Add tests

.pre-commit-config.yaml

@@ -28,7 +28,7 @@ repos:
       - id: ruff
         name: ruff
         language: system
-        entry: ruff
+        entry: ruff --fix
         types: [python]
       - id: mypy
         name: mypy

README.md

@@ -3,6 +3,7 @@
 Denoisers is a denoising library for audio with a focus on simplicity and ease of use. There are two major types of architectures available. WaveUNet for waveform denoising and UNet for spectrogram denoising.
 
 ## Demo
+
 Gradio demo [here](https://wrice-denoisers.hf.space)
 
 ## Usage/Examples
@@ -11,7 +12,6 @@ Gradio demo [here](https://wrice-denoisers.hf.space)
 pip install denoisers
 ```
 
-
 ```python
 import torch
 import torchaudio

denoisers/modeling/modules.py

@@ -195,53 +195,6 @@ def forward(self, x: Tensor) -> Tensor:
         return x
 
 
-class ResidualDownBlock1D(nn.Module):
-    """1d residual down block."""
-
-    def __init__(
-        self,
-        in_channels: int,
-        out_channels: int,
-        kernel_size: int = 3,
-        activation: str = "leaky_relu",
-        dropout: float = 0.0,
-    ) -> None:
-        super().__init__()
-        self.conv_1 = DownsampleBlock1D(
-            in_channels=in_channels,
-            out_channels=in_channels,
-            kernel_size=kernel_size,
-            dropout=dropout,
-            activation=activation,
-            bias=False,
-        )
-
-        self.conv_2 = DownsampleBlock1D(
-            in_channels=in_channels,
-            out_channels=out_channels,
-            kernel_size=kernel_size,
-            dropout=dropout,
-            activation=activation,
-            bias=False,
-        )
-        self.res_conv = Downsample1D(
-            in_channels=in_channels,
-            out_channels=out_channels,
-            kernel_size=1,
-            stride=2,
-            use_conv=True,
-            bias=False,
-        )
-
-    def forward(self, x: Tensor) -> Tensor:
-        """Forward pass."""
-        residual = x
-        x = self.conv_1(x)
-        x = self.conv_2(x)
-        x = x + self.res_conv(residual)
-        return x
-
-
 class Activation(nn.Module):
     """Activation function."""
 

denoisers/testing/__init__.py

@@ -0,0 +1,19 @@
+"""Utility functions for tests."""
+import torch
+from torch import Tensor
+
+
+def sine_wave(frequency: float, duration: float, sample_rate: int) -> Tensor:
+    """Generate a sine wave.
+
+    Args:
+        frequency: Frequency of the sine wave.
+        duration: Duration of the sine wave in seconds.
+        sample_rate: Sample rate of the sine wave.
+
+    Returns:
+        A torch tensor containing the sine wave.
+    """
+    return torch.sin(
+        2 * torch.pi * torch.arange(sample_rate * duration) * frequency / sample_rate
+    ).unsqueeze(0)