UmeAiRT commited on
Commit
a4dcf32
·
verified ·
1 Parent(s): f3472c8

Add AGENTS.md

Browse files
Files changed (1) hide show
  1. AGENTS.md +45 -0
AGENTS.md ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # AGENTS.md — ComfyUI-Auto_installer-Assets
2
+
3
+ ## Purpose
4
+
5
+ Pure assets repository (HuggingFace LFS) for model files and installation binaries used by [ComfyUI-Auto_installer](https://github.com/UmeAiRT/ComfyUI-Auto_installer). No code — only large binary files.
6
+
7
+ ## Ecosystem
8
+
9
+ This repo is consumed by the **Auto_installer** project's PowerShell download scripts. Any file rename or move here **must** be reflected in the corresponding download scripts.
10
+
11
+ ## Folder Structure
12
+
13
+ | Folder | Contents | Notes |
14
+ |---|---|---|
15
+ | `models/diffusion_models/` | Backbone models (GGUF + safetensors) | Official ComfyUI folder. Organized by family (`FLUX/`, `WAN/`, etc.) |
16
+ | `models/text_encoders/` | Text encoders (T5, QWEN, GEMMA, LLAMA) | Official ComfyUI folder. **NOT** `clip/` |
17
+ | `models/clip/` | CLIP models only (clip_l, clip_g, ViT, longclip) | Legacy alias of `text_encoders/`. Do NOT put text encoders here |
18
+ | `models/checkpoints/` | Full merged checkpoints | |
19
+ | `models/loras/` | LoRA adapters | |
20
+ | `models/vae/` | VAE models (transverse, shared across families) | |
21
+ | `models/controlnet/` | ControlNet models | |
22
+ | `models/xlabs/` | XLabs custom node models | ⚠️ Hardcoded path — do NOT move |
23
+ | `models/onnx/` | ONNX upscalers for TensorRT | ⚠️ Hardcoded path — do NOT move |
24
+ | `bin/` | Installation executables (Git, Python, CUDA, etc.) | |
25
+ | `whl/` | Pre-compiled Python wheels | |
26
+ | `packages/` | Zip archives (extensions, build tools, configs) | |
27
+
28
+ ## Naming Conventions
29
+
30
+ ### GGUF files — Follow the [official GGUF spec](https://github.com/ggml-org/ggml/blob/master/docs/gguf.md)
31
+
32
+ Format: `<BaseName>-<SizeLabel>-<FineTune>-<Encoding>.gguf`
33
+ - **CamelCase** BaseName: `Flux1`, `Wan2.1`, `HiDream`, `LTXV`, `LTX-2`, `Qwen`
34
+ - **Uppercase** SizeLabel: `14B`, `1.3B`, `13B`
35
+ - **Uppercase** Encoding: `Q4_K_S`, `Q8_0`, `Q6_K`
36
+ - Hyphens between components
37
+
38
+ Examples: `Flux1-Dev-Q4_K_S.gguf`, `Wan2.2-T2V-HighNoise-14B-Q8_0.gguf`
39
+
40
+ ### Safetensors files — Community convention
41
+
42
+ - **Lowercase** with **hyphens** as separators
43
+ - Precision suffixes: `fp16`, `bf16`, `fp8_e4m3fn`, `fp8_scaled`
44
+
45
+ Examples: `wan2.1-t2v-14b-fp16.safetensors`, `qwen-image-edit-fp8_e4m3fn.safetensors`