initial setup: ComfyUI + kohya_ss scripts, LoRA config, workflows

2026-03-13 22:12:04 +01:00
commit 4c2972e7a2
9 changed files with 419 additions and 0 deletions
--- a/.gitignore
+++ b/.gitignore
@@ -0,0 +1,31 @@
+# models - too large for git
+models/
+*.safetensors
+*.ckpt
+*.pt
+*.pth
+*.bin
+
+# generated outputs
+output/
+outputs/
+
+# venvs / installs
+comfyui/venv/
+comfyui/__pycache__/
+kohya_ss/venv/
+kohya_ss/__pycache__/
+kohya_ss/logs/
+
+# training data (images)
+training_data/
+
+# ComfyUI user data / cache
+comfyui/user/
+comfyui/temp/
+comfyui/custom_nodes/**/node_modules/
+
+# misc
+*.log
+.DS_Store
+Thumbs.db
--- a/README.md
+++ b/README.md
@@ -0,0 +1,72 @@
+# animepics
+
+Anime image generation + LoRA training setup.
+
+## Stack
+
+- **ComfyUI** — image generation UI
+- **NoobAI-XL** — base anime model (SDXL-based, SFW+NSFW)
+- **kohya_ss** — LoRA training
+
+## Requirements
+
+- Python 3.10+ installed and in PATH
+- Git installed
+- NVIDIA GPU with CUDA (4070 recommended)
+- ~20GB free disk space for models
+
+## Setup
+
+Run once to install everything:
+
+```powershell
+.\setup.ps1
+```
+
+This will:
+1. Clone ComfyUI and install dependencies
+2. Clone kohya_ss and install dependencies
+3. Install ComfyUI custom nodes (AnimateDiff, ControlNet, etc.)
+4. Download the NoobAI-XL base model + VAE
+
+## Launch
+
+```powershell
+# Start image generation UI (opens browser at localhost:8188)
+.\launch_comfyui.ps1
+
+# Start LoRA training UI (opens browser at localhost:7860)
+.\launch_kohya.ps1
+```
+
+## LoRA Training
+
+1. Put your training images in `training_data/<your_lora_name>/img/10_<trigger_word>/`
+2. Copy `training/example_lora_config.toml` and edit it
+3. Launch kohya and use the GUI, or run `.\train_lora.ps1 <config_file>`
+
+## Directory Structure
+
+```
+animepics/
+├── comfyui/          # ComfyUI install (gitignored venv)
+├── kohya_ss/         # kohya_ss install (gitignored venv)
+├── models/           # shared model storage (gitignored)
+│   ├── checkpoints/  # base models (.safetensors)
+│   ├── loras/        # trained LoRAs
+│   ├── vae/          # VAE models
+│   ├── embeddings/   # textual inversions
+│   └── controlnet/   # ControlNet models
+├── training_data/    # LoRA training images (gitignored)
+├── output/           # generated images (gitignored)
+├── training/         # LoRA training configs
+└── workflows/        # ComfyUI workflow JSON files
+```
+
+## Model Downloads
+
+Base model (NoobAI-XL Vpred):
+- https://civitai.com/models/833294
+
+Good NSFW VAE:
+- Already baked into NoobAI, but sdxl_vae.safetensors from stabilityai works too
--- a/config/extra_model_paths.yaml
+++ b/config/extra_model_paths.yaml
@@ -0,0 +1,11 @@
+# ComfyUI extra model paths — points to shared models/ directory
+# so both ComfyUI and kohya_ss use the same model storage
+
+animepics:
+    base_path: E:/animepics/models
+    checkpoints: checkpoints
+    loras: loras
+    vae: vae
+    embeddings: embeddings
+    controlnet: controlnet
+    upscale_models: upscale_models
--- a/launch_comfyui.ps1
+++ b/launch_comfyui.ps1
@@ -0,0 +1,12 @@
+#!/usr/bin/env pwsh
+# launch_comfyui.ps1 — start ComfyUI
+
+$root = $PSScriptRoot
+Set-Location "$root/comfyui"
+
+.\venv\Scripts\Activate.ps1
+python main.py `
+    --listen 127.0.0.1 `
+    --port 8188 `
+    --extra-model-paths-config "$root/comfyui/extra_model_paths.yaml" `
+    --preview-method auto
--- a/launch_kohya.ps1
+++ b/launch_kohya.ps1
@@ -0,0 +1,8 @@
+#!/usr/bin/env pwsh
+# launch_kohya.ps1 — start kohya_ss GUI
+
+$root = $PSScriptRoot
+Set-Location "$root/kohya_ss"
+
+.\venv\Scripts\Activate.ps1
+python kohya_gui.py --listen 127.0.0.1 --server_port 7860
--- a/setup.ps1
+++ b/setup.ps1
@@ -0,0 +1,88 @@
+#!/usr/bin/env pwsh
+# setup.ps1 — one-time install of ComfyUI + kohya_ss
+
+$root = $PSScriptRoot
+Set-Location $root
+
+Write-Host "==> setting up animepics environment" -ForegroundColor Cyan
+
+# create model dirs
+$model_dirs = @(
+    "models/checkpoints",
+    "models/loras",
+    "models/vae",
+    "models/embeddings",
+    "models/controlnet",
+    "models/upscale_models",
+    "training_data",
+    "output"
+)
+foreach ($dir in $model_dirs) {
+    New-Item -ItemType Directory -Force -Path "$root/$dir" | Out-Null
+}
+Write-Host "  created model directories" -ForegroundColor Green
+
+# --- ComfyUI ---
+Write-Host "`n==> installing ComfyUI" -ForegroundColor Cyan
+if (-not (Test-Path "$root/comfyui")) {
+    git clone https://github.com/comfyanonymous/ComfyUI.git "$root/comfyui"
+} else {
+    Write-Host "  comfyui already cloned, pulling latest" -ForegroundColor Yellow
+    git -C "$root/comfyui" pull
+}
+
+Push-Location "$root/comfyui"
+python -m venv venv
+.\venv\Scripts\Activate.ps1
+pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124
+pip install -r requirements.txt
+deactivate
+Pop-Location
+Write-Host "  ComfyUI installed" -ForegroundColor Green
+
+# ComfyUI custom nodes
+Write-Host "`n==> installing ComfyUI custom nodes" -ForegroundColor Cyan
+$custom_nodes = @(
+    @{ name="ComfyUI-Manager";      url="https://github.com/ltdrdata/ComfyUI-Manager.git" },
+    @{ name="ComfyUI_IPAdapter_plus"; url="https://github.com/cubiq/ComfyUI_IPAdapter_plus.git" },
+    @{ name="ComfyUI-Advanced-ControlNet"; url="https://github.com/Kosinkadink/ComfyUI-Advanced-ControlNet.git" },
+    @{ name="was-node-suite-comfyui"; url="https://github.com/WASasquatch/was-node-suite-comfyui.git" }
+)
+New-Item -ItemType Directory -Force -Path "$root/comfyui/custom_nodes" | Out-Null
+foreach ($node in $custom_nodes) {
+    $node_path = "$root/comfyui/custom_nodes/$($node.name)"
+    if (-not (Test-Path $node_path)) {
+        git clone $node.url $node_path
+        Write-Host "  cloned $($node.name)" -ForegroundColor Green
+    } else {
+        git -C $node_path pull
+        Write-Host "  updated $($node.name)" -ForegroundColor Yellow
+    }
+}
+
+# copy extra_model_paths config so ComfyUI finds our shared models dir
+Copy-Item -Force "$root/config/extra_model_paths.yaml" "$root/comfyui/extra_model_paths.yaml"
+Write-Host "  copied extra_model_paths.yaml" -ForegroundColor Green
+
+# --- kohya_ss ---
+Write-Host "`n==> installing kohya_ss" -ForegroundColor Cyan
+if (-not (Test-Path "$root/kohya_ss")) {
+    git clone https://github.com/bmaltais/kohya_ss.git "$root/kohya_ss"
+} else {
+    Write-Host "  kohya_ss already cloned, pulling latest" -ForegroundColor Yellow
+    git -C "$root/kohya_ss" pull
+}
+
+Push-Location "$root/kohya_ss"
+python -m venv venv
+.\venv\Scripts\Activate.ps1
+pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124
+pip install -r requirements.txt
+deactivate
+Pop-Location
+Write-Host "  kohya_ss installed" -ForegroundColor Green
+
+Write-Host "`n==> setup complete!" -ForegroundColor Cyan
+Write-Host "  next: download your base model into models/checkpoints/" -ForegroundColor White
+Write-Host "  recommended: NoobAI-XL from https://civitai.com/models/833294" -ForegroundColor White
+Write-Host "`n  then run: .\launch_comfyui.ps1" -ForegroundColor White
--- a/train_lora.ps1
+++ b/train_lora.ps1
@@ -0,0 +1,23 @@
+#!/usr/bin/env pwsh
+# train_lora.ps1 — run LoRA training from a config file
+# usage: .\train_lora.ps1 training/my_lora.toml
+
+param(
+    [Parameter(Mandatory=$true)]
+    [string]$config_file
+)
+
+$root = $PSScriptRoot
+
+if (-not (Test-Path "$root/$config_file")) {
+    Write-Host "config file not found: $config_file" -ForegroundColor Red
+    exit 1
+}
+
+Set-Location "$root/kohya_ss"
+.\venv\Scripts\Activate.ps1
+
+accelerate launch `
+    --num_cpu_threads_per_process 2 `
+    train_network.py `
+    --config_file "$root/$config_file"
--- a/training/example_lora_config.toml
+++ b/training/example_lora_config.toml
@@ -0,0 +1,54 @@
+# example LoRA training config for kohya_ss (SDXL)
+# copy this, rename it, and edit the paths/settings for your lora
+# then train with: .\train_lora.ps1 training/my_lora.toml
+
+[model_arguments]
+pretrained_model_name_or_path = "E:/animepics/models/checkpoints/noobai-xl.safetensors"
+# set to true for vpred models (NoobAI-XL uses vpred)
+v_parameterization = true
+zero_terminal_snr = true
+
+[saving_arguments]
+save_every_n_epochs = 1
+save_model_as = "safetensors"
+output_dir = "E:/animepics/models/loras"
+output_name = "my_lora_v1"
+
+[dataset_arguments]
+# dataset dir structure: training_data/<lora_name>/img/<repeats>_<trigger>/
+train_data_dir = "E:/animepics/training_data/my_lora/img"
+resolution = "1024,1024"
+enable_bucket = true
+min_bucket_reso = 512
+max_bucket_reso = 2048
+bucket_reso_steps = 64
+caption_extension = ".txt"
+shuffle_caption = true
+keep_tokens = 1
+
+[training_arguments]
+output_dir = "E:/animepics/models/loras"
+logging_dir = "E:/animepics/kohya_ss/logs"
+max_train_epochs = 10
+train_batch_size = 1
+gradient_accumulation_steps = 1
+gradient_checkpointing = true
+mixed_precision = "bf16"
+save_precision = "bf16"
+seed = 42
+max_token_length = 225
+xformers = true
+# learning rates — good defaults for NoobAI-XL
+learning_rate = 0.0001
+unet_lr = 0.0001
+text_encoder_lr = 0.00005
+lr_scheduler = "cosine_with_restarts"
+lr_warmup_steps = 100
+optimizer_type = "AdamW8bit"
+
+[network_arguments]
+network_module = "networks.lora"
+network_dim = 32        # rank — higher = more capacity, 16-64 is typical
+network_alpha = 16      # usually half of dim
+# optional: train only specific layers
+# network_args = ["conv_dim=16", "conv_alpha=8"]
--- a/workflows/anime_txt2img.json
+++ b/workflows/anime_txt2img.json
@@ -0,0 +1,120 @@
+{
+  "last_node_id": 9,
+  "last_link_id": 9,
+  "nodes": [
+    {
+      "id": 4,
+      "type": "CheckpointLoaderSimple",
+      "pos": [26, 474],
+      "size": [315, 98],
+      "flags": {},
+      "order": 0,
+      "mode": 0,
+      "outputs": [
+        {"name": "MODEL", "type": "MODEL", "links": [1], "slot_index": 0},
+        {"name": "CLIP", "type": "CLIP", "links": [3, 5], "slot_index": 1},
+        {"name": "VAE", "type": "VAE", "links": [8], "slot_index": 2}
+      ],
+      "properties": {"Node name for S&R": "CheckpointLoaderSimple"},
+      "widgets_values": ["noobai-xl.safetensors"]
+    },
+    {
+      "id": 6,
+      "type": "CLIPTextEncode",
+      "pos": [415, 186],
+      "size": [422, 164],
+      "flags": {},
+      "order": 2,
+      "mode": 0,
+      "inputs": [{"name": "clip", "type": "CLIP", "link": 3}],
+      "outputs": [{"name": "CONDITIONING", "type": "CONDITIONING", "links": [4], "slot_index": 0}],
+      "properties": {"Node name for S&R": "CLIPTextEncode"},
+      "widgets_values": ["1girl, anime, masterpiece, best quality, highly detailed, beautiful face, solo"]
+    },
+    {
+      "id": 7,
+      "type": "CLIPTextEncode",
+      "pos": [415, 400],
+      "size": [422, 164],
+      "flags": {},
+      "order": 3,
+      "mode": 0,
+      "inputs": [{"name": "clip", "type": "CLIP", "link": 5}],
+      "outputs": [{"name": "CONDITIONING", "type": "CONDITIONING", "links": [6], "slot_index": 0}],
+      "properties": {"Node name for S&R": "CLIPTextEncode"},
+      "widgets_values": ["worst quality, low quality, bad anatomy, extra limbs, poorly drawn face, ugly, watermark, text"]
+    },
+    {
+      "id": 5,
+      "type": "EmptyLatentImage",
+      "pos": [473, 609],
+      "size": [315, 106],
+      "flags": {},
+      "order": 1,
+      "mode": 0,
+      "outputs": [{"name": "LATENT", "type": "LATENT", "links": [2], "slot_index": 0}],
+      "properties": {"Node name for S&R": "EmptyLatentImage"},
+      "widgets_values": [832, 1216, 1]
+    },
+    {
+      "id": 3,
+      "type": "KSampler",
+      "pos": [863, 186],
+      "size": [315, 474],
+      "flags": {},
+      "order": 4,
+      "mode": 0,
+      "inputs": [
+        {"name": "model", "type": "MODEL", "link": 1},
+        {"name": "positive", "type": "CONDITIONING", "link": 4},
+        {"name": "negative", "type": "CONDITIONING", "link": 6},
+        {"name": "latent_image", "type": "LATENT", "link": 2}
+      ],
+      "outputs": [{"name": "LATENT", "type": "LATENT", "links": [7], "slot_index": 0}],
+      "properties": {"Node name for S&R": "KSampler"},
+      "widgets_values": [42, "fixed", 28, 7, "dpmpp_2m", "karras", 1]
+    },
+    {
+      "id": 8,
+      "type": "VAEDecode",
+      "pos": [1209, 186],
+      "size": [210, 46],
+      "flags": {},
+      "order": 5,
+      "mode": 0,
+      "inputs": [
+        {"name": "samples", "type": "LATENT", "link": 7},
+        {"name": "vae", "type": "VAE", "link": 8}
+      ],
+      "outputs": [{"name": "IMAGE", "type": "IMAGE", "links": [9], "slot_index": 0}],
+      "properties": {"Node name for S&R": "VAEDecode"}
+    },
+    {
+      "id": 9,
+      "type": "SaveImage",
+      "pos": [1451, 186],
+      "size": [210, 58],
+      "flags": {},
+      "order": 6,
+      "mode": 0,
+      "inputs": [{"name": "images", "type": "IMAGE", "link": 9}],
+      "properties": {"Node name for S&R": "SaveImage"},
+      "widgets_values": ["anime"]
+    }
+  ],
+  "links": [
+    [1, 4, 0, 3, 0, "MODEL"],
+    [2, 5, 0, 3, 3, "LATENT"],
+    [3, 4, 1, 6, 0, "CLIP"],
+    [4, 6, 0, 3, 1, "CONDITIONING"],
+    [5, 4, 1, 7, 0, "CLIP"],
+    [6, 7, 0, 3, 2, "CONDITIONING"],
+    [7, 3, 0, 8, 0, "LATENT"],
+    [8, 4, 2, 8, 1, "VAE"],
+    [9, 8, 0, 9, 0, "IMAGE"]
+  ],
+  "groups": [],
+  "config": {},
+  "extra": {"ds": {"scale": 0.8, "offset": [0, 0]}},
+  "version": 0.4
+}