Initial text-based TTS tokenizer

README.md

@@ -4,7 +4,7 @@ This tokenizer is designed for text-based Text-to-Speech models.
 
 ## Vocabulary Structure
 
-- **Total Size**: 13,016
+- **Total Size**: 20,831
 - **Text Tokens**: ~8,000 (BPE-trained)
 - **Audio Codes**: 12,801 (`<|code_0|>` to `<|code_12800|>`)
 - **Special Tokens**: 22
@@ -54,7 +54,7 @@ BPE subword units trained on English text:
 
 - **Tokenizer Type**: BPE (Byte-Pair Encoding)
 - **Text Vocab Size**: 8,000
-- **Training Data**: 1,000,000+ text samples
+- **Training Data**: 100,000+ text samples
 - **Min Frequency**: 2
 
 ## Model Compatibility