | 3.300 | 2.711 | 16,081 |
+| WordPiece | 8k | 2.343 | 3.486 | 
+| **GAT (ours)** | 16k | 2.400 | 1.986 |  
+| SentencePiece | 16k | 3.78 | 1.917 |  
+| BPE | 16k | 3.840 | 3.940 | 347,656 |
+| WordPiece | 16k | 3.243 | 2.676 |  
+| **GAT (ours)** | 32k | 4.806 | 1.827 |  
+| SentencePiece | 32k | 3.855 | 1.675 |  
+| BPE | 32k | 3.512 | 1.769 |  
+| WordPiece | 32k | 3.143 | 1.708 |  
 
 # Usage Example