update README

README.md

@@ -1,25 +1,32 @@
 ## 🎙️ VoxCPM: Tokenizer-Free TTS for Context-Aware Speech Generation and True-to-Life Voice Cloning
 
 
-[![Project Page](https://img.shields.io/badge/Project%20Page-GitHub-blue)](https://github.com/OpenBMB/VoxCPM/) [![Hugging Face](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-OpenBMB-yellow)](hhttps://huggingface.co/openbmb/VoxCPM-0.5B) [![Live Playground](https://img.shields.io/badge/Live%20PlayGround-Demo-orange)](https://huggingface.co/spaces/OpenBMB/VoxCPM-Demo) [![Samples](https://img.shields.io/badge/Page-Samples-red)](https://thuhcsi.github.io/VoxCPM/)
+[![Project Page](https://img.shields.io/badge/Project%20Page-GitHub-blue)](https://github.com/OpenBMB/VoxCPM/) [![Hugging Face](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-OpenBMB-yellow)](https://huggingface.co/openbmb/VoxCPM-0.5B) [![ModelScope](https://img.shields.io/badge/ModelScope-OpenBMB-purple)](https://modelscope.cn/models/OpenBMB/VoxCPM-0.5B)  [![Live Playground](https://img.shields.io/badge/Live%20PlayGround-Demo-orange)](https://huggingface.co/spaces/OpenBMB/VoxCPM-Demo) [![Samples](https://img.shields.io/badge/Page-Samples-red)](https://openbmb.github.io/VoxCPM-demopage)
+
 
 
 <div align="center">
   <img src="assets/voxcpm_logo.png" alt="VoxCPM Logo" width="40%">
 </div>
 
+<div align="center">
+
+👋 Contact us on [WeChat](assets/wechat.png)
+
+</div>
+
 ## News 
-* [2025.09.16] 🔥 🔥 🔥  We Open Source the VoxCPM-0.5B weights!
+* [2025.09.16] 🔥 🔥 🔥  We Open Source the VoxCPM-0.5B [weights](https://huggingface.co/openbmb/VoxCPM-0.5B)!
 * [2025.09.16] 🎉 🎉 🎉  We Provide the [Gradio PlayGround](https://huggingface.co/spaces/OpenBMB/VoxCPM-Demo) for VoxCPM-0.5B, try it now! 
 
 ## Overview
 
 VoxCPM is a novel tokenizer-free Text-to-Speech (TTS) system that redefines realism in speech synthesis. By modeling speech in a continuous space, it overcomes the limitations of discrete tokenization and enables two flagship capabilities: context-aware speech generation and true-to-life zero-shot voice cloning.
 
-Unlike mainstream approaches that convert speech to discrete tokens, VoxCPM uses an end-to-end diffusion autoregressive architecture that directly generates continuous speech representations from text. Built on [MiniCPM-4](https://huggingface.co/openbmb/MiniCPM4-0.5B), it achieves implicit semantic-acoustic decoupling through hierachical language modeling and FSQ constraints, greatly enhancing both expressiveness and generation stability.
+Unlike mainstream approaches that convert speech to discrete tokens, VoxCPM uses an end-to-end diffusion autoregressive architecture that directly generates continuous speech representations from text. Built on [MiniCPM-4](https://huggingface.co/openbmb/MiniCPM4-0.5B) backbone, it achieves implicit semantic-acoustic decoupling through hierachical language modeling and FSQ constraints, greatly enhancing both expressiveness and generation stability.
 
 <div align="center">
-  <img src="assets/voxcpm_model.png" alt="VoxCPM Model Architecture" width="500">
+  <img src="assets/voxcpm_model.png" alt="VoxCPM Model Architecture" width="90%">
 </div>
 
 
@@ -30,6 +37,8 @@ Unlike mainstream approaches that convert speech to discrete tokens, VoxCPM uses
 
 
 
+
+
 ##  Quick Start
 
 ### 🔧 Install from PyPI
@@ -61,13 +70,13 @@ wav = model.generate(
     text="VoxCPM is an innovative end-to-end TTS model from ModelBest, designed to generate highly expressive speech.",
     prompt_wav_path=None,      # optional: path to a prompt speech for voice cloning
     prompt_text=None,          # optional: reference text
-    cfg_value=2.0,             
+    cfg_value=2.0,             # LM guidance on LocDiT, higher for better adherence to the prompt, but maybe worse
-    inference_timesteps=10,
+    inference_timesteps=10,   # LocDiT inference timesteps, higher for better result, lower for fast speed
-    normalize=True, 
+    normalize=True,           # enable external TN tool
-    denoise=True,
+    denoise=True,             # enable external Denoise tool
-    retry_badcase=True,        # optional: enable retrying mode
+    retry_badcase=True,        # enable retrying mode for some bad cases (unstoppable)
-    retry_badcase_max_times=3,
+    retry_badcase_max_times=3,  # maximum retrying times
-    retry_badcase_ratio_threshold=6.0,
+    retry_badcase_ratio_threshold=6.0, # maximum length restriction for bad case detection (simple but effective), it could be adjusted for slow pace speech
 )
 
 sf.write("output.wav", wav, 16000)
@@ -80,10 +89,10 @@ After installation, the entry point is `voxcpm` (or use `python -m voxcpm.cli`).
 
 ```bash
 # 1) Direct synthesis (single text)
-voxcpm --text "Hello VoxCPM" --output out.wav
+voxcpm --text "VoxCPM is an innovative end-to-end TTS model from ModelBest, designed to generate highly expressive speech." --output out.wav
 
 # 2) Voice cloning (reference audio + transcript)
-voxcpm --text "Hello" \
+voxcpm --text "VoxCPM is an innovative end-to-end TTS model from ModelBest, designed to generate highly expressive speech." \
   --prompt-audio path/to/voice.wav \
   --prompt-text "reference transcript" \
   --output out.wav \
@@ -175,41 +184,41 @@ VoxCPM achieves competitive results on public zero-shot TTS benchmarks:
 | Model | Parameters | Open-Source | test-EN | | test-ZH | | test-Hard | |
 |------|------|------|:------------:|:--:|:------------:|:--:|:-------------:|:--:|
 | | | | WER/%⬇ | SIM/%⬆| CER/%⬇| SIM/%⬆ | CER/%⬇ | SIM/%⬆ |
+| MegaTTS3 | 0.5B | ❌ | 2.79 | 77.1 | 1.52 | 79.0 | - | - |
+| DiTAR | 0.6B | ❌ | 1.69 | 73.5 | 1.02 | 75.3 | - | - |
+| CosyVoice3 | 0.5B | ❌ | 2.02 | 71.8 | 1.16 | 78.0 | 6.08 | 75.8 |
+| CosyVoice3 | 1.5B | ❌ | 2.22 | 72.0 | 1.12 | 78.1 | 5.83 | 75.8 |
+| Seed-TTS | - | ❌ | 2.25 | 76.2 | 1.12 | 79.6 | 7.59 | 77.6 |
+| MiniMax-Speech | - | ❌ | 1.65 | 69.2 | 0.83 | 78.3 | - | - |
 | CosyVoice | 0.3B | ✅ | 4.29 | 60.9 | 3.63 | 72.3 | 11.75 | 70.9 |
-| CosyVoice2 | 0.5B | ✅ | 3.09 | 65.9 | 1.38 | 75.7 | 6.83 | 72.4 |
+| CosyVoice2 | 0.5B | ✅ | 3.09 | 65.9 | 1.38 | 75.7 | **6.83** | 72.4 |
 | F5-TTS | 0.3B | ✅ | 2.00 | 67.0 | 1.53 | 76.0 | 8.67 | 71.3 |
 | SparkTTS | 0.5B | ✅ | 3.14 | 57.3 | 1.54 | 66.0 | - | - |
 | FireRedTTS | 0.5B | ✅ | 3.82 | 46.0 | 1.51 | 63.5 | 17.45 | 62.1 |
 | FireRedTTS-2 | 1.5B | ✅ | 1.95 | 66.5 | 1.14 | 73.6 | - | - |
-| Qwen2.5-Omni | 7B | ✅ | 2.72 | 63.2 | 1.70 | 75.2 | 7.97 | 74.7 |
+| Qwen2.5-Omni | 7B | ✅ | 2.72 | 63.2 | 1.70 | 75.2 | 7.97 | **74.7** |
 | OpenAudio-s1-mini | 0.5B | ✅ | 1.94 | 55.0 | 1.18 | 68.5 | - | - |
 | IndexTTS2 | 1.5B | ✅ | 2.23 | 70.6 | 1.03 | 76.5 | - | - |
 | VibeVoice | 1.5B | ✅ | 3.04 | 68.9 | 1.16 | 74.4 | - | - |
 | HiggsAudio-v2 | 3B | ✅ | 2.44 | 67.7 | 1.50 | 74.0 | - | - |
-| CosyVoice3 | 0.5B | ❌ | 2.02 | 71.8 | 1.16 | 78.0 | 6.08 | 75.8 |
+| **VoxCPM** | 0.5B | ✅ | **1.85** | **72.9** | **0.93** | **77.2** | 8.87 | 73.0 |
-| CosyVoice3 | 1.5B | ❌ | 2.22 | 72.0 | 1.12 | 78.1 | 5.83 | 75.8 |
-| MegaTTS3 | 0.5B | ❌ | 2.79 | 77.1 | 1.52 | 79.0 | - | - |
-| DiTAR | 0.6B | ❌ | 1.69 | 73.5 | 1.02 | 75.3 | - | - |
-| Seed-TTS | - | ❌ | 2.25 | 76.2 | 1.12 | 79.6 | 7.59 | 77.6 |
-| MiniMax-Speech | - | ❌ | 1.65 | 69.2 | 0.83 | 78.3 | - | - |
-| **VoxCPM** | **0.5B** | **✅** | **1.85** | **72.9** | **0.93** | **77.2** | 8.87 | 73.0 |
 
 
 ###  CV3-eval Benchmark
 
-| Model | zh | en | hard-zh | | | hard-en | | | |
+| Model | zh | en | hard-zh | | | hard-en | | |
-|-------|:--:|:--:|:-------:|:--:|:--:|:-------:|:--:|:--:|:--:|
+|-------|:--:|:--:|:-------:|:--:|:--:|:-------:|:--:|:--:|
-| | CER/%⬇ | WER/%⬇ | CER/%⬇ | SIM/%⬆ | DNSMOS⬆ | WER/%⬇ | SIM/%⬆ | DNSMOS⬆ | |
+| | CER/%⬇ | WER/%⬇ | CER/%⬇ | SIM/%⬆ | DNSMOS⬆ | WER/%⬇ | SIM/%⬆ | DNSMOS⬆ |
-| F5-TTS | 5.47 | 8.90 | - | - | - | - | - | - | |
+| F5-TTS | 5.47 | 8.90 | - | - | - | - | - | - |
-| SparkTTS | 5.15 | 11.0 | - | - | - | - | - | - | |
+| SparkTTS | 5.15 | 11.0 | - | - | - | - | - | - |
-| GPT-SoVits | 7.34 | 12.5 | - | - | - | - | - | - | |
+| GPT-SoVits | 7.34 | 12.5 | - | - | - | - | - | - |
-| CosyVoice2 | 4.08 | 6.32 | 12.58 | 72.6 | 3.81 | 11.96 | 66.7 | 3.95 | |
+| CosyVoice2 | 4.08 | 6.32 | 12.58 | 72.6 | 3.81 | 11.96 | 66.7 | 3.95 |
-| OpenAudio-s1-mini | 4.00 | 5.54 | 18.1 | 58.2 | 3.77 | 12.4 | 55.7 | 3.89 | |
+| OpenAudio-s1-mini | 4.00 | 5.54 | 18.1 | 58.2 | 3.77 | 12.4 | 55.7 | 3.89 |
-| IndexTTS2 | 3.58 | 4.45 | 12.8 | 74.6 | 3.65 | fail | fail | fail | |
+| IndexTTS2 | 3.58 | 4.45 | 12.8 | 74.6 | 3.65 | - | - | - |
-| HiggsAudio-v2 | 9.54 | 7.89 | 41.0 | 60.2 | 3.39 | 10.3 | 61.8 | 3.68 | |
+| HiggsAudio-v2 | 9.54 | 7.89 | 41.0 | 60.2 | 3.39 | 10.3 | 61.8 | 3.68 |
-| CosyVoice3-0.5B | 3.89 | 5.24 | 14.15 | 78.6 | 3.75 | 9.04 | 75.9 | 3.92 | |
+| CosyVoice3-0.5B | 3.89 | 5.24 | 14.15 | 78.6 | 3.75 | 9.04 | 75.9 | 3.92 |
-| CosyVoice3-1.5B | 3.91 | 4.99 | 9.77 | 78.5 | 3.79 | 10.55 | 76.1 | 3.95 | |
+| CosyVoice3-1.5B | 3.91 | 4.99 | 9.77 | 78.5 | 3.79 | 10.55 | 76.1 | 3.95 |
-| **VoxCPM** | **3.40** | **4.04** | 12.9 | 66.1 | 3.59 | **7.89** | 64.3 | 3.74 | |
+| **VoxCPM** | **3.40** | **4.04** | 12.9 | 66.1 | 3.59 | **7.89** | 64.3 | 3.74 |
 
 
 
@@ -231,6 +240,13 @@ VoxCPM achieves competitive results on public zero-shot TTS benchmarks:
 
 
 
+## 📝TO-DO List
+Please stay tuned for updates!
+- [ ] Release the VoxCPM technical report.
+- [ ] Support higher sampling rate (next version).
+
+
+
 ## 📄 License
 The VoxCPM model weights and code are open-sourced under the [Apache-2.0](LICENSE) license.
 
@@ -255,6 +271,8 @@ This project is developed by the following institutions:
 
 ## 📚 Citation
 
+The techical report is coming soon, please wait for the release 😊
+
 If you find our model helpful, please consider citing our projects 📝 and staring us ⭐️！
 
 ```bib

app.py

@@ -170,7 +170,7 @@ def create_demo_interface(demo: VoxCPMDemo):
 
         # Pro Tips
         with gr.Accordion("💡 Pro Tips ｜使用建议", open=False, elem_id="acc_tips"):
-            gr.Markdown(f"""
+            gr.Markdown("""
             ### Prompt Speech Enhancement｜参考语音降噪
             - **Enable** to remove background noise for a clean, studio-like voice, with an external ZipEnhancer component.  
               **启用**：通过 ZipEnhancer 组件消除背景噪音，获得更好的音质。

pyproject.toml

@@ -27,22 +27,21 @@ classifiers = [
 ]
 requires-python = ">=3.8"
 dependencies = [
-    "torch==2.5.1",
+    "torch>=2.5.0",
-    "torchaudio==2.5.1",
+    "torchaudio>=2.5.0",
-    "transformers==4.50.1",
+    "transformers>=4.36.2",
     "einops",
     "gradio",
     "inflect",
-    "WeTextProcessing",
     "addict",
-    "modelscope==1.22.0",
+    "wetext",
-    "simplejson",
+    "modelscope>=1.22.0",
-    "datasets==2.18.0",
+    "datasets>=2,<4",
-    "sortedcontainers",
-    "librosa",
     "huggingface-hub",
     "pydantic",
     "tqdm",
+    "simplejson",
+    "sortedcontainers",
     "soundfile",
     "funasr",
     "spaces"
@@ -69,7 +68,7 @@ Documentation = "https://github.com/OpenBMB/VoxCPM#readme"
 
 [tool.setuptools.packages.find]
 where = ["src"]
-include = ["voxcpm"]
+include = ["voxcpm*"]
 
 [tool.setuptools.package-dir]
 "" = "src"

requirements.txt

@@ -1,16 +0,0 @@
-torch==2.5.1
-torchaudio==2.5.1
-transformers==4.50.1
-einops
-gradio
-inflect
-WeTextProcessing
-addicts
-modelscope==1.22.0
-simplejson 
-datasets==2.18.0 
-addicts
-sortedcontainers
-librosa
-huggingface-hub
-spaces

src/voxcpm/core.py

@@ -2,12 +2,8 @@ import torch
 import torchaudio
 import os
 import tempfile
-from modelscope.pipelines import pipeline
-from modelscope.utils.constant import Tasks
 from huggingface_hub import snapshot_download
 from .model.voxcpm import VoxCPMModel
-from .utils.text_normalize import TextNormalizer
-
 
 class VoxCPM:
     def __init__(self,
@@ -27,16 +23,16 @@ class VoxCPM:
         """
         print(f"voxcpm_model_path: {voxcpm_model_path}, zipenhancer_model_path: {zipenhancer_model_path}, enable_denoiser: {enable_denoiser}")
         self.tts_model = VoxCPMModel.from_local(voxcpm_model_path)
-        self.text_normalizer = TextNormalizer()
+        self.text_normalizer = None
         if enable_denoiser and zipenhancer_model_path is not None:
-            self.denoiser = pipeline(
+            from .zipenhancer import ZipEnhancer
-                Tasks.acoustic_noise_suppression,
+            self.denoiser = ZipEnhancer(zipenhancer_model_path)
-                model=zipenhancer_model_path)
         else:
             self.denoiser = None
         print("Warm up VoxCPMModel...")
         self.tts_model.generate(
-            target_text="Hello, this is the first test sentence."
+            target_text="Hello, this is the first test sentence.",
+            max_len=10,
         )
 
     @classmethod
@@ -50,7 +46,7 @@ class VoxCPM:
         """Instantiate ``VoxCPM`` from a Hugging Face Hub snapshot.
 
         Args:
-            hf_model_id: Explicit Hugging Face repository id (e.g. "org/repo").
+            hf_model_id: Explicit Hugging Face repository id (e.g. "org/repo") or local path.
             load_denoiser: Whether to initialize the denoiser pipeline.
             zipenhancer_model_id: Denoiser model id or path for ModelScope
                 acoustic noise suppression.
@@ -67,14 +63,19 @@ class VoxCPM:
                 ``hf_model_id`` is provided.
         """
         repo_id = hf_model_id
-        if not repo_id or repo_id.strip() == "":
+        if not repo_id:
-            raise ValueError("You must provide a valid hf_model_id")
+            raise ValueError("You must provide hf_model_id")
         
-        local_path = snapshot_download(
+        # Load from local path if provided
-            repo_id=repo_id,
+        if os.path.isdir(repo_id):
-            cache_dir=cache_dir,
+            local_path = repo_id
-            local_files_only=local_files_only,
+        else:
-        )
+            # Otherwise, try from_pretrained (Hub); exit on failure
+            local_path = snapshot_download(
+                repo_id=repo_id,
+                cache_dir=cache_dir,
+                local_files_only=local_files_only,
+            )
 
         return cls(
             voxcpm_model_path=local_path,
@@ -82,12 +83,6 @@ class VoxCPM:
             enable_denoiser=load_denoiser,
         )
 
-    def _normalize_loudness(self, wav_path: str):
-        audio, sr = torchaudio.load(wav_path)
-        loudness = torchaudio.functional.loudness(audio, sr)
-        normalized_audio = torchaudio.functional.gain(audio, -20-loudness)
-        torchaudio.save(wav_path, normalized_audio, sr)
-
     def generate(self, 
             text : str,
             prompt_wav_path : str = None,
@@ -135,9 +130,7 @@ class VoxCPM:
                 if denoise and self.denoiser is not None:
                     with tempfile.NamedTemporaryFile(delete=False, suffix='.wav') as tmp_file:
                         temp_prompt_wav_path = tmp_file.name
-                    
+                    self.denoiser.enhance(prompt_wav_path, output_path=temp_prompt_wav_path)
-                    self.denoiser(prompt_wav_path, output_path=temp_prompt_wav_path)
-                    self._normalize_loudness(temp_prompt_wav_path)
                     prompt_wav_path = temp_prompt_wav_path
                 fixed_prompt_cache = self.tts_model.build_prompt_cache(
                     prompt_wav_path=prompt_wav_path,
@@ -151,6 +144,9 @@ class VoxCPM:
                     continue
                 print("sub_text:", sub_text)
                 if normalize:
+                    if self.text_normalizer is None:
+                        from .utils.text_normalize import TextNormalizer
+                        self.text_normalizer = TextNormalizer()
                     sub_text = self.text_normalizer.normalize(sub_text)
                 wav, target_text_token, generated_audio_feat = self.tts_model.generate_with_prompt_cache(
                                 target_text=sub_text,

src/voxcpm/model/voxcpm.py

@@ -148,12 +148,15 @@ class VoxCPMModel(nn.Module):
 
     
     def optimize(self):
-        if self.device == "cuda":
+        try:
+            if self.device != "cuda":
+                raise ValueError("VoxCPMModel can only be optimized on CUDA device")
             self.base_lm.forward_step = torch.compile(self.base_lm.forward_step, mode="reduce-overhead", fullgraph=True)
             self.residual_lm.forward_step = torch.compile(self.residual_lm.forward_step, mode="reduce-overhead", fullgraph=True)
             self.feat_encoder_step = torch.compile(self.feat_encoder, mode="reduce-overhead", fullgraph=True)
             self.feat_decoder.estimator = torch.compile(self.feat_decoder.estimator, mode="reduce-overhead", fullgraph=True)
-        else:
+        except:
+            print("VoxCPMModel can not be optimized by torch.compile, using original forward_step functions")
             self.base_lm.forward_step = self.base_lm.forward_step
             self.residual_lm.forward_step = self.residual_lm.forward_step
             self.feat_encoder_step = self.feat_encoder

src/voxcpm/modules/locdit/unified_cfm.py

@@ -88,7 +88,7 @@ class UnifiedCFM(torch.nn.Module):
                 shape: (n_timesteps + 1,)
             mu (torch.Tensor): output of encoder
                 shape: (batch_size, n_feats)
-            cond: Not used but kept for future purposes
+            cond: condition -- prefix prompt
             cfg_value (float, optional): cfg value for guidance. Defaults to 1.0.
         """
         t, _, dt = t_span[0], t_span[-1], t_span[0] - t_span[1]

src/voxcpm/utils/text_normalize.py

@@ -3,40 +3,7 @@ import re
 import regex
 import inflect
 from functools import partial
-from tn.chinese.normalizer import Normalizer as ZhNormalizer
+from wetext import Normalizer
-from tn.english.normalizer import Normalizer as EnNormalizer
-
-def normal_cut_sentence(text):
-    # 先处理括号内的逗号，将其替换为特殊标记
-    text = re.sub(r'([（(][^）)]*)([,，])([^）)]*[）)])', r'\1&&&\3', text)
-    text = re.sub('([。！，？\?])([^’”])',r'\1\n\2',text)#普通断句符号且后面没有引号
-    text = re.sub('(\.{6})([^’”])',r'\1\n\2',text)#英文省略号且后面没有引号
-    text = re.sub('(\…{2})([^’”])',r'\1\n\2',text)#中文省略号且后面没有引号
-    text = re.sub('([. ,。！；？\?\.{6}\…{2}][’”])([^’”])',r'\1\n\2',text)#断句号+引号且后面没有引号
-    # 处理英文句子的分隔
-    text = re.sub(r'([.,!?])([^’”\'"])', r'\1\n\2', text)  # 句号、感叹号、问号后面没有引号
-    text = re.sub(r'([.!?][’”\'"])([^’”\'"])', r'\1\n\2', text)  # 句号、感叹号、问号加引号后面的部分
-    text = re.sub(r'([（(][^）)]*)(&&&)([^）)]*[）)])', r'\1，\3', text)
-    text = [t for t in text.split("\n") if t]
-    return text
-
-
-def cut_sentence_with_fix_length(text : str, length : int):
-    sentences = normal_cut_sentence(text)
-    cur_length = 0
-    res = ""
-    for sentence in sentences:
-        if not sentence:
-            continue
-        if cur_length > length or cur_length + len(sentence) > length:
-            yield res
-            res = ""
-            cur_length = 0
-        res += sentence
-        cur_length += len(sentence)
-    if res:
-        yield res
-
 
 chinese_char_pattern = re.compile(r'[\u4e00-\u9fff]+')
 
@@ -195,8 +162,8 @@ def clean_text(text):
 class TextNormalizer:
     def __init__(self, tokenizer=None):
         self.tokenizer = tokenizer
-        self.zh_tn_model = ZhNormalizer(remove_erhua=False, full_to_half=False, remove_interjections=False, overwrite_cache=True)
+        self.zh_tn_model = Normalizer(lang="zh", operator="tn", remove_erhua=True)
-        self.en_tn_model = EnNormalizer()
+        self.en_tn_model = Normalizer(lang="en", operator="tn")
         self.inflect_parser = inflect.engine()
     
     def normalize(self, text, split=False):
@@ -207,38 +174,12 @@ class TextNormalizer:
             text = text.replace("=", "等于") # 修复 ”550 + 320 等于 870 千卡。“ 被错误正则为 ”五百五十加三百二十等于八七十千卡.“
             if re.search(r'([\d$%^*_+≥≤≠×÷?=])', text): # 避免 英文连字符被错误正则为减
                 text = re.sub(r'(?<=[a-zA-Z0-9])-(?=\d)', ' - ', text) # 修复 x-2 被正则为 x负2
-                text = self.zh_tn_model.normalize(text)
-            text = re.sub(r'(?<=[a-zA-Z0-9])-(?=\d)', ' - ', text) # 修复 x-2 被正则为 x负2
             text = self.zh_tn_model.normalize(text)
             text = replace_blank(text)
             text = replace_corner_mark(text)
             text = remove_bracket(text)
-            text = re.sub(r'[，,]+$', '。', text)
         else:
             text = self.en_tn_model.normalize(text)
             text = spell_out_number(text, self.inflect_parser)
         if split is False:
             return text
-        
-        
-if __name__ == "__main__":
-    text_normalizer = TextNormalizer()
-    text = r"""今天我们学习一元二次方程。一元二次方程的标准形式是：
-ax2+bx+c=0ax^2 + bx + c = 0ax2+bx+c=0 
-其中，aaa、bbb 和 ccc 是常数，xxx 是变量。这个方程的解可以通过求根公式来找到。
-一元二次方程的解法有几种：
-  - 因式分解法：通过将方程因式分解来求解。我们首先尝试将方程表达成两个括号的形式，解决方程的解。比如，方程x2−5x+6=0x^2 - 5x + 6 = 0x2−5x+6=0可以因式分解为(x−2)(x−3)=0(x - 2)(x - 3) = 0(x−2)(x−3)=0，因此根为2和3。
-  - 配方法：通过配方将方程转化为完全平方的形式，从而解出。我们通过加上或减去适当的常数来完成这一过程，使得方程可以直接写成一个完全平方的形式。
-  - 求根公式：我们可以使用求根公式直接求出方程的解。这个公式适用于所有的一元二次方程，即使我们无法通过因式分解或配方法来解决时，也能使用该公式。
-公式：x=−b±b2−4ac2ax = \frac{-b \pm \sqrt{b^2 - 4ac}}{2a}x=2a−b±b2−4ac​​这个公式可以帮助我们求解任何一元二次方程的根。
-对于一元二次方程，我们需要了解判别式。判别式的作用是帮助我们判断方程的解的个数和性质。判别式 Δ\DeltaΔ 由下式给出：Δ=b2−4ac\Delta = b^2 - 4acΔ=b2−4ac 根据判别式的值，我们可以知道：
-  - 如果 Δ>0\Delta > 0Δ>0，方程有两个不相等的实数解。这是因为判别式大于0时，根号内的值是正数，所以我们可以得到两个不同的解。
-  - 如果 Δ=0\Delta = 0Δ=0，方程有一个实数解。这是因为根号内的值为零，导致两个解相等，也就是说方程有一个解。
-  - 如果 Δ<0\Delta < 0Δ<0，方程没有实数解。这意味着根号内的值是负数，无法进行实数运算，因此方程没有实数解，可能有复数解。"""
-    texts = ["这是一个公式 (a+b)³=a³+3a²b+3ab²+b³ S=(a×b)÷2", "这样的发展为AI仅仅作为“工具”这一观点提出了新的挑战，", "550 + 320 = 870千卡。", "解一元二次方程：3x^2+x-2=0", "你好啊"]
-    texts = [text]
-    for text in texts:
-        text = text_normalizer.normalize(text)
-        print(text)
-        for t in cut_sentence_with_fix_length(text, 15):
-            print(t)

src/voxcpm/zipenhancer.py

@@ -0,0 +1,76 @@
+"""
+ZipEnhancer Module - Audio Denoising Enhancer
+
+Provides on-demand import ZipEnhancer functionality for audio denoising processing.
+Related dependencies are imported only when denoising functionality is needed.
+"""
+
+import os
+import tempfile
+from typing import Optional, Union
+import torchaudio
+import torch
+from modelscope.pipelines import pipeline
+from modelscope.utils.constant import Tasks
+
+
+class ZipEnhancer:
+    """ZipEnhancer Audio Denoising Enhancer"""
+    def __init__(self, model_path: str = "iic/speech_zipenhancer_ans_multiloss_16k_base"):
+        """
+        Initialize ZipEnhancer
+        Args:
+            model_path: ModelScope model path or local path
+        """
+        self.model_path = model_path
+        self._pipeline = pipeline(
+                Tasks.acoustic_noise_suppression,
+                model=self.model_path
+            )
+        
+    def _normalize_loudness(self, wav_path: str):
+        """
+        Audio loudness normalization
+        
+        Args:
+            wav_path: Audio file path
+        """
+        audio, sr = torchaudio.load(wav_path)
+        loudness = torchaudio.functional.loudness(audio, sr)
+        normalized_audio = torchaudio.functional.gain(audio, -20-loudness)
+        torchaudio.save(wav_path, normalized_audio, sr)
+    
+    def enhance(self, input_path: str, output_path: Optional[str] = None, 
+                normalize_loudness: bool = True) -> str:
+        """
+        Audio denoising enhancement
+        Args:
+            input_path: Input audio file path
+            output_path: Output audio file path (optional, creates temp file by default)
+            normalize_loudness: Whether to perform loudness normalization
+        Returns:
+            str: Output audio file path
+        Raises:
+            RuntimeError: If pipeline is not initialized or processing fails
+        """
+        if not os.path.exists(input_path):
+            raise FileNotFoundError(f"Input audio file does not exist: {input_path}")
+        # Create temporary file if no output path is specified
+        if output_path is None:
+            with tempfile.NamedTemporaryFile(delete=False, suffix='.wav') as tmp_file:
+                output_path = tmp_file.name
+        try:
+            # Perform denoising processing
+            self._pipeline(input_path, output_path=output_path)
+            # Loudness normalization
+            if normalize_loudness:
+                self._normalize_loudness(output_path)
+            return output_path
+        except Exception as e:
+            # Clean up possibly created temporary files
+            if output_path and os.path.exists(output_path):
+                try:
+                    os.unlink(output_path)
+                except OSError:
+                    pass
+            raise RuntimeError(f"Audio denoising processing failed: {e}")