Clarify language-specific inference modes

README.md

@@ -52,6 +52,15 @@ This model repository hosts the public release checkpoints for **SE-Bridge-TTS**
 
 The released files are CosyVoice2 LLM checkpoints. They are intended to be loaded with a CosyVoice2-compatible checkout and the standard CosyVoice2 base model assets. The base model directory should contain the normal CosyVoice2 configuration and acoustic/vocoder weights, while this repository supplies the Thai or Lao LLM checkpoint.
 
+Recommended inference mode by language:
+
+| Checkpoint | Language | Recommended mode |
+| --- | --- | --- |
+| `thai_tts.pt` | Thai (`th`) | Zero-shot inference with `inference_zero_shot`. |
+| `lao_tts.pt` | Lao (`lo`) | Cross-lingual inference only with `inference_cross_lingual`. |
+
+For this release, Thai is the default zero-shot use case. Lao should be run through the cross-lingual path; do not treat `lao_tts.pt` as a zero-shot checkpoint.
+
 Install or prepare CosyVoice first:
 
 ```bash
@@ -61,7 +70,7 @@ pip install -r requirements.txt
 pip install huggingface_hub torchaudio
 ```
 
-Minimal zero-shot inference example:
+Language-aware inference example:
 
 ```python
 import sys
@@ -80,7 +89,7 @@ from cosyvoice.utils.file_utils import load_wav
 HF_REPO_ID = "isabeth/SE-Bridge-TTS"
 BASE_MODEL_DIR = Path("pretrained_models/CosyVoice2-0.5B")
 
-language = "lao"  # choose "thai" or "lao"
+language = "thai"  # choose "thai" for zero-shot, or "lao" for cross-lingual only
 checkpoint_name = {
     "thai": "thai_tts.pt",
     "lao": "lao_tts.pt",
@@ -100,36 +109,30 @@ state_dict = torch.load(checkpoint_path, map_location="cpu")
 cosyvoice.model.llm.load_state_dict(state_dict, strict=False)
 
 prompt_speech_16k = load_wav("prompt.wav", 16000)
-prompt_text = "Transcript of prompt.wav."
 tts_text = "Text to synthesize in the selected language."
 
-for idx, output in enumerate(
-    cosyvoice.inference_zero_shot(
+if language == "thai":
+    prompt_text = "Transcript of prompt.wav."
+    outputs = cosyvoice.inference_zero_shot(
         tts_text,
         prompt_text,
         prompt_speech_16k,
         stream=False,
     )
-):
-    torchaudio.save(
-        f"se_bridge_tts_{language}_{idx}.wav",
-        output["tts_speech"],
-        cosyvoice.sample_rate,
-    )
-```
-
-For cross-lingual prompting, use the same loaded model and replace the generation loop with:
-
-```python
-for idx, output in enumerate(
-    cosyvoice.inference_cross_lingual(
+    output_prefix = "se_bridge_tts_thai_zero_shot"
+elif language == "lao":
+    outputs = cosyvoice.inference_cross_lingual(
         tts_text,
         prompt_speech_16k,
         stream=False,
     )
-):
+    output_prefix = "se_bridge_tts_lao_cross_lingual"
+else:
+    raise ValueError("language must be either 'thai' or 'lao'")
+
+for idx, output in enumerate(outputs):
     torchaudio.save(
-        f"se_bridge_tts_{language}_cross_lingual_{idx}.wav",
+        f"{output_prefix}_{idx}.wav",
         output["tts_speech"],
         cosyvoice.sample_rate,
     )