初始化换发型项目：3个微服务代码 + 部署脚本

包含： - hair_service_sd: 换发型/换发色算法服务 (端口 8801) - photo_service: LoRA 训练调度服务 (端口 32678) - stable-diffusion-webui: SD WebUI 推理服务 (端口 57860) - kohya_ss_home: 训练环境代码 - meidaojia: 监控测试脚本 - setup.sh: 一键部署脚本 (conda环境恢复 + 配置生成 + 完整性检查) - start_all_services.sh: 启动3个服务 - configure.ini.template: 路径模板化 (BASE_DIR自动推导) - conda_envs/py310.yml: py310 环境定义大文件 (weights/, models/, data/, conda_envs/*.tar.gz 等) 通过 .gitignore 排除，由网盘单独上传。
2026-07-11 18:11:49 +08:00
commit 0eb61f3e60
628 changed files with 120882 additions and 0 deletions
@@ -0,0 +1,144 @@
+"""
+Tiny AutoEncoder for Stable Diffusion
+(DNN for encoding / decoding SD's latent space)
+
+https://github.com/madebyollin/taesd
+"""
+import os
+import torch
+import torch.nn as nn
+
+from modules import devices, paths_internal, shared
+
+sd_vae_taesd_models = {}
+
+
+def conv(n_in, n_out, **kwargs):
+    return nn.Conv2d(n_in, n_out, 3, padding=1, **kwargs)
+
+
+class Clamp(nn.Module):
+    @staticmethod
+    def forward(x):
+        return torch.tanh(x / 3) * 3
+
+
+class Block(nn.Module):
+    def __init__(self, n_in, n_out):
+        super().__init__()
+        self.conv = nn.Sequential(conv(n_in, n_out), nn.ReLU(), conv(n_out, n_out), nn.ReLU(), conv(n_out, n_out))
+        self.skip = nn.Conv2d(n_in, n_out, 1, bias=False) if n_in != n_out else nn.Identity()
+        self.fuse = nn.ReLU()
+
+    def forward(self, x):
+        return self.fuse(self.conv(x) + self.skip(x))
+
+
+def decoder(latent_channels=4):
+    return nn.Sequential(
+        Clamp(), conv(latent_channels, 64), nn.ReLU(),
+        Block(64, 64), Block(64, 64), Block(64, 64), nn.Upsample(scale_factor=2), conv(64, 64, bias=False),
+        Block(64, 64), Block(64, 64), Block(64, 64), nn.Upsample(scale_factor=2), conv(64, 64, bias=False),
+        Block(64, 64), Block(64, 64), Block(64, 64), nn.Upsample(scale_factor=2), conv(64, 64, bias=False),
+        Block(64, 64), conv(64, 3),
+    )
+
+
+def encoder(latent_channels=4):
+    return nn.Sequential(
+        conv(3, 64), Block(64, 64),
+        conv(64, 64, stride=2, bias=False), Block(64, 64), Block(64, 64), Block(64, 64),
+        conv(64, 64, stride=2, bias=False), Block(64, 64), Block(64, 64), Block(64, 64),
+        conv(64, 64, stride=2, bias=False), Block(64, 64), Block(64, 64), Block(64, 64),
+        conv(64, latent_channels),
+    )
+
+
+class TAESDDecoder(nn.Module):
+    latent_magnitude = 3
+    latent_shift = 0.5
+
+    def __init__(self, decoder_path="taesd_decoder.pth", latent_channels=None):
+        """Initialize pretrained TAESD on the given device from the given checkpoints."""
+        super().__init__()
+
+        if latent_channels is None:
+            latent_channels = 16 if "taesd3" in str(decoder_path) else 4
+
+        self.decoder = decoder(latent_channels)
+        self.decoder.load_state_dict(
+            torch.load(decoder_path, map_location='cpu' if devices.device.type != 'cuda' else None))
+
+
+class TAESDEncoder(nn.Module):
+    latent_magnitude = 3
+    latent_shift = 0.5
+
+    def __init__(self, encoder_path="taesd_encoder.pth", latent_channels=None):
+        """Initialize pretrained TAESD on the given device from the given checkpoints."""
+        super().__init__()
+
+        if latent_channels is None:
+            latent_channels = 16 if "taesd3" in str(encoder_path) else 4
+
+        self.encoder = encoder(latent_channels)
+        self.encoder.load_state_dict(
+            torch.load(encoder_path, map_location='cpu' if devices.device.type != 'cuda' else None))
+
+
+def download_model(model_path, model_url):
+    if not os.path.exists(model_path):
+        os.makedirs(os.path.dirname(model_path), exist_ok=True)
+
+        print(f'Downloading TAESD model to: {model_path}')
+        torch.hub.download_url_to_file(model_url, model_path)
+
+
+def decoder_model():
+    if shared.sd_model.is_sd3:
+        model_name = "taesd3_decoder.pth"
+    elif shared.sd_model.is_sdxl:
+        model_name = "taesdxl_decoder.pth"
+    else:
+        model_name = "taesd_decoder.pth"
+
+    loaded_model = sd_vae_taesd_models.get(model_name)
+
+    if loaded_model is None:
+        model_path = os.path.join(paths_internal.models_path, "VAE-taesd", model_name)
+        download_model(model_path, 'https://github.com/madebyollin/taesd/raw/main/' + model_name)
+
+        if os.path.exists(model_path):
+            loaded_model = TAESDDecoder(model_path)
+            loaded_model.eval()
+            loaded_model.to(devices.device, devices.dtype)
+            sd_vae_taesd_models[model_name] = loaded_model
+        else:
+            raise FileNotFoundError('TAESD model not found')
+
+    return loaded_model.decoder
+
+
+def encoder_model():
+    if shared.sd_model.is_sd3:
+        model_name = "taesd3_encoder.pth"
+    elif shared.sd_model.is_sdxl:
+        model_name = "taesdxl_encoder.pth"
+    else:
+        model_name = "taesd_encoder.pth"
+
+    loaded_model = sd_vae_taesd_models.get(model_name)
+
+    if loaded_model is None:
+        model_path = os.path.join(paths_internal.models_path, "VAE-taesd", model_name)
+        download_model(model_path, 'https://github.com/madebyollin/taesd/raw/main/' + model_name)
+
+        if os.path.exists(model_path):
+            loaded_model = TAESDEncoder(model_path)
+            loaded_model.eval()
+            loaded_model.to(devices.device, devices.dtype)
+            sd_vae_taesd_models[model_name] = loaded_model
+        else:
+            raise FileNotFoundError('TAESD model not found')
+
+    return loaded_model.encoder