Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,58 @@
|
|
| 1 |
---
|
| 2 |
license: creativeml-openrail-m
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: creativeml-openrail-m
|
| 3 |
+
library_name: diffusers
|
| 4 |
+
tags:
|
| 5 |
+
- stable-diffusion
|
| 6 |
+
- sd-1.5
|
| 7 |
+
- hyper-sd
|
| 8 |
+
- onnx
|
| 9 |
+
- text-to-image
|
| 10 |
+
- 4-step
|
| 11 |
+
- midjourney-style
|
| 12 |
+
base_model: prompthero/openjourney
|
| 13 |
+
pipeline_tag: text-to-image
|
| 14 |
---
|
| 15 |
+
|
| 16 |
+
# Openjourney + Hyper-SD (4-step) — ONNX
|
| 17 |
+
|
| 18 |
+
ONNX export of [prompthero/openjourney](https://huggingface.co/prompthero/openjourney) with the [ByteDance/Hyper-SD](https://huggingface.co/ByteDance/Hyper-SD) 4-step LoRA fused into the UNet. SD 1.5 architecture, 512×512 native, Euler scheduler, CFG = 1, **4 steps**.
|
| 19 |
+
|
| 20 |
+
Openjourney is prompthero's Midjourney-v4-style fine-tune. Activator phrase: `mdjrny-v4 style`. Pick this one when you want the painterly / illustrative "MJ look."
|
| 21 |
+
|
| 22 |
+
Converted artifact. Training credit: prompthero (Openjourney), ByteDance (Hyper-SD).
|
| 23 |
+
|
| 24 |
+
## What this repo contains
|
| 25 |
+
|
| 26 |
+
```
|
| 27 |
+
model_index.json
|
| 28 |
+
feature_extractor/
|
| 29 |
+
scheduler/
|
| 30 |
+
text_encoder/
|
| 31 |
+
tokenizer/
|
| 32 |
+
unet/ # Openjourney UNet + Hyper-SD-15 4-step LoRA fused in
|
| 33 |
+
vae_decoder/
|
| 34 |
+
vae_encoder/
|
| 35 |
+
```
|
| 36 |
+
|
| 37 |
+
## How it was produced
|
| 38 |
+
|
| 39 |
+
1. Load `prompthero/openjourney` via `diffusers`.
|
| 40 |
+
2. Fuse `ByteDance/Hyper-SD/Hyper-SD15-4steps-lora.safetensors`.
|
| 41 |
+
3. `optimum-cli export onnx`.
|
| 42 |
+
|
| 43 |
+
Toolchain: `optimum 1.24.0`, `diffusers 0.31.0`, `transformers 4.45.2`, `torch 2.4.x` (CUDA 12.4). Conversion script: [`scripts/export-openjourney-hyper.ps1`](https://github.com/HeliosophLLC/DatumIngest/blob/main/scripts/export-openjourney-hyper.ps1).
|
| 44 |
+
|
| 45 |
+
## Inference notes
|
| 46 |
+
|
| 47 |
+
| Setting | Value |
|
| 48 |
+
|---|---|
|
| 49 |
+
| Scheduler | Euler |
|
| 50 |
+
| Steps | 4 |
|
| 51 |
+
| CFG / guidance scale | 1.0 |
|
| 52 |
+
| Negative prompt | Skip |
|
| 53 |
+
| Resolution | 512×512 native |
|
| 54 |
+
| Activator | Include `mdjrny-v4 style` in prompts |
|
| 55 |
+
|
| 56 |
+
## License
|
| 57 |
+
|
| 58 |
+
CreativeML OpenRAIL-M (SD 1.5 + Openjourney + Hyper-SD). License files included. By using this model you accept those terms.
|