refactor: move torch/deepfilternet to optional [full] extra

.github/workflows/ci.yml

@@ -54,4 +54,4 @@ jobs:
       - name: Test with unittest
         working-directory: ./tests
         run: |
-          uv run python -m unittest discover -s . -p 'test_*.py'
+          uv run --extra full python -m unittest discover -s . -p 'test_*.py'

README.md

@@ -77,10 +77,10 @@ into `.lrc` subtitles with LLMs such as
 
 ## Installation ⚙️
 
-1. Install CUDA 11.x and [cuDNN 8 for CUDA 11](https://developer.nvidia.com/cudnn) first according
+1. Install [CUDA](https://developer.nvidia.com/cuda-toolkit) and [cuDNN](https://developer.nvidia.com/cudnn) according
    to https://opennmt.net/CTranslate2/installation.html to enable `faster-whisper`.
 
-   `faster-whisper` also needs [cuBLAS for CUDA 11](https://developer.nvidia.com/cublas) installed.
+   `faster-whisper` also needs [cuBLAS](https://developer.nvidia.com/cublas) installed.
    <details>
    <summary>For Windows Users (click to expand)</summary> 
 
@@ -103,7 +103,7 @@ into `.lrc` subtitles with LLMs such as
 3. Install [ffmpeg](https://ffmpeg.org/download.html) and add `bin` directory
    to your `PATH`.
 
-4. This project can be installed from PyPI:
+4. Install from PyPI:
 
     ```shell
     pip install openlrc
@@ -115,20 +115,12 @@ into `.lrc` subtitles with LLMs such as
     pip install git+https://github.com/zh-plus/openlrc
     ```
 
-5. Install the latest [faster-whisper](https://github.com/guillaumekln/faster-whisper) from source:
-   ```shell
-   pip install "faster-whisper @ https://github.com/SYSTRAN/faster-whisper/archive/8327d8cc647266ed66f6cd878cf97eccface7351.tar.gz"
-   ```
+5. **(Optional)** If you need noise suppression (`noise_suppress=True`), install the full extras
+   which includes torch and DeepFilterNet:
 
-6. Install [PyTorch](https://pytorch.org/get-started/locally/):
-   ```shell
-   pip install --force-reinstall torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124
-   ```
-
-7. Fix the `typing-extensions` issue:
-   ```shell
-   pip install typing-extensions -U
-   ```
+    ```shell
+    pip install openlrc[full]
+    ```
 
 ## Lightweight Imports
 
@@ -156,8 +148,9 @@ Heavy dependencies are loaded only when the corresponding features are first use
 - `lingua` is loaded when language detection helpers are used.
 
 > [!NOTE]
-> Lightweight imports improve import-time behavior only. They do not change installation requirements:
-> `pip install openlrc` still installs the full dependency set declared by the package.
+> The base `pip install openlrc` does **not** include torch or DeepFilterNet.
+> These are only installed with `pip install openlrc[full]` and are only needed
+> for noise suppression (`noise_suppress=True`).
 
 ## Usage 🐍
 
@@ -213,7 +206,7 @@ if __name__ == '__main__':
     lrcer = LRCer(transcription=TranscriptionConfig(vad_options=vad_options))
     lrcer.run('./data/test.mp3', target_lang='zh-cn')
 
-    # Enhance the audio using noise suppression (consume more time).
+    # Enhance the audio using noise suppression (requires openlrc[full], consumes more time).
     lrcer.run('./data/test.mp3', target_lang='zh-cn', noise_suppress=True)
 
     # Change the translation model

openlrc/preprocess.py

@@ -61,8 +61,14 @@ def noise_suppression(self, audio_paths: list[Path], atten_lim_db: int = 15):
         if not audio_paths:
             return []
 
-        import torch
-        from df.enhance import enhance, init_df, load_audio, save_audio
+        try:
+            import torch
+            from df.enhance import enhance, init_df, load_audio, save_audio
+        except ImportError:
+            raise ImportError(
+                "Noise suppression requires torch and deepfilternet. "
+                "Install them with: pip install openlrc[full]"
+            )
 
         if "atten_lim_db" in self.options:
             atten_lim_db = self.options["atten_lim_db"]

openlrc/utils.py

@@ -12,7 +12,6 @@
 
 if TYPE_CHECKING:
     from spacy.language import Language as SpacyLanguage
-    import torch
 
 from openlrc.defaults import supported_languages_lingua
 from openlrc.logger import logger
@@ -102,12 +101,14 @@ def get_audio_duration(path: str | Path) -> float:
         return audio.duration
 
 
-def release_memory(model: torch.nn.Module) -> None:
-    import torch
+def release_memory(model: Any) -> None:
+    try:
+        import torch
+    except ImportError:
+        return
 
-    # gc.collect()
-    torch.cuda.empty_cache()
-    del model
+    if isinstance(model, torch.nn.Module):
+        torch.cuda.empty_cache()
 
 
 def normalize(text):

pyproject.toml

@@ -36,15 +36,21 @@ dependencies = [
     "pysbd>=0.3.4,<0.4",
     "faster-whisper>=1.1.1,<2",
     "ffmpeg-normalize>=1.27.5,<2",
-    "deepfilternet>=0.5.6,<0.6",
     "google-genai==1.11.0",
     "json_repair==0.25.2",
     "onnxruntime>=1.20.0,<1.24; python_version < '3.11'",
     "onnxruntime>=1.20.0,<2; python_version >= '3.11'",
+    "pip>=25.1",
+]
+
+[project.optional-dependencies]
+# Noise suppression via DeepFilterNet (requires torch).
+# Install with: pip install openlrc[full]
+# Only needed when using noise_suppress=True in LRCer.run() or Preprocessor.run().
+full = [
     "torch>=2.6.0",
-    "torchvision>=0.21.0",
     "torchaudio>=2.0.0",
-    "pip>=25.1",
+    "deepfilternet>=0.5.6,<0.6",
 ]
 
 [project.urls]