⚗️ | Small fixes here and there

.gitignore

@@ -166,3 +166,4 @@ dataset/
 old-output/
 output/
 *.wav
+models/

AudioUtils.py

@@ -0,0 +1,40 @@
+import torch
+import torch.nn.functional as F
+
+
+def stereo_tensor_to_mono(waveform: torch.Tensor) -> torch.Tensor:
+    mono_tensor = torch.mean(waveform, dim=0, keepdim=True)
+    return mono_tensor
+
+
+def pad_tensor(audio_tensor: torch.Tensor, target_length: int = 512) -> torch.Tensor:
+    current = audio_tensor.size(-1)
+    padding_amount = target_length - current
+    if padding_amount <= 0:
+        return audio_tensor
+
+    return F.pad(audio_tensor, (0, padding_amount))
+
+
+def split_audio(audio_tensor: torch.Tensor, chunk_size: int = 512, pad_last_tensor: bool = False) -> list[torch.Tensor]:
+    chunks = list(torch.split(audio_tensor, chunk_size, dim=1))
+
+    if pad_last_tensor:
+        last_chunk = chunks[-1]
+
+        if last_chunk.size(-1) < chunk_size:
+                    chunks[-1] = pad_tensor(last_chunk, chunk_size)
+
+    return chunks
+
+
+def reconstruct_audio(chunks: list[torch.Tensor]) -> torch.Tensor:
+    reconstructed_tensor = torch.cat(chunks, dim=-1)
+    return reconstructed_tensor
+
+
+def normalize(audio_tensor: torch.Tensor, eps: float = 1e-8) -> torch.Tensor:
+    max_val = torch.max(torch.abs(audio_tensor))
+    if max_val < eps:
+        return audio_tensor
+    return audio_tensor / max_val

README.md

@@ -18,6 +18,7 @@ SISU (Super Ingenious Sound Upscaler) is a project that uses GANs (Generative Ad
 1. **Set Up**:
    - Make sure you have Python installed (version 3.8 or higher).
    - Install needed packages: `pip install -r requirements.txt`
+   - Install current version of PyTorch (CUDA/ROCm/What ever your device supports)
 
 2. **Prepare Audio Data**:
    - Put your audio files in the `dataset/good` folder.

app.py

@@ -0,0 +1,128 @@
+import argparse
+
+import torch
+import torchaudio
+import torchcodec
+import tqdm
+from accelerate import Accelerator
+
+import AudioUtils
+from generator import SISUGenerator
+
+# Init script argument parser
+parser = argparse.ArgumentParser(description="Training script")
+parser.add_argument("--device", type=str, default="cpu", help="Select device")
+parser.add_argument("--model", type=str, help="Model to use for upscaling")
+parser.add_argument(
+    "--clip_length",
+    type=int,
+    default=8000,
+    help="Internal clip length, leave unspecified if unsure",
+)
+parser.add_argument(
+    "--sample_rate", type=int, default=44100, help="Output clip sample rate"
+)
+parser.add_argument(
+    "--bitrate",
+    type=int,
+    default=192000,
+    help="Output clip bitrate",
+)
+parser.add_argument("-i", "--input", type=str, help="Input audio file")
+parser.add_argument("-o", "--output", type=str, help="Output audio file")
+
+args = parser.parse_args()
+
+if args.sample_rate < 8000:
+    print(
+        "Sample rate cannot be lower than 8000! (44100 is recommended for base models)"
+    )
+    exit()
+
+# ---------------------------
+# Init accelerator
+# ---------------------------
+
+accelerator = Accelerator(mixed_precision="bf16")
+
+# ---------------------------
+# Models
+# ---------------------------
+generator = SISUGenerator()
+
+accelerator.print("🔨 | Compiling models...")
+
+generator = torch.compile(generator)
+
+accelerator.print("✅ | Compiling done!")
+
+# ---------------------------
+# Prepare accelerator
+# ---------------------------
+
+generator = accelerator.prepare(generator)
+
+# ---------------------------
+# Checkpoint helpers
+# ---------------------------
+
+models_dir = args.model
+clip_length = args.clip_length
+input_audio = args.input
+output_audio = args.output
+
+if models_dir:
+    ckpt = torch.load(models_dir)
+
+    accelerator.unwrap_model(generator).load_state_dict(ckpt["G"])
+
+    accelerator.print("💾 | Loaded model!")
+else:
+    print(
+        "Generator model (--model) isn't specified. Do you have the trained model? If not, you need to train it OR acquire it from somewhere (DON'T ASK ME, YET!)"
+    )
+
+
+def start():
+    # To Mono!
+    decoder = torchcodec.decoders.AudioDecoder(input_audio)
+
+    decoded_samples = decoder.get_all_samples()
+    audio = decoded_samples.data
+    original_sample_rate = decoded_samples.sample_rate
+
+    # Support for multichannel audio
+    # audio = AudioUtils.stereo_tensor_to_mono(audio)
+    audio = AudioUtils.normalize(audio)
+
+    resample_transform = torchaudio.transforms.Resample(
+        original_sample_rate, args.sample_rate
+    )
+
+    audio = resample_transform(audio)
+
+    splitted_audio = AudioUtils.split_audio(audio, clip_length)
+    splitted_audio_on_device = [t.view(1, t.shape[0], t.shape[-1]).to(accelerator.device) for t in splitted_audio]
+    processed_audio = []
+    with torch.no_grad():
+        for clip in tqdm.tqdm(splitted_audio_on_device, desc="Processing..."):
+            channels = []
+            for audio_channel in torch.split(clip, 1, dim=1):
+                output_piece = generator(audio_channel)
+                channels.append(output_piece.detach().cpu())
+            output_clip = torch.cat(channels, dim=1)
+            processed_audio.append(output_clip)
+
+    reconstructed_audio = AudioUtils.reconstruct_audio(processed_audio)
+    reconstructed_audio = reconstructed_audio.squeeze(0)
+    print(f"🔊 | Saving {output_audio}!")
+    torchaudio.save_with_torchcodec(
+        uri=output_audio,
+        src=reconstructed_audio,
+        sample_rate=args.sample_rate,
+        channels_first=True,
+        compression=args.bitrate,
+    )
+
+
+start()

data.py

@@ -1,50 +1,72 @@
-from torch.utils.data import Dataset
-import torch.nn.functional as F
-import torchaudio
 import os
 import random
 
+import torch
+import torchaudio
+import torchcodec.decoders as decoders
+import tqdm
+from torch.utils.data import Dataset
+
+import AudioUtils
+
 
 class AudioDataset(Dataset):
-    audio_sample_rates = [8000, 11025, 16000, 22050]
+    audio_sample_rates = [8000, 11025, 12000, 16000, 22050, 24000, 32000, 44100]
 
-    def __init__(self, input_dir, target_duration=None, padding_mode='constant', padding_value=0.0):
-        self.input_files = [os.path.join(input_dir, f) for f in os.listdir(input_dir) if f.endswith('.wav')]
-        self.target_duration = target_duration  # Duration in seconds or None if not set
-        self.padding_mode = padding_mode
-        self.padding_value = padding_value
+    def __init__(self, input_dir, clip_length: int = 512, normalize: bool = True):
+        self.clip_length = clip_length
+        self.normalize = normalize
+
+        input_files = [
+            os.path.join(input_dir, f)
+            for f in os.listdir(input_dir)
+            if os.path.isfile(os.path.join(input_dir, f))
+            and f.lower().endswith((".wav", ".mp3", ".flac"))
+        ]
+
+        data = []
+        for audio_clip in tqdm.tqdm(
+            input_files, desc=f"Processing {len(input_files)} audio file(s)"
+        ):
+            decoder = decoders.AudioDecoder(audio_clip)
+            decoded_samples = decoder.get_all_samples()
+
+            audio = decoded_samples.data.float()
+            original_sample_rate = decoded_samples.sample_rate
+
+            if normalize:
+                audio = AudioUtils.normalize(audio)
+
+            splitted_high_quality_audio = AudioUtils.split_audio(audio, clip_length, True)
+
+            if not splitted_high_quality_audio:
+                continue
+
+            for splitted_audio_clip in splitted_high_quality_audio:
+                for audio_clip in torch.split(splitted_audio_clip, 1):
+                    data.append((audio_clip, original_sample_rate))
+
+        self.audio_data = data
 
     def __len__(self):
-        return len(self.input_files)
-
+        return len(self.audio_data)
 
     def __getitem__(self, idx):
-        high_quality_wav, sr_original = torchaudio.load(self.input_files[idx], normalize=True)
+        audio_clip = self.audio_data[idx]
+        mangled_sample_rate = random.choice(self.audio_sample_rates)
 
-        sample_rate = random.choice(self.audio_sample_rates)
-        resample_transform = torchaudio.transforms.Resample(sr_original, sample_rate)
-        low_quality_wav = resample_transform(high_quality_wav)
-        low_quality_wav = low_quality_wav
+        resample_transform_low = torchaudio.transforms.Resample(
+            audio_clip[1], mangled_sample_rate
+        )
 
-        # Calculate target length based on desired duration and 16000 Hz
-        if self.target_duration is not None:
-            target_length = int(self.target_duration * 44100)
-        else:
-            # Calculate duration of original high quality audio
-            target_length = high_quality_wav.size(1)
+        resample_transform_high = torchaudio.transforms.Resample(
+            mangled_sample_rate, audio_clip[1]
+        )
 
-        # Pad both to the calculated target length
-        high_quality_wav = self.stretch_tensor(high_quality_wav, target_length)
-        low_quality_wav = self.stretch_tensor(low_quality_wav, target_length)
-
-
-        return low_quality_wav, high_quality_wav
-
-    def stretch_tensor(self, tensor, target_length):
-        current_length = tensor.size(1)
-        scale_factor = target_length / current_length
-
-        # Resample the tensor using linear interpolation
-        tensor = F.interpolate(tensor.unsqueeze(0), scale_factor=scale_factor, mode='linear', align_corners=False).squeeze(0)
-
-        return tensor
+        low_audio_clip = resample_transform_high(resample_transform_low(audio_clip[0]))
+        if audio_clip[0].shape[1] < low_audio_clip.shape[1]:
+            low_audio_clip = low_audio_clip[:, :audio_clip[0].shape[1]]
+        elif audio_clip[0].shape[1] > low_audio_clip.shape[1]:
+            target_len = audio_clip[0].shape[1]
+            low_audio_clip = AudioUtils.pad_tensor(low_audio_clip, target_len)
+        return ((audio_clip[0], low_audio_clip), (audio_clip[1], mangled_sample_rate))

discriminator.py

@@ -1,24 +1,179 @@
+import torch
 import torch.nn as nn
+import torch.nn.functional as F
+from torch.nn.utils.parametrizations import weight_norm, spectral_norm
+
+# -------------------------------------------------------------------
+# 1. Multi-Period Discriminator (MPD)
+#    Captures periodic structures (pitch/timbre) by folding audio.
+# -------------------------------------------------------------------
+
+class DiscriminatorP(nn.Module):
+    def __init__(self, period, kernel_size=5, stride=3, use_spectral_norm=False):
+        super(DiscriminatorP, self).__init__()
+        self.period = period
+        self.use_spectral_norm = use_spectral_norm
+
+        # Use spectral_norm for stability, or weight_norm for performance
+        norm_f = spectral_norm if use_spectral_norm else weight_norm
+
+        # We use 2D convs because we "fold" the 1D audio into 2D (Period x Time)
+        self.convs = nn.ModuleList([
+            norm_f(nn.Conv2d(1, 32, (kernel_size, 1), (stride, 1), padding=(2, 0))),
+            norm_f(nn.Conv2d(32, 128, (kernel_size, 1), (stride, 1), padding=(2, 0))),
+            norm_f(nn.Conv2d(128, 512, (kernel_size, 1), (stride, 1), padding=(2, 0))),
+            norm_f(nn.Conv2d(512, 1024, (kernel_size, 1), (stride, 1), padding=(2, 0))),
+            norm_f(nn.Conv2d(1024, 1024, (kernel_size, 1), 1, padding=(2, 0))),
+        ])
+
+        self.conv_post = norm_f(nn.Conv2d(1024, 1, (3, 1), 1, padding=(1, 0)))
+
+    def forward(self, x):
+        fmap = []
+
+        # 1d to 2d conversion: [B, C, T] -> [B, C, T/P, P]
+        b, c, t = x.shape
+        if t % self.period != 0: # Pad if not divisible by period
+            n_pad = self.period - (t % self.period)
+            x = F.pad(x, (0, n_pad), "reflect")
+            t = t + n_pad
+
+        x = x.view(b, c, t // self.period, self.period)
+
+        for l in self.convs:
+            x = l(x)
+            x = F.leaky_relu(x, 0.1)
+            fmap.append(x) # Store feature map for Feature Matching Loss
+
+        x = self.conv_post(x)
+        fmap.append(x)
+
+        # Flatten back to 1D for score
+        x = torch.flatten(x, 1, -1)
+
+        return x, fmap
+
+
+class MultiPeriodDiscriminator(nn.Module):
+    def __init__(self, periods=[2, 3, 5, 7, 11]):
+        super(MultiPeriodDiscriminator, self).__init__()
+        self.discriminators = nn.ModuleList([
+            DiscriminatorP(p) for p in periods
+        ])
+
+    def forward(self, y, y_hat):
+        y_d_rs = [] # Real scores
+        y_d_gs = [] # Generated (Fake) scores
+        fmap_rs = [] # Real feature maps
+        fmap_gs = [] # Generated (Fake) feature maps
+
+        for i, d in enumerate(self.discriminators):
+            y_d_r, fmap_r = d(y)
+            y_d_g, fmap_g = d(y_hat)
+
+            y_d_rs.append(y_d_r)
+            fmap_rs.append(fmap_r)
+            y_d_gs.append(y_d_g)
+            fmap_gs.append(fmap_g)
+
+        return y_d_rs, y_d_gs, fmap_rs, fmap_gs
+
+
+# -------------------------------------------------------------------
+# 2. Multi-Scale Discriminator (MSD)
+#    Captures structure at different audio resolutions (raw, x0.5, x0.25).
+# -------------------------------------------------------------------
+
+class DiscriminatorS(nn.Module):
+    def __init__(self, use_spectral_norm=False):
+        super(DiscriminatorS, self).__init__()
+        norm_f = spectral_norm if use_spectral_norm else weight_norm
+
+        # Standard 1D Convolutions with large receptive field
+        self.convs = nn.ModuleList([
+            norm_f(nn.Conv1d(1, 16, 15, 1, padding=7)),
+            norm_f(nn.Conv1d(16, 64, 41, 4, groups=4, padding=20)),
+            norm_f(nn.Conv1d(64, 256, 41, 4, groups=16, padding=20)),
+            norm_f(nn.Conv1d(256, 1024, 41, 4, groups=64, padding=20)),
+            norm_f(nn.Conv1d(1024, 1024, 41, 4, groups=256, padding=20)),
+            norm_f(nn.Conv1d(1024, 1024, 5, 1, padding=2)),
+        ])
+        self.conv_post = norm_f(nn.Conv1d(1024, 1, 3, 1, padding=1))
+
+    def forward(self, x):
+        fmap = []
+        for l in self.convs:
+            x = l(x)
+            x = F.leaky_relu(x, 0.1)
+            fmap.append(x)
+        x = self.conv_post(x)
+        fmap.append(x)
+        x = torch.flatten(x, 1, -1)
+        return x, fmap
+
+
+class MultiScaleDiscriminator(nn.Module):
+    def __init__(self):
+        super(MultiScaleDiscriminator, self).__init__()
+        # 3 Scales: Original, Downsampled x2, Downsampled x4
+        self.discriminators = nn.ModuleList([
+            DiscriminatorS(use_spectral_norm=True),
+            DiscriminatorS(),
+            DiscriminatorS(),
+        ])
+        self.meanpools = nn.ModuleList([
+            nn.AvgPool1d(4, 2, padding=2),
+            nn.AvgPool1d(4, 2, padding=2)
+        ])
+
+    def forward(self, y, y_hat):
+        y_d_rs = []
+        y_d_gs = []
+        fmap_rs = []
+        fmap_gs = []
+
+        for i, d in enumerate(self.discriminators):
+            if i != 0:
+                # Downsample input for subsequent discriminators
+                y = self.meanpools[i-1](y)
+                y_hat = self.meanpools[i-1](y_hat)
+
+            y_d_r, fmap_r = d(y)
+            y_d_g, fmap_g = d(y_hat)
+
+            y_d_rs.append(y_d_r)
+            fmap_rs.append(fmap_r)
+            y_d_gs.append(y_d_g)
+            fmap_gs.append(fmap_g)
+
+        return y_d_rs, y_d_gs, fmap_rs, fmap_gs
+
+
+# -------------------------------------------------------------------
+# 3. Master Wrapper
+#    Combines MPD and MSD into one class to fit your training script.
+# -------------------------------------------------------------------
 
 class SISUDiscriminator(nn.Module):
     def __init__(self):
         super(SISUDiscriminator, self).__init__()
-        self.model = nn.Sequential(
-            nn.Conv1d(2, 128, kernel_size=3, padding=1),
-            nn.LeakyReLU(0.2, inplace=True),
-            nn.Conv1d(128, 256, kernel_size=3, padding=1),
-            nn.LeakyReLU(0.2, inplace=True),
-            nn.Conv1d(256, 128, kernel_size=3, padding=1),
-            nn.LeakyReLU(0.2, inplace=True),
-            nn.Conv1d(128, 64, kernel_size=3, padding=1),
-            nn.LeakyReLU(0.2, inplace=True),
-            nn.Conv1d(64, 1, kernel_size=3, padding=1),
-            nn.LeakyReLU(0.2, inplace=True),
-        )
-        self.global_avg_pool = nn.AdaptiveAvgPool1d(1)  # Output size (1,)
+        self.mpd = MultiPeriodDiscriminator()
+        self.msd = MultiScaleDiscriminator()
 
-    def forward(self, x):
-        x = self.model(x)
-        x = self.global_avg_pool(x)
-        x = x.view(-1, 1)  # Flatten to (batch_size, 1)
-        return x
+    def forward(self, y, y_hat):
+        # Return format:
+        # scores_real, scores_fake, features_real, features_fake
+
+        # Run Multi-Period
+        mpd_y_d_rs, mpd_y_d_gs, mpd_fmap_rs, mpd_fmap_gs = self.mpd(y, y_hat)
+
+        # Run Multi-Scale
+        msd_y_d_rs, msd_y_d_gs, msd_fmap_rs, msd_fmap_gs = self.msd(y, y_hat)
+
+        # Combine all results
+        return (
+            mpd_y_d_rs + msd_y_d_rs, # All real scores
+            mpd_y_d_gs + msd_y_d_gs, # All fake scores
+            mpd_fmap_rs + msd_fmap_rs, # All real feature maps
+            mpd_fmap_gs + msd_fmap_gs  # All fake feature maps
+        )

generator.py

@@ -1,23 +1,126 @@
+import torch
 import torch.nn as nn
+from torch.nn.utils.parametrizations import weight_norm
 
-class SISUGenerator(nn.Module):
-    def __init__(self, upscale_scale=1): # No noise_dim parameter
-        super(SISUGenerator, self).__init__()
-        self.model = nn.Sequential(
-            nn.Conv1d(2, 128, kernel_size=3, padding=1),
-            nn.LeakyReLU(0.2, inplace=True),
-            nn.Conv1d(128, 256, kernel_size=3, padding=1),
-            nn.LeakyReLU(0.2, inplace=True),
+def GeneratorBlock(in_channels, out_channels, kernel_size=3, stride=1, dilation=1):
+    padding = (kernel_size - 1) // 2 * dilation
 
-            nn.Upsample(scale_factor=upscale_scale, mode='nearest'),
+    return nn.Sequential(
 
-            nn.Conv1d(256, 128, kernel_size=3, padding=1),
-            nn.LeakyReLU(0.2, inplace=True),
-            nn.Conv1d(128, 64, kernel_size=3, padding=1),
-            nn.LeakyReLU(0.2, inplace=True),
-            nn.Conv1d(64, 2, kernel_size=3, padding=1),
-            nn.Tanh()
+        weight_norm(nn.Conv1d(
+            in_channels,
+            out_channels,
+            kernel_size=kernel_size,
+            stride=stride,
+            dilation=dilation,
+            padding=padding
+        )),
+        nn.PReLU(num_parameters=1, init=0.1),
+    )
+
+
+class AttentionBlock(nn.Module):
+    def __init__(self, channels):
+        super(AttentionBlock, self).__init__()
+        self.attention = nn.Sequential(
+            weight_norm(nn.Conv1d(channels, channels // 4, kernel_size=1)),
+            nn.ReLU(inplace=True),
+            weight_norm(nn.Conv1d(channels // 4, channels, kernel_size=1)),
+            nn.Sigmoid(),
         )
 
     def forward(self, x):
-        return self.model(x)
+        attention_weights = self.attention(x)
+        return x + (x * attention_weights)
+
+
+class ResidualInResidualBlock(nn.Module):
+    def __init__(self, channels, num_convs=3):
+        super(ResidualInResidualBlock, self).__init__()
+
+        self.conv_layers = nn.Sequential(
+            *[GeneratorBlock(channels, channels) for _ in range(num_convs)]
+        )
+
+        self.attention = AttentionBlock(channels)
+
+    def forward(self, x):
+        residual = x
+        x = self.conv_layers(x)
+        x = self.attention(x)
+        return x + residual
+
+def UpsampleBlock(in_channels, out_channels, scale_factor=2):
+    return nn.Sequential(
+        nn.Upsample(scale_factor=scale_factor, mode='nearest'),
+        weight_norm(nn.Conv1d(
+            in_channels=in_channels,
+            out_channels=out_channels,
+            kernel_size=3,
+            stride=1,
+            padding=1
+        )),
+        nn.PReLU(num_parameters=1, init=0.1)
+    )
+
+class SISUGenerator(nn.Module):
+    def __init__(self, channels=32, num_rirb=4):
+        super(SISUGenerator, self).__init__()
+
+        self.first_conv = GeneratorBlock(1, channels)
+
+        self.downsample = GeneratorBlock(channels, channels * 2, stride=2)
+        self.downsample_attn = AttentionBlock(channels * 2)
+        self.downsample_2 = GeneratorBlock(channels * 2, channels * 4, stride=2)
+        self.downsample_2_attn = AttentionBlock(channels * 4)
+
+        self.rirb = nn.Sequential(
+            *[ResidualInResidualBlock(channels * 4) for _ in range(num_rirb)]
+        )
+
+        self.upsample = UpsampleBlock(channels * 4, channels * 2)
+        self.upsample_attn = AttentionBlock(channels * 2)
+        self.compress_1 = GeneratorBlock(channels * 4, channels * 2)
+
+        self.upsample_2 = UpsampleBlock(channels * 2, channels)
+        self.upsample_2_attn = AttentionBlock(channels)
+        self.compress_2 = GeneratorBlock(channels * 2, channels)
+
+        self.final_conv = nn.Sequential(
+            weight_norm(nn.Conv1d(channels, 1, kernel_size=7, padding=3)),
+            nn.Tanh()
+        )
+
+
+    def forward(self, x):
+        residual_input = x
+
+        # Encoding
+        x1 = self.first_conv(x)
+
+        x2 = self.downsample(x1)
+        x2 = self.downsample_attn(x2)
+
+        x3 = self.downsample_2(x2)
+        x3 = self.downsample_2_attn(x3)
+
+        # Bottleneck (Deep Residual processing)
+        x_rirb = self.rirb(x3)
+
+        # Decoding with Skip Connections
+        up1 = self.upsample(x_rirb)
+        up1 = self.upsample_attn(up1)
+
+        cat1 = torch.cat((up1, x2), dim=1)
+        comp1 = self.compress_1(cat1)
+
+        up2 = self.upsample_2(comp1)
+        up2 = self.upsample_2_attn(up2)
+
+        cat2 = torch.cat((up2, x1), dim=1)
+        comp2 = self.compress_2(cat2)
+
+        learned_residual = self.final_conv(comp2)
+
+        output = residual_input + learned_residual
+        return output

requirements.txt

@@ -1,12 +0,0 @@
-filelock>=3.16.1
-fsspec>=2024.10.0
-Jinja2>=3.1.4
-MarkupSafe>=2.1.5
-mpmath>=1.3.0
-networkx>=3.4.2
-numpy>=2.1.2
-pillow>=11.0.0
-setuptools>=70.2.0
-sympy>=1.13.1
-tqdm>=4.67.1
-typing_extensions>=4.12.2

test.py

@@ -1,10 +0,0 @@
-import torch.nn as nn
-import torch
-from discriminator import SISUDiscriminator
-
-
-discriminator = SISUDiscriminator()
-test_input = torch.randn(1, 2, 1000) # Example input (batch_size, channels, frames)
-output = discriminator(test_input)
-print(output)
-print("Output shape:", output.shape)

training.py

@@ -1,135 +1,251 @@
+import argparse
+import datetime
+import os
+
 import torch
 import torch.nn as nn
 import torch.optim as optim
-
-import torch.nn.functional as F
-import torchaudio
 import tqdm
-
-from torch.utils.data import random_split
-from torch.utils.data import DataLoader
+from accelerate import Accelerator
+from torch.utils.data import DataLoader, DistributedSampler
 
 from data import AudioDataset
-from generator import SISUGenerator
 from discriminator import SISUDiscriminator
+from generator import SISUGenerator
+from utils.TrainingTools import discriminator_train, generator_train
 
-# Check for CUDA availability
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-print(f"Using device: {device}")
+# ---------------------------
+# Argument parsing
+# ---------------------------
+parser = argparse.ArgumentParser(description="Training script (safer defaults)")
+parser.add_argument("--resume", action="store_true", help="Resume training")
+parser.add_argument(
+    "--epochs", type=int, default=5000, help="Number of training epochs"
+)
+parser.add_argument("--batch_size", type=int, default=8, help="Batch size")
+parser.add_argument("--num_workers", type=int, default=4, help="DataLoader num_workers") # Increased workers slightly
+parser.add_argument("--debug", action="store_true", help="Print debug logs")
+parser.add_argument(
+    "--no_pin_memory", action="store_true", help="Disable pin_memory even on CUDA"
+)
+args = parser.parse_args()
 
-# Initialize dataset and dataloader
-dataset_dir = './dataset/good'
-dataset = AudioDataset(dataset_dir, target_duration=2.0)
+# ---------------------------
+# Init accelerator
+# ---------------------------
 
-dataset_size = len(dataset)
-train_size = int(dataset_size * .9)
-val_size = int(dataset_size-train_size)
+try:
+    accelerator = Accelerator(mixed_precision="bf16")
+except Exception:
+    accelerator = Accelerator(mixed_precision="fp16")
+    accelerator.print("⚠️ | bf16 unavailable — falling back to fp16")
 
-train_dataset, val_dataset = random_split(dataset, [train_size, val_size])
-
-train_data_loader = DataLoader(train_dataset, batch_size=8, shuffle=True)
-val_data_loader = DataLoader(val_dataset, batch_size=8, shuffle=True)
-
-# Initialize models and move them to device
+# ---------------------------
+# Models
+# ---------------------------
 generator = SISUGenerator()
+# Note: SISUDiscriminator is now an Ensemble (MPD + MSD)
 discriminator = SISUDiscriminator()
 
-generator = generator.to(device)
-discriminator = discriminator.to(device)
+accelerator.print("🔨 | Compiling models...")
 
-# Loss
-criterion_g = nn.L1Loss()
-criterion_d = nn.BCEWithLogitsLoss()
+# Torch compile is great, but if you hit errors with the new List/Tuple outputs
+# of the discriminator, you might need to disable it for D.
+generator = torch.compile(generator)
+discriminator = torch.compile(discriminator)
 
-# Optimizers
-optimizer_g = optim.Adam(generator.parameters(), lr=0.0001, betas=(0.5, 0.999))
-optimizer_d = optim.Adam(discriminator.parameters(), lr=0.0001, betas=(0.5, 0.999))
+accelerator.print("✅ | Compiling done!")
 
-# Scheduler
-scheduler_g = torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer_g, mode='min', factor=0.5, patience=5)
-scheduler_d = torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer_d, mode='min', factor=0.5, patience=5)
+# ---------------------------
+# Dataset / DataLoader
+# ---------------------------
+accelerator.print("📊 | Fetching dataset...")
+dataset = AudioDataset("./dataset", 8192)
 
-def snr(y_true, y_pred):
-    noise = y_true - y_pred
-    signal_power = torch.mean(y_true ** 2)
-    noise_power = torch.mean(noise ** 2)
-    snr_db = 10 * torch.log10(signal_power / noise_power)
-    return snr_db
+sampler = DistributedSampler(dataset) if accelerator.num_processes > 1 else None
+pin_memory = torch.cuda.is_available() and not args.no_pin_memory
 
-def discriminator_train(discriminator, optimizer, criterion, generator, real_labels, fake_labels, high_quality, low_quality):
-    optimizer.zero_grad()
+train_loader = DataLoader(
+    dataset,
+    sampler=sampler,
+    batch_size=args.batch_size,
+    shuffle=(sampler is None),
+    num_workers=args.num_workers,
+    pin_memory=pin_memory,
+    persistent_workers=pin_memory,
+)
 
-    discriminator_decision_from_real = discriminator(high_quality)
-    d_loss_real = criterion(discriminator_decision_from_real, real_labels)
+if not train_loader or not train_loader.batch_size or train_loader.batch_size == 0:
+    accelerator.print("🪹 | There is no data to train with! Exiting...")
+    exit()
 
-    generator_output = generator(low_quality)
-    discriminator_decision_from_fake = discriminator(generator_output.detach())
-    d_loss_fake = criterion(discriminator_decision_from_fake, fake_labels)
+loader_batch_size = train_loader.batch_size
 
-    d_loss = (d_loss_real + d_loss_fake) / 2.0
+accelerator.print("✅ | Dataset fetched!")
 
-    d_loss.backward()
-    nn.utils.clip_grad_norm_(discriminator.parameters(), max_norm=1.0) #Gradient Clipping
-    optimizer.step()
-    # print(f"Discriminator Loss: {d_loss.item():.4f}, Mean Real Logit: {discriminator_decision_from_real.mean().item():.2f}, Mean Fake Logit: {discriminator_decision_from_fake.mean().item():.2f}")
+# ---------------------------
+# Losses / Optimizers / Scalers
+# ---------------------------
 
-def start_training():
+optimizer_g = optim.AdamW(
+    generator.parameters(), lr=0.0003, betas=(0.5, 0.999), weight_decay=0.0001
+)
+optimizer_d = optim.AdamW(
+    discriminator.parameters(), lr=0.0003, betas=(0.5, 0.999), weight_decay=0.0001
+)
 
-    # Training loop
-    # discriminator_epochs = 1000
-    generator_epochs = 500
-    for generator_epoch in range(generator_epochs):
-        low_quality_audio = torch.empty((1))
-        high_quality_audio = torch.empty((1))
-        ai_enhanced_audio = torch.empty((1))
+scheduler_g = torch.optim.lr_scheduler.ReduceLROnPlateau(
+    optimizer_g, mode="min", factor=0.5, patience=5
+)
+scheduler_d = torch.optim.lr_scheduler.ReduceLROnPlateau(
+    optimizer_d, mode="min", factor=0.5, patience=5
+)
 
-        # Training
-        for low_quality, high_quality in tqdm.tqdm(train_data_loader, desc=f"Epoch {generator_epoch+1}/{generator_epochs}"):
-            high_quality = high_quality.to(device)
-            low_quality = low_quality.to(device)
+# ---------------------------
+# Prepare accelerator
+# ---------------------------
 
-            batch_size = high_quality.size(0)
-            real_labels = torch.ones(batch_size, 1).to(device)
-            fake_labels = torch.zeros(batch_size, 1).to(device)
+generator, discriminator, optimizer_g, optimizer_d, train_loader = accelerator.prepare(
+    generator, discriminator, optimizer_g, optimizer_d, train_loader
+)
 
-            # Train Discriminator
-            discriminator.train()
+# ---------------------------
+# Checkpoint helpers
+# ---------------------------
+models_dir = "./models"
+os.makedirs(models_dir, exist_ok=True)
 
-            for _ in range(3):
-                discriminator_train(discriminator, optimizer_d, criterion_d, generator, real_labels, fake_labels, high_quality, low_quality)
 
-            # Train Generator
-            generator.train()
-            optimizer_g.zero_grad()
+def save_ckpt(path, epoch, loss=None, is_best=False):
+    accelerator.wait_for_everyone()
+    if accelerator.is_main_process:
+        state = {
+            "epoch": epoch,
+            "G": accelerator.unwrap_model(generator).state_dict(),
+            "D": accelerator.unwrap_model(discriminator).state_dict(),
+            "optG": optimizer_g.state_dict(),
+            "optD": optimizer_d.state_dict(),
+            "schedG": scheduler_g.state_dict(),
+            "schedD": scheduler_d.state_dict()
+        }
 
-            # Generator loss: how well fake data fools the discriminator
-            generator_output = generator(low_quality)
-            discriminator_decision = discriminator(generator_output)  # No detach here
-            g_loss = criterion_g(discriminator_decision, real_labels)  # Train generator to produce real-like outputs
+        accelerator.save(state, os.path.join(models_dir, "last.pt"))
 
-            g_loss.backward()
+        if is_best:
+            accelerator.save(state, os.path.join(models_dir, "best.pt"))
+            accelerator.print(f"🌟 | New best model saved with G Loss: {loss:.4f}")
+
+
+start_epoch = 0
+if args.resume:
+    ckpt_path = os.path.join(models_dir, "last.pt")
+    if os.path.exists(ckpt_path):
+        ckpt = torch.load(ckpt_path)
+
+        accelerator.unwrap_model(generator).load_state_dict(ckpt["G"])
+        accelerator.unwrap_model(discriminator).load_state_dict(ckpt["D"])
+        optimizer_g.load_state_dict(ckpt["optG"])
+        optimizer_d.load_state_dict(ckpt["optD"])
+        scheduler_g.load_state_dict(ckpt["schedG"])
+        scheduler_d.load_state_dict(ckpt["schedD"])
+
+        start_epoch = ckpt.get("epoch", 1)
+        accelerator.print(f"🔁 | Resumed from epoch {start_epoch}!")
+    else:
+        accelerator.print("⚠️ | Resume requested but no checkpoint found. Starting fresh.")
+
+
+accelerator.print("🏋️ | Started training...")
+
+try:
+    for epoch in range(start_epoch, args.epochs):
+        generator.train()
+        discriminator.train()
+
+        discriminator_time = 0
+        generator_time = 0
+
+        running_d, running_g, steps = 0.0, 0.0, 0
+
+        progress_bar = tqdm.tqdm(train_loader, desc=f"Epoch {epoch} | D {discriminator_time}μs | G {generator_time}μs")
+
+        for i, (
+            (high_quality, low_quality),
+            (high_sample_rate, low_sample_rate),
+        ) in enumerate(progress_bar):
+            with accelerator.autocast():
+                generator_output = generator(low_quality)
+
+            # --- Discriminator ---
+            d_time = datetime.datetime.now()
+            optimizer_d.zero_grad(set_to_none=True)
+            with accelerator.autocast():
+                d_loss = discriminator_train(
+                    high_quality,
+                    discriminator,
+                    generator_output.detach()
+                )
+
+            accelerator.backward(d_loss)
+            optimizer_d.step()
+            discriminator_time = (datetime.datetime.now() - d_time).microseconds
+
+            # --- Generator ---
+            g_time = datetime.datetime.now()
+            optimizer_g.zero_grad(set_to_none=True)
+            with accelerator.autocast():
+                g_total, g_adv = generator_train(
+                    low_quality,
+                    high_quality,
+                    generator,
+                    discriminator,
+                    generator_output
+                )
+
+            accelerator.backward(g_total)
+            torch.nn.utils.clip_grad_norm_(generator.parameters(), 1)
             optimizer_g.step()
+            generator_time = (datetime.datetime.now() - g_time).microseconds
 
-            low_quality_audio = low_quality
-            high_quality_audio = high_quality
-            ai_enhanced_audio = generator_output
+            d_val = accelerator.gather(d_loss.detach()).mean()
+            g_val = accelerator.gather(g_total.detach()).mean()
 
-        metric = snr(high_quality_audio, ai_enhanced_audio)
-        print(f"Generator metric {metric}!")
-        scheduler_g.step(metric)
+            if torch.isfinite(d_val):
+                running_d += d_val.item()
+            else:
+                accelerator.print(
+                    f"🫥 | NaN in discriminator loss at step {i}, skipping update."
+                )
 
-        if generator_epoch % 10 == 0:
-            print(f"Saved epoch {generator_epoch}!")
-            torchaudio.save(f"./output/epoch-{generator_epoch}-audio-crap.wav", low_quality_audio[0].cpu(), 44100)
-            torchaudio.save(f"./output/epoch-{generator_epoch}-audio-ai.wav", ai_enhanced_audio[0].cpu(), 44100)
-            torchaudio.save(f"./output/epoch-{generator_epoch}-audio-orig.wav", high_quality_audio[0].cpu(), 44100)
+            if torch.isfinite(g_val):
+                running_g += g_val.item()
+            else:
+                accelerator.print(
+                    f"🫥 | NaN in generator loss at step {i}, skipping update."
+                )
 
-        if generator_epoch % 50 == 0:
-            torch.save(discriminator.state_dict(), "discriminator.pt")
-            torch.save(generator.state_dict(), "generator.pt")
+            steps += 1
+            progress_bar.set_description(f"Epoch {epoch} | D {discriminator_time}μs | G {generator_time}μs")
 
-    torch.save(discriminator.state_dict(), "discriminator.pt")
-    torch.save(generator.state_dict(), "generator.pt")
-    print("Training complete!")
+        if steps == 0:
+            accelerator.print("🪹 | No steps in epoch (empty dataloader?). Exiting.")
+            break
 
-start_training()
+        mean_d = running_d / steps
+        mean_g = running_g / steps
+
+        scheduler_d.step(mean_d)
+        scheduler_g.step(mean_g)
+
+        save_ckpt(os.path.join(models_dir, "last.pt"), epoch)
+        accelerator.print(f"🤝 | Epoch {epoch} done | D {mean_d:.4f} | G {mean_g:.4f}")
+
+except Exception:
+    try:
+        save_ckpt(os.path.join(models_dir, "crash_last.pt"), epoch)
+        accelerator.print(f"💾 | Saved crash checkpoint for epoch {epoch}")
+    except Exception as e:
+        accelerator.print("😬 | Failed saving crash checkpoint:", e)
+    raise
+
+accelerator.print("🏁 | Training finished.")

utils/MultiResolutionSTFTLoss.py

@@ -0,0 +1,68 @@
+from typing import Dict, List
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torchaudio.transforms as T
+
+
+class MultiResolutionSTFTLoss(nn.Module):
+    def __init__(
+        self,
+        fft_sizes: List[int] = [512, 1024, 2048, 4096, 8192],
+        hop_sizes: List[int] = [64, 128, 256, 512, 1024],
+        win_lengths: List[int] = [256, 512, 1024, 2048, 4096],
+        eps: float = 1e-7,
+        center: bool = True
+    ):
+        super().__init__()
+
+        self.eps = eps
+        self.n_resolutions = len(fft_sizes)
+
+        self.stft_transforms = nn.ModuleList()
+        for i, (n_fft, hop_len, win_len) in enumerate(zip(fft_sizes, hop_sizes, win_lengths)):
+            stft = T.Spectrogram(
+                n_fft=n_fft,
+                hop_length=hop_len,
+                win_length=win_len,
+                window_fn=torch.hann_window,
+                power=None,
+                center=center,
+                pad_mode="reflect",
+                normalized=False,
+            )
+            self.stft_transforms.append(stft)
+
+    def forward(
+        self, y_true: torch.Tensor, y_pred: torch.Tensor
+    ) -> Dict[str, torch.Tensor]:
+        if y_true.dim() == 3 and y_true.size(1) == 1:
+            y_true = y_true.squeeze(1)
+        if y_pred.dim() == 3 and y_pred.size(1) == 1:
+            y_pred = y_pred.squeeze(1)
+
+        sc_loss = 0.0
+        mag_loss = 0.0
+
+        for stft in self.stft_transforms:
+            stft.window = stft.window.to(y_true.device)
+            stft_true = stft(y_true)
+            stft_pred = stft(y_pred)
+
+            stft_mag_true = torch.abs(stft_true)
+            stft_mag_pred = torch.abs(stft_pred)
+
+            norm_true = torch.linalg.norm(stft_mag_true, dim=(-2, -1))
+            norm_diff = torch.linalg.norm(stft_mag_true - stft_mag_pred, dim=(-2, -1))
+            sc_loss += torch.mean(norm_diff / (norm_true + self.eps))
+
+            log_mag_pred = torch.log(stft_mag_pred + self.eps)
+            log_mag_true = torch.log(stft_mag_true + self.eps)
+            mag_loss += F.l1_loss(log_mag_pred, log_mag_true)
+
+        sc_loss /= self.n_resolutions
+        mag_loss /= self.n_resolutions
+        total_loss = sc_loss + mag_loss
+
+        return {"total": total_loss, "sc": sc_loss, "mag": mag_loss}

utils/TrainingTools.py

@@ -0,0 +1,88 @@
+import torch
+import torch.nn.functional as F
+from utils.MultiResolutionSTFTLoss import MultiResolutionSTFTLoss
+
+# Keep STFT settings as is
+stft_loss_fn = MultiResolutionSTFTLoss(
+    fft_sizes=[512, 1024, 2048],
+    hop_sizes=[64, 128, 256],
+    win_lengths=[256, 512, 1024]
+)
+
+
+def feature_matching_loss(fmap_r, fmap_g):
+    """
+    Computes L1 distance between real and fake feature maps.
+    """
+    loss = 0
+    for dr, dg in zip(fmap_r, fmap_g):
+        for rl, gl in zip(dr, dg):
+            rl = rl.detach()
+            loss += torch.mean(torch.abs(rl - gl))
+
+    return loss * 2
+
+
+def discriminator_loss(disc_real_outputs, disc_generated_outputs):
+    """
+    Least Squares GAN Loss (LSGAN) for the Discriminator.
+    Objective: Real -> 1, Fake -> 0
+    """
+    loss = 0
+    r_losses = []
+    g_losses = []
+
+    for dr, dg in zip(disc_real_outputs, disc_generated_outputs):
+        r_loss = torch.mean((dr - 1) ** 2)
+        g_loss = torch.mean(dg ** 2)
+
+        loss += (r_loss + g_loss)
+        r_losses.append(r_loss.item())
+        g_losses.append(g_loss.item())
+
+    return loss, r_losses, g_losses
+
+
+def generator_adv_loss(disc_generated_outputs):
+    loss = 0.0
+    for dg in disc_generated_outputs:
+        loss += torch.mean((dg - 1) ** 2)
+    return loss
+
+
+def discriminator_train(
+    high_quality,
+    discriminator,
+    generator_output
+):
+    y_d_rs, y_d_gs, _, _ = discriminator(high_quality, generator_output.detach())
+
+    d_loss, _, _ = discriminator_loss(y_d_rs, y_d_gs)
+
+    return d_loss
+
+
+def generator_train(
+    low_quality,
+    high_quality,
+    generator,
+    discriminator,
+    generator_output
+):
+    y_d_rs, y_d_gs, fmap_rs, fmap_gs = discriminator(high_quality, generator_output)
+
+    loss_gen_adv = generator_adv_loss(y_d_gs)
+
+    loss_fm = feature_matching_loss(fmap_rs, fmap_gs)
+
+    stft_loss = stft_loss_fn(high_quality, generator_output)["total"]
+
+    lambda_stft = 45.0
+    lambda_fm = 2.0
+    lambda_adv = 1.0
+
+    combined_loss = (lambda_stft * stft_loss) + \
+                    (lambda_fm * loss_fm) + \
+                    (lambda_adv * loss_gen_adv)
+
+    return combined_loss, loss_gen_adv