fix: 改用soundfile保存音频，解决torchaudio格式问题

README.md

@@ -30,15 +30,29 @@ pip install -r requirements.txt
 ### 2. 启动服务
 
 ```bash
-# 默认端口 12002
+# 方式一：自动下载模型
 python3 server.py
 
-# 或使用脚本
-PORT=12002 ./start.sh
+# 方式二：使用已下载的模型（指定路径）
+MODEL_PATH=/path/to/ChatTTS python3 server.py
+
+# 方式三：使用本地缓存模型
+MODEL_SOURCE=local python3 server.py
 ```
 
 首次启动会自动下载 ChatTTS 模型（约 2GB）。
 
+**模型路径配置**：
+
+| 环境变量 | 说明 | 示例 |
+|---------|------|------|
+| `MODEL_PATH` | 模型目录路径 | `/data/models/ChatTTS` |
+| `MODEL_SOURCE` | 加载方式 | `auto` / `local` / `download` |
+
+ChatTTS 模型默认下载位置：
+- Linux: `~/.cache/modelscope/hub/ChatTTS/`
+- 或手动下载后解压到任意目录
+
 ### 3. 验证服务
 
 ```bash

requirements.txt

@@ -4,4 +4,5 @@ python-multipart==0.0.9
 torch==2.2.0
 torchaudio==2.2.0
 transformers==4.38.0
-ChatTTS==0.1.1
+ChatTTS
+soundfile==0.12.1

server.py

@@ -24,6 +24,13 @@ from pydantic import BaseModel
 PORT = int(os.getenv("PORT", "12002"))
 SAMPLE_RATE = 24000  # ChatTTS 默认采样率
 AUDIO_DIR = os.getenv("AUDIO_DIR", "audio_output")
+
+# ChatTTS 模型路径配置
+# 如果已下载模型，设置 MODEL_PATH 环境变量
+# 例如: MODEL_PATH=/path/to/ChatTTS
+MODEL_PATH = os.getenv("MODEL_PATH", None)
+MODEL_SOURCE = os.getenv("MODEL_SOURCE", "auto")  # auto / local / download
+
 os.makedirs(AUDIO_DIR, exist_ok=True)
 
 # 日志
@@ -74,13 +81,33 @@ def load_model():
             import ChatTTS
             chat_model = ChatTTS.Chat()
             
-            # 加载模型
-            # 方式一：从本地加载（如果有）
-            # 方式二：自动下载
-            chat_model.load(
-                compile=True,  # 编译优化
-                device="cuda" if torch.cuda.is_available() else "cpu"
-            )
+            device = "cuda" if torch.cuda.is_available() else "cpu"
+            logger.info(f"Using device: {device}")
+            
+            # 加载模型方式
+            if MODEL_PATH and os.path.exists(MODEL_PATH):
+                # 从本地路径加载
+                logger.info(f"Loading model from: {MODEL_PATH}")
+                chat_model.load(
+                    source=MODEL_PATH,
+                    compile=True,
+                    device=device
+                )
+            elif MODEL_SOURCE == "local":
+                # 从默认本地缓存加载
+                logger.info("Loading from local cache...")
+                chat_model.load(
+                    compile=True,
+                    device=device,
+                    source="local"
+                )
+            else:
+                # 自动下载
+                logger.info("Auto downloading model...")
+                chat_model.load(
+                    compile=True,
+                    device=device
+                )
             
             logger.info("ChatTTS model loaded successfully")
         except Exception as e:
@@ -97,11 +124,15 @@ def save_audio(audio_tensor: torch.Tensor, filename: str) -> str:
     filepath = os.path.join(AUDIO_DIR, filename)
     
     # 确保 tensor 正确形状
-    if audio_tensor.dim() == 1:
-        audio_tensor = audio_tensor.unsqueeze(0)
+    if audio_tensor.dim() == 2:
+        audio_tensor = audio_tensor.squeeze(0)
     
-    # 保存为 WAV
-    torchaudio.save(filepath, audio_tensor, SAMPLE_RATE, format="wav")
+    # 转换为 numpy
+    audio_np = audio_tensor.cpu().numpy() if audio_tensor.is_cuda else audio_tensor.numpy()
+    
+    # 使用 soundfile 保存
+    import soundfile as sf
+    sf.write(filepath, audio_np, SAMPLE_RATE)
     
     return filepath
 
@@ -158,22 +189,29 @@ async def synthesize(
         # 生成唯一文件名
         filename = f"{uuid.uuid4().hex}.wav"
         
-        # 合成参数
-        params = {
-            'temperature': temperature,
-            'top_p': top_p,
-            'top_k': top_k,
-            'spk_emb': None,  # 可选：说话人嵌入
-        }
-        
         # 合成语音
         logger.info(f"Synthesizing: {text[:50]}...")
         
-        # ChatTTS 生成
-        audio_tensor = model.infer(
-            [text],
-            params=params
-        )[0]  # 返回是列表，取第一个
+        # ChatTTS 基本调用（简化版）
+        # 返回: list of audio tensors
+        result = model.infer(text)
+        
+        # 处理返回结果
+        if isinstance(result, list):
+            audio_tensor = result[0]
+        elif isinstance(result, tuple):
+            audio_tensor = result[0]
+        else:
+            audio_tensor = result
+        
+        # 转换为 torch tensor（如果是 numpy）
+        import numpy as np
+        if isinstance(audio_tensor, np.ndarray):
+            audio_tensor = torch.from_numpy(audio_tensor).float()
+        
+        # 确保 tensor 正确形状
+        if audio_tensor.dim() == 1:
+            audio_tensor = audio_tensor.unsqueeze(0)
         
         # 保存音频
         filepath = save_audio(audio_tensor, filename)
@@ -208,14 +246,15 @@ async def synthesize_batch(requests: list[SynthesizeRequest]):
         texts = [r.text for r in requests]
         
         # 统一参数
-        params = {
-            'temperature': requests[0].temperature,
-            'top_p': requests[0].top_p,
-            'top_k': requests[0].top_k,
-        }
+        infer_params = {}
         
         # 批量生成
-        audio_tensors = model.infer(texts, params=params)
+        audio_tensors = model.infer(
+            texts,
+            temperature=requests[0].temperature,
+            top_P=requests[0].top_p,
+            top_K=requests[0].top_k,
+        )
         
         results = []
         for i, audio_tensor in enumerate(audio_tensors):
@@ -310,6 +349,19 @@ async def synthesize_with_emotion(
         raise HTTPException(status_code=500, detail=str(e))
 
 
+if __name__ == "__main__":
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=PORT)  audio_url=f"/audio/{filename}",
+            duration=round(duration, 2),
+            text=text,
+            timestamp=datetime.now().isoformat()
+        )
+        
+    except Exception as e:
+        logger.error(f"Emotion synthesis error: {e}")
+        raise HTTPException(status_code=500, detail=str(e))
+
+
 if __name__ == "__main__":
     import uvicorn
     uvicorn.run(app, host="0.0.0.0", port=PORT)

start.sh

@@ -4,8 +4,17 @@
 cd "$(dirname "$0")"
 
 PORT=${PORT:-12002}
+MODEL_PATH=${MODEL_PATH:-""}
+MODEL_SOURCE=${MODEL_SOURCE:-"auto"}
 
 echo "Starting ChatTTS service on port $PORT..."
 echo "Device: $(python3 -c 'import torch; print("cuda" if torch.cuda.is_available() else "cpu")')"
 
+if [ -n "$MODEL_PATH" ]; then
+    echo "Model path: $MODEL_PATH"
+fi
+
+export MODEL_PATH
+export MODEL_SOURCE
+
 python3 server.py