Update README.md

README.md

@@ -35,10 +35,10 @@ tags:
 This model is designed to support **phoneme-based text-to-speech (TTS) systems**, including but not limited to [StyleTTS2](https://github.com/yl4579/StyleTTS2). Thanks to its Catalan-specific phoneme vocabulary and contextual embedding capabilities, it can serve as a phoneme encoder for any TTS architecture requiring phoneme-level features.
 
 Features of our PL-BERT:
-- It is trained **exclusively on Catalan** phonemized text.
-- It uses a reduced **phoneme vocabulary of 178 tokens**.
-- It uses a simple tokenizer for words.
-- It includes a custom `token_maps.pkl` and adapted `util.py`.
+- It is trained **exclusively on Catalan** phonemized text
+- It uses a reduced **phoneme vocabulary of 178 tokens**
+- It uses a simple tokenizer for words
+- It includes a custom `token_maps.pkl` and adapted `util.py`
 
 ---
 
@@ -46,15 +46,15 @@ Features of our PL-BERT:
 
 ### Intended uses
 
-- Integration into phoneme-based TTS pipelines such as StyleTTS2, Matxa-TTS, or custom diffusion-based synthesizers.
-- Accent-aware synthesis and phoneme embedding extraction for Catalan.
-- Research on phoneme-level language modeling in low-resource or multi-accent settings.
-
+- Integration into phoneme-based TTS pipelines such as StyleTTS2, Matxa-TTS, or custom diffusion-based synthesizers
+- Accent-aware synthesis and phoneme embedding extraction for Catalan
+- Research on phoneme-level language modeling in low-resource or multi-accent settings
+- 
 ### Limitations
 
-- Not designed for general NLP tasks like classification or sentiment analysis.
-- Only supports Catalan phoneme tokens.
-- Some accents may be underrepresented in the training data.
+- Not designed for general NLP tasks like classification or sentiment analysis
+- Only supports Catalan phoneme tokens
+- Some accents may be underrepresented in the training data
 
 ---
 
@@ -63,7 +63,7 @@ Features of our PL-BERT:
 Here is an example of how to use this model within the StyleTTS2 framework:
 
 1. Clone the StyleTTS2 repository: https://github.com/yl4579/StyleTTS2
-2. Inside the `Utils` directory, create a new folder, for example: `PLBERT_cat_multiaccent`.
+2. Inside the `Utils` directory, create a new folder, for example: `PLBERT_cat_multiaccent`
 3. Copy the following files into that folder:
    - `config.yml` (training configuration)
    - `step_1000000.t7` (trained checkpoint)
@@ -78,7 +78,7 @@ Here is an example of how to use this model within the StyleTTS2 framework:
 
    `from Utils.PLBERT_cat_multiaccent.util import load_plbert`
 
-6. Use `espeak-ng` with the language code `ca` to phonemize your Catalan text files for training and validation.
+6. Phonemize your Catalan text files for training and validation (if you use `espeak-ng` use the language code `ca`)
 
 Note: Although this example uses StyleTTS2, the model is compatible with other TTS architectures that operate on phoneme sequences. You can use the contextualized phoneme embeddings from PL-BERT in any compatible synthesis system.
 
@@ -88,7 +88,7 @@ Note: Although this example uses StyleTTS2, the model is compatible with other T
 
 ### Training data
 
-The model was trained on a Catalan corpus phonemized using espeak-ng. The dataset includes sentences from speakers across Catalonia, Balearic Islands, and Valencia. It uses a consistent phoneme token set with boundary markers and masking tokens.
+The model was trained on a phonemized Catalan corpus (any phonemizer can be used) extracted from the [CATalog](https://huggingface.co/datasets/projecte-aina/CATalog) corpus. The dataset includes sentences from speakers across Catalonia, Balearic Islands, and Valencia. It uses a consistent phoneme token set with boundary markers and masking tokens.
 
 Tokenizer: custom (split using whitespaces)  
 Phoneme masking strategy: word-level and phoneme-level masking and replacement