Upload 4 files

README.md

@@ -10,13 +10,14 @@ This project provides a Grapheme to Phoneme (G2P) conversion tool that first che
     - **Stress Addition**: The second Transformer model adds stress markers to the phonemes.
 3. **ARPAbet Output**: Outputs phonemes in ARPAbet format.
 4. **Phoneme Integer Indices**: Converts graphemes to phoneme integer indices.
+5. A BPE tokenizer was used, which led to a better translation quality
 
 ## Installation
 
 1. Clone the repository:
     ```sh
-    git clone https://github.com/NikiPshg/G2P_en_lex.git
-    cd G2P_en_lex
+    git clone https://github.com/NikiPshg/Grapheme-to-Phoneme-G2P-with-Stress.git
+    cd Grapheme-to-Phoneme-G2P-with-Stress
     ```
 
 2. Install the required dependencies:
@@ -31,15 +32,17 @@ This project provides a Grapheme to Phoneme (G2P) conversion tool that first che
 from G2P_lexicon import g2p_en_lexicon
 
 # Initialize the G2P converter
-g2p_converter = g2p_en_lexicon()
-
+g2p = g2p_en_lexicon()
 # Convert a word to phonemes
 text = "text, numbers, and some strange symbols !№;% 21"
-phonemes = G2P_en_lex(text, with_stress=False)
-['T', 'EH', 'K', 'S', 'T', ' ', ',', ' ', 'N', 'AH', 'M', 'B', 'ER', 'Z', ' ', ',', ' ',
-'AH', 'N', 'D', ' ', 'S', 'AH', 'M', ' ', 'S', 'T', 'R', 'EY', 'N', 'JH', ' ',
-'S', 'IH', 'M', 'B', 'AH', 'L', 'Z', ' ',
-'T', 'W', 'EH', 'N', 'IY', 'W', 'AH', 'N']
+phonemes = g2p(text, with_stress=False)
+['T', 'EH', 'K', 'S', 'T', ' ', ',', ' ',
+'N', 'AH', 'M', 'B', 'ER', 'Z',' ', ',', ' ', 
+'AE', 'N', 'D', ' ', 'S', 'AH', 'M', ' ',
+'S', 'T', 'R', 'EY', 'N', 'JH',' ', 
+'S', 'IH', 'M', 'B', 'AH', 'L', 'Z',' ', 
+'T', 'W', 'EH', 'N', 'IY', ' ', 'W', 'AH', 'N']
+